SAJCH Vol 10, No 1 (2016)

CHILD HEALTH THE SOUTH AFRICAN JOURNAL OF

March 2016

Volume 10

No. 1

• Mothers’ reasons for refusing to give consent to HIV testing on their infants • Audiological findings in neurologically compromised children • Haemophilia: A disease of women as well • Nutrition in the first 1 000 days of life in vulnerable communities • Analysis of paediatric poisoning • Childhood and adolescent fatalities at the Pretoria Medico-Legal Laboratory • Screening for retinopathy of prematurity in VLBW babies

CHILD HEALTH THE SOUTH AFRICAN JOURNAL OF

March 2016

Volume 10

No. 1

CONTENTS

Editorial

3 A global consensus on the prevention and management of nutritional rickets:

How does this change policy in South Africa?

EDITOR J M Pettifor FOUNDING EDITOR N P Khumalo EDITORIAL BOARD Prof. M Adhikari (University of KwaZuluNatal, Durban) Prof. M Kruger (Stellenbosch University) Prof. H Rode (Red Cross War Memorial Children's Hospital, Cape Town) Prof. L Spitz (Emeritus Nuffield Professor of Paediatric Surgery, London) Prof. A Venter (University of the Free State, Bloemfontein) Dr T Westwood (Red Cross War Memorial Children's Hospital, Cape Town) Prof. D F Wittenberg (University of Pretoria) HEALTH & MEDICAL PUBLISHING GROUP:

J M Pettifor

CEO AND PUBLISHER Hannah Kikaya

Short Reports

EXECUTIVE EDITOR Bridget Farham

N Shipalana, T S Ntuli

MANAGING EDITORS Ingrid Nye Claudia Naidu

5

Mothers’ reasons for refusing to give consent to HIV testing and the outcome in the children

8 A comparison of clinician and caregiver assessment of functioning in patients attending a child and adolescent mental health clinic in Nigeria

A O Okewole, M U Dada, M Bello-Mojeed, O C Ogun

Research

S Tayob, K Pillay, B Tlou, Y Ganie

12

Prevalence of positive coeliac serology in a cohort of South African children with type 1 diabetes mellitus

16 Effect of community integrated management of childhood illness on mothers’ healthcare-seeking behaviour and home management of childhood illness in Ile Ife, South-West Nigeria: A household survey

O A Ogundele, T Ogundele, O S Olajide, O I Agunbiade

20

Audiological findings in a group of neurologically compromised children: A retrospective study

K Baillieu, K Khoza-Shangase, L Jacklin

25 Risks for communication delays and disorders in infants in an urban primary healthcare clinic

D Claassen, J Pieterse, J van der Linde, E Kruger, B Vinck

29

Haemophilia: A disease of women as well

T Naicker, C Aldous, R Thejpal

33

Serious bacterial infections in febrile young children: Lack of value of biomarkers

M Karsas, P J Becker, R J Green

37 Selected facets of nutrition during the first 1 000 days of life in vulnerable

South African communities

L M du Plessis, M G Herselman, M H McLachlan, J H Nel

43 A 3-year survey of acute poisoning exposures in infants reported in telephone calls made to the Tygerberg Poison Information Centre, South Africa

C J Marks, D J van Hoving

47 Exploring sibling attitudes towards participation when the younger sibling has a severe speech and language disability

M Hansen, M Harty, J Bornman

52

Childhood and adolescent fatalities at the Pretoria Medico-Legal Laboratory: 2005 - 2009

G van den Ordel, L du Toit-Prinsloo, G Saayman

57

Factors present on admission associated with increased mortality in children admitted to a paediatric intensive care unit (PICU)

C L Hendricks, N H McKerrow, R J Hendricks

TECHNICAL EDITORS Emma Buchanan Paula van der Bijl NEWS EDITOR Chris Bateman PRODUCTION MANAGER Emma Jane Couzens DTP AND DESIGN Carl Sampson HEAD OF SALES AND MARKETING Diane Smith | Tel. 012 481 2069 Email: dianes@hmpg.co.za ONLINE SUPPORT Gertrude Fani | Tel. 072 463 2159 Email: publishing@hmpg.co.za FINANCE Tshepiso Mokoena HMPG BOARD OF DIRECTORS Prof. M Lukhele (Chair), Dr M R Abbas, Dr M J Grootboom, Mrs H Kikaya, Prof. E L Mazwai, Dr M Mbokota, Dr G Wolvaardt HEAD OFFICE Block F, Castle Walk Corporate Park, Nossob Street, Erasmuskloof Ext. 3, Pretoria, 0181 EDITORIAL OFFICE Suites 9 & 10, Lonsdale Building, Gardener Way, Pinelands, 7405 Tel. 021 532 1281 | Cell. 072 635 9825 Email: publishing@hmpg.co.za ISSN 1994-3032

Use of editorial material is subject to the Creative Commons Attribution – Noncommercial Works Licence. http://creativecommons.org/licenses/by-nc/3.0

63 The association between chronic undernutrition and malaria among Ethiopian children aged 6 - 59 months: A facility-based case-control study

H Y Hassen, J H Ali

68

Epidemiology of paediatric poisoning reporting to a tertiary hospital in Ghana

D Ansong, C Nkyi, C O Appiah, E X Amuzu, C A Frimpong, I Nyanor, S B Nguah, J Sylverken

71 An exploratory study of the implementation of early intervention workshops for primary caregivers in Johannesburg

S Medhurst, S Abdoola, L Duncan

75 Usefulness of ultrasonography and biochemical features in the diagnosis of cholestatic jaundice in infants

M S Choopa, C Kock, S O M Manda, A J Terblanche, D F Wittenberg

79 An evaluation of the screening for retinopathy of prematurity in very-low-birth-weight babies at a tertiary hospital in Johannesburg, South Africa

Z Dadoo, D E Ballot

83

Perceptions of community-based human milk banks before and after training in a resource-limited South African setting

H E Goodfellow, P Reimers, K Israel-Ballard, A Coutsoudis

87

Prevalence and nature of communication delays in a South African primary healthcare context

J van der Linde, D W Swanepoel, J Sommerville, F Glascoe, B Vinck, E M Louw

92 An analysis of national data on care-seeking behaviour by parents of children with suspected pneumonia in Nigeria

M B Abdulkadir, Z A Abdulkadir, W B R Johnson

Case Reports

T Padayachi, I E Haffejee

96

Too much diarrhoea, too many infections, and too few neutrophils

99 Ecthyma gangrenosum caused by Stenotrophomonas maltophilia in a neutropenic leukaemic infant: A case report

D K Das, S Shukla

101

CPD Questions

apers for publication should be addressed to the Editor, P via the website: www.sajch.org.za Tel: 072 635 9825 E-mail: publishing@hmpg.co.za

Cover: 'Child's drawing' by Aleksander Mijatovic

ŠCopyright: Health and Medical Publishing Group (Pty) Ltd

EDITORIAL

A global consensus on the prevention and management of nutritional rickets: How does this change policy in South Africa?

Globally, rickets secondary to vitamin D deficiency and/or low dietary calcium intake remains a significant public health problem, not only in resource-poor countries but also in those that are better resourced, despite effective means being available for its prevention and treat­ ment. The disease may have both short- and long-term consequences for the affected child, the latter frequently persisting into adulthood.

Although there have been renewed scientific interest and focus on the role of vitamin D status in health and disease, the same cannot be said for the implementation of public health guidance for the eradication of rickets due to vitamin D deficiency and/or low dietary calcium intake in our most vulnerable populations. Recently, there have been a number of international initiatives to address these issues.[1] At the beginning of this year, a consensus group comprising representatives from paediatric endocrine societies, nutrition societies and individual experts published a consensus statement on the prevention and management of nutritional rickets in the Journal of Clinical Endocrinology and Metabolism[2] and Hormone Research in Paediatrics.[3] This consensus statement was the culmination of a review of the literature by five task teams and a three-day workshop held in 2014. It is hoped that this statement will be the beginning of a concerted effort to convince relevant international organisations and national governments that the eradication of nutritional rickets is not only feasible but also of considerable benefit to child health globally.

At the outset, it should be emphasised that the recommendations pertain only to the prevention and treatment of nutritional rickets and do not attempt to address other possible clinical effects of a poor vitamin D status. It is not possible in this editorial to highlight all the recommendations made by the group, but there are a number of recommendations that are central to the global prevention of rickets that need to be emphasised: • Vitamin D deficiency is defined by a 25-hydroxyvitamin D (25OHD) of <30 nmol/L. • For children >12 months of age, a dietary calcium intake of <300 mg/ day increases the risk of rickets independent of serum 25OHD levels.

• All infants <12 months of age should be supplemented with vita­ min D 400 IU/day, irrespective of the mode of feeding.

• Children >12 months of age should be considered for supple­ mentation (600 IU/day) if they are at risk of vitamin D deficiency by being a member of a high-risk community or having social or environmental factors that reduce their access to skin synthesis of vitamin D or to dietary vitamin D intake.

• In healthy children, routine 25OHD screening is not recommended and therefore there is no specific threshold for vitamin D supple­ mentation in this population.

• All pregnant women should be supplemented with vitamin D 600 IU/ day to reduce the risk of neonates being born with a low 25OHD concen­ tration, but evidence does not suggest that dietary calcium require­ ments are increased during pregnancy and lactation.

• Fortification of staple foods should be considered as a means of improving the vitamin D status of populations.

The most important question to be asked by South African (SA) child health professionals is how do these recommendations relate to the situation in SA? Certainly, what little information there is available would suggest that the recommendations are very relevant. Firstly, despite the good average daily hours of sunshine that most areas of the country experience throughout the year, vitamin D production 3

SAJCH

through ultraviolet radiation is limited in the southern-most regions during the autumn and winter months from April through September.[4] Secondly, symptomatic vitamin D deficiency (mainly hypocalcaemia) is not an uncommon presentation in young infants at Chris Hani Baragwanath Academic Hospital in Johannesburg, and severe rickets and bone deformities are frequently seen in toddlers and young children living in the overcrowded high-rise buildings of the inner city region of Johannesburg (unpublished data). Clear seasonal variations in circulating 25OHD have been reported in adults from both Johannesburg[5] and CapeTown,[6] and the prevalence of vitamin D deficiency has been reported to be high in Indian females living in Johannesburg[5] and in blacks living in Cape Town.[6] Because of the high prevalence of rickets in infants in Cape Town and to a lesser extent in Johannesburg in the mid-1900s, vitamin D supplementation (400 IU/day) of all infants during the first 12 months of life was routine practice in primary healthcare clinics; however, this practice was later discontinued. It is clear from the renewed relatively common occurrence of symptomatic vitamin D deficiency in young infants in Johannesburg that the practice of vitamin D supplementation of all infants should be re-initiated. Furthermore, toddlers and young children in the inner city areas of Johannesburg should be considered as high risk and, therefore, supplementation should be continued for this group at 600 IU/day. This group of high-risk children is also at risk of dietary calcium deficiency, which aggravates and accentuates the risk of nutritional rickets. Although there is limited information on the vitamin D status of pregnant mothers in SA, indications from a study at Chris Hani Baragwanath Academic Hospital are that the prevalence of vitamin D deficiency is high (Velaphi S, unpublished data); therefore, attention should be given to providing all mothers with a vitamin D supplement of 600 IU/day during pregnancy to ensure that infants at birth are born vitamin D replete.

In conclusion, the guidelines recommend that all infants in SA should receive vitamin D supplements, while in the older age groups, those who are at high risk of rickets should be targeted. All pregnant mothers should also receive vitamin D as part of their micronutrient supplements.

John M Pettifor MB BCh, FCPaed (SA), PhD (Med), MASSAf Editor, South African Journal of Child Health References 1. Schoenmakers I, Pettifor JM, Pena-Rosas JP, et al. Prevention and consequences of vitamin D deficiency in pregnant and lactating women and children: A symposium to prioritise vitamin D on the global agenda. Steroid Biochem Mol Biol 2015;pii:S0960-0760(15)30130-8. [http://dx.doi.org/10.1016/j. jsbmb.2015.11.004] 2. Munns CF, Shaw N, Kiely M, et al. Global consensus recommendations on prevention and management of nutritional rickets. J Clin Endocrinol Metab 2016:101(2):394-415. [http://dx.doi.org/10.1210/jc.2015-2175] 3. Munns CF, Shaw N, Kiely M, et al. Global consensus recommendations on prevention and management of nutritional rickets. Horm Res Paediatr 2016;85(2):83-106. [http://dx.doi.org/10.1159/000443136] 4. Pettifor JM, Moodley GP, Hough FS, et al. The effect of season and latitude on in vitro vitamin D formation by sunlight in South Africa. S Afr Med J 1996;86(10):1270-1272.

MARCH 2016 Vol. 10 No. 1

EDITORIAL 5. George JA, Norris SA, van Deventer HE, Pettifor JM, Crowther NJ. Effect of adiposity, season, diet and calcium or vitamin D supplementation on the vitamin D status of healthy urban African and Asian-Indian adults. Br J Nutr 2014;112(4):590-599. [http://dx.doi.org/10.1017/50007114514001202] 6. Martineau AR, Nhamoyebonde S, Oni T, et al. Reciprocal seasonal variation in vitamin D status and tuberculosis notifications in Cape Town, South Africa.

4

Proc Natl Acad Sci USA 2011;108(47):19013-19017. [http://dx.doi.org/10.1073/ pnas.1111825108]

S Afr J Child Health 2016;10(1):3-4. DOI:10.7196/SAJCH.2016.v10i1.1165

SAJCH MARCH 2016 Vol. 10 No. 1

SHORT REPORT

Mothers’ reasons for refusing to give consent to HIV testing and the outcome in the children N Shipalana,1 MB ChB, FCPaed (SA), MMed (Paed); T S Ntuli,2 BSc, BSc Hons, MSc 1

Department of Paediatrics, University of Limpopo (Polokwane Campus), Polokwane, South Africa Research Development and Administration, University of Limpopo (Turfloop Campus); and Department of Public Health Medicine, University of Limpopo (Polokwane Campus), Polokwane, South Africa

2

Corresponding author: N Shipalana (shipalanan@vodamail.co.za) Background. HIV/AIDS is one of the most common underlying causes of death in children between the ages of 3 months and 5 years in sub-Saharan Africa. In Limpopo Province, South Africa, the prevention of mother-to-child transmission (PMTCT) programme introduced in early 2000 and paediatric ARV roll-out have had a poor uptake due to various factors.

Objective. To establish the reasons why mothers decline HIV testing for their children.

Methods. A cross-sectional descriptive study was conducted at the paediatric ward, Mankweng Hospital, Limpopo, for a period of 1 year (June 2009 - June 2010). All mothers who had declined HIV testing on their children were requested to participate. All the participants gave informed consent.

Results. A total of 30 mothers participated. All women had attended antenatal care, 28 (93%) stated that their HIV results were negative and 2 (8%) had undergone PMTCT. The reasons mothers refused HIV testing on their children included the following: did not want to be stressed with a positive result (67%), did not want to know their status (7%) and could not consent as their partners had declined tests on both baby and mother (7%); 20% had other reasons including fear of HIV stigma. The median age of the children was 13 months (interquartile range 2 months - 10 years). Twenty-one (70%) of children were discharged home after treatment without HIV testing, five (16%) mothers signed refusal of hospital treatment, three (12%) started ARV after the mother reconsidered and signed consent, with good response to highly active antiretroviral treatment (HAART) and one child died after a month in the hospital.

Conclusion. Fear of being stressed by a positive result was the main reason mothers refused an HIV test on their children. Better education about HIV transmission, prevention and the good response to HAART is needed to increase the uptake of HIV testing and antiretroviral therapy.

S Afr J Child Health 2016:10(1):5-7. DOI:10.7196/SAJCH.2016.v10i1.952

The World Health Organization (WHO) estimated global reduction of under-five deaths from 12.6 million in 1990 to 6.6 million (47%) in 2012.[1] Most countries have shown good progress by reducing the under-five mortality rate by more than half since 1990. HIV/AIDS is the most common cause of morbidity and mortality in under-five children in sub-Saharan Africa.[2-6] A greater proportion of these children get infected through mother-to-child transmission. Studies have demonstrated that early diagnosis and treatment can reduce morbidity and mortality among these children.[7-11] Prevention of mother-to-child transmission (PMTCT) pro­ grammes have made it possible to reduce the risk of infection. [12- 16] In South Africa (SA), the PMTCT guidelines recommend that diagnostic testing in infants be performed as early as 4 - 6 weeks of age, so that highly active antiretroviral treatment (HAART) is initiated early, thereby reducing infant deaths.[16] Although PMTCT has been successful, a high proportion of HIV-exposed and HIV-infected infants remain unidentified.[15] In SA, provinces with better health resources have a higher intake of children on HIV/AIDS programmes. Hsiao et al.,[16] in their study conducted in the Western Cape Province of SA, illustrated an increase in the proportion of infected infants successfully linked to HIV care and treatment.[16] HIV/AIDS is a chronic disease that needs the full understanding and cooperation of a parent and/or guardian to take the child for lifelong treatment. Children with signs of HIV/AIDS need the consent of a parent and/or guardian to perform specific screening or tests for HIV. However, women are still reluctant to give consent for testing on their sick children. The objective of this study is to investigate the reasons why mothers decline HIV testing for their children in a tertiary hospital.

5

SAJCH

Methods

A cross-sectional descriptive study was conducted over a period of 12 months from June 2009 to June 2010 at the paediatric ward of Mankweng Hospital in Limpopo Province, SA. The hospital is a combined district/regional hospital for the local population and a tertiary referral centre, with 46 beds in the paediatric ward. Children under 13 years of age with illness and requiring hospitalisation are admitted to the paediatric ward from the outpatient and emergency departments. The study included all mothers who had declined HIV testing on their children.

During the study period, the practice in the ward was to do HIV testing on all children that had clinical stigma of HIV infection, acute or chronic medical conditions that mimic HIV infection or history of HIV exposure, as well as those whose mothers requested voluntary testing. Where HIV testing was requested by a health worker, the mother was given pretest counselling by an HIV councillor or occasionally by a registered nurse or doctor, especially when she declined the test. The mothers who refused HIV testing on their children were also given pretest counselling.

Ethics approval for this study was obtained from the University of Limpopo Ethics Committee, Polokwane Campus (Ref: 033/2008). Anonymity and confidentiality of the participants’ personal information were protected, and the participants signed informed consent forms before participating in the study. The data for the study were collected by the paediatrician main researcher, and included the mother’s age, education, marital and employment status, antenatal care (ANC) visits, voluntary counselling and testing (VCT) results, PMTCT (prevention of mother-to-child transmission), reasons for refusal of an HIV test on their children, baby’s age and outcome of the child

MARCH 2016 Vol. 10 No. 1

SHORT REPORT during the current admission. Categorical data were displayed as percentages. Statistical software (STATA 9.0; Stata Corp, USA) was used for data analysis.

Table 1. Outcome of the children (N=30)

Results

Discharged home after treating admission illness – no HIV test

21 (70)

Reconsidered and signed consent for HIV testing and started HAART

3 (10)

Died in hospital without knowing HIV status

1 (3)

Mothers signed RHT

5 (17)

A total of 1 563 children were admitted during the 12-month period of the study. Mothers who were offered testing according to the testing policy of the ward at that time totalled 241. Two hundred and eleven (88%) signed consent for HIV testing of their children and 30 (12%) refused.

Of the 30 mothers who declined HIV testing on their children, 28 (96%) were in the age group 18 - 35 years, 27 (93%) had attempted or passed grade 12, 13 (43%) were single mothers and 18 (60%) were unemployed. All 30 mothers had attended ANC. Most (n=28, 96%) stated that their HIV results were negative, and only two (8%) had undergone PMTCT. Verification of the HIV results was not possible as most ANC HIV tests are rapid tests with no hard copies given to clients and documentation of the results in the child’s Road to Health Card (RHC) was not always done. The median age of the children was 13 months (interquartile range (IQR) 2 months - 10 years).

With regard to the children’s outcomes, 21 (70%) were discharged home after treatment of the admission illness without confirmation of their HIV status, 3 (10%) started antiretrovirals (ARVs) with good response to HAART after the mother reconsidered and signed consent, and 1 (3%) child died after a month in the hospital. For the remaining 5 (17%) children, their mothers signed refusal of hospital treatment (RHT) and they left the hospital before being well enough to be discharged (Table 1).

Reasons mothers refused HIV testing on their children included the following: did not want to be stressed with a positive result (67%), did not want to know their status (7%), and could not consent as their partners had declined tests on both baby and mother (7%). The other reasons for refusal are varied due to the open-ended design of the questionnaire (Table 2).

Discussion

This study explored why some mothers refuse HIV testing for their sick children. The SA national strategic plan for HIV and sexually transmitted infections is to expand access to care, treatment and support (80%) of people infected with HIV.[17] Women and children are the most vulnerable groups that need special attention. HIV/ AIDS is a leading factor that has contributed to the rising under-5 morbidity and mortality rates and had made it impossible for SA to achieve Millenium Development Goal 4 by 2015.[3,5,18] The roll-out of PMTCT has made it easy for pregnant women to access voluntary counselling and testing (VCT) at their local clinics.[14,16] However, studies have shown that healthcare providers experienced challenges of women unwilling to consent to HIV testing and treatment.[20-22] In our study, all mothers attended ANC, showing that routine ANC is acceptable, and this opportunity can be used to significantly increase HIV testing rates. However, while 96% of women in this study had VCT, their HIV results could not be verified as HIV testing at most ANC sites is done by rapid testing and results are recorded in the mother’s antenatal records; no hard copy is given to the client.

Moodley et al.[22] in their study reported that up to 84% of pregnant women had received HIV counselling and testing, and ~90% of them had been tested for HIV. Khan et al.[23] carried out an audit of HIV information on the RHC at rural clinics in Limpopo Province. Their study found that 62% of the pregnant women had had their HIV status documented on their RHC.[23] Despite the high rate of antenatal visits during pregnancy, some barriers for VCT/HIV testing still exist. Studies have shown that the stigmatisation of HIV/AIDS, overexaggeration of HIV treatment side-effects and women’s worry about their partner’s reaction if they find out they had tested positive can contribute to unwillingness to have HIV testing and treatment. [21,22] 6

SAJCH

n (%)

Table 2. Reasons mothers refuse HIV testing (N=30) n (%) Does not want to be stressed by a positive result

20 (67)

Knows about HIV transmission but doesn't want to know status

2 (7)

Partner has declined HIV testing

2 (7)

HIV stigma

1 (3)

Fear of knowing long-term effects of disease on their baby

1 (3)

Fear of feeling guilty if the child was positive

1 (3)

In the process of divorce

1 (3)

Fear of losing hope if child found HIV-positive

1 (3)

Father of child had just died

1 (3)

Recently, a study conducted at Mankweng primary healthcare (PHC) facilities in Limpopo revealed that the clients who attended these clinics lacked knowledge regarding VCT, the prevention of HIV infection and support systems available to HIV-positive people.[25] There are few studies that assessed the reasons why mothers refused HIV testing for their babies. In high-income countries, mothers who refused HIV testing for their children have cited various reasons, including: the perception that a physically well child cannot be infected with HIV, inability to cope with a positive diagnosis in a child, and feeling of guilt if the child tested positive. [26,27] In the present study, the main reason mothers refused HIV testing of their children was that they did not want to be stressed by the positive HIV result of the child.

The involvement of partners in HIV prevention and treatment has been investigated previously in sub-Saharan Africa. A systematic review by the University of Tampere, Finland, showed that men had a positive attitude towards PMTCT programmes even though some barriers existed, such as lack of knowledge or time, stigma of HIV/ AIDS, problem with healthcare services and cultural issues.[29] The Child Care Act states that the Minister of Social Development can give consent for HIV testing in situations where parents refuse consent or cannot be found.[30] Although healthcare workers have the power to consult and make use of this recommendation, it is not advisable to take this route immediately and do HIV testing where the parent has not given consent, as HIV/AIDS is a chronic disease that needs the cooperation of the parent/guardian to enable continuous treatment of the child. Providing the mother with education and support could help tackle some of the barriers raised in this study.

Study limitations

The major limitation of this study was the small sample size. Another limitation was that while most mothers had had HIV VCT, their HIV status could not be verified as they did not have proof of their test results, which would normally be recorded in their antenatal records or the children’s RHCs.

MARCH 2016 Vol. 10 No. 1

SHORT REPORT Conclusion

The study explored why mothers refuse HIV testing for their children admitted to a paediatric ward with an acute or chronic illness. Fear of being stressed by a positive result was the most common reason cited for refusal to give consent for HIV testing. The results of our study suggest that better parent/guardian education about HIV transmission and prevention, and the effectiveness of HAART, is needed to increase the uptake of HIV testing.

Acknowledgements. We thank the staff of the paediatric ward for assisting during the data collection, and the mothers of the children admitted in the ward for their cooperation during this study. In addition, thanks to Dr R Muloiwa for providing useful comments and suggestions.

References 1. The United Nations Inter-agency Group for Child Mortality Estimation. Levels and Trends in Child Mortality. Report 2013. New York: UNICEF, 2013. http:// www.unicef.org (accessed 21 February 2014).

2. Violari A, Cotton MF, Gibb DM, et al. Early antiretroviral therapy and mortality among HIV-infected infants. N Engl J Med 2008;359(21):2233-2244. [http:// dx.doi.org/10.1056/NEJMoa0800971] 3. Newell M-L, Coovadia H, Cortina-Borja M, et al. Mortality of infected and uninfected infants born to HIV-infected mothers in Africa: A pooled analysis. Lancet 2004;364(9441):1236-1243. [http://dx.doi.org/10.1016/S01406736(04)17140-7] 4. De Martino M, Tovo PA, Balducci M, et al. Reduction in mortality with availability of antiretroviral therapy for children with perinatal HIV-1 infection.

JAMA 2000;248(2):2871-2872. [http://dx.doi.org/10.1001/jama.284.2.190] 5. Van Deventer JD, Carter CL, Schoeman CJ, Joubert G. The impact of HIV on the profile of paediatric admissions and deaths at Pelonomi Hospital, Bloemfontein, South Africa. J Trop Paediatr 2005;51(6):391-392. [http://dx.doi. org/10.1093/tropej/fmi038] 6. Ntuli ST, Malangu N, Alberts M. Causes of deaths in children under-five years old at a tertiary hospital in Limpopo Province of South Africa. Glob J Health Sci 2013;5(3):95-100. [http://dx.doi.org/10.5539/gjhs.v5n3p95] 7. Penazzato M, Prendergast A, Tierney J, Cotton M, Gibb D. Effectiveness of antiretroviral therapy in HIV-infected children under 2 years of age. Cochrane Database Syst Rev 2012;7:CD004772. [http://dx.doi.org/10.1002/14651858.

CD004772.pub3] 8. Boyle DS, Lehman DA, Lillis L, et al. Rapid detection of HIV-1 proviral DNA for early infant diagnosis using recombinant polymerase amplification. MBio 2013;4(2):pii: e00135-13. [http://dx.doi.org/10.1128/mBio.00135-13] 9. Wamalwa DC, Farquhar C, Obimbo EM, et al. Early response to highly active antiretroviral therapy in HIV-1-infected children. J Acquir Immune Defic Syndr 2007;45(3):311-317. [http://dx.doi.org/10.1097/QAI.0b013e318042d613] 10. Kilewo C, Karlsson K, Ngarina M, et al. Prevention of mother-to-child transmission of HIV-1 through breastfeeding by treating mothers with triple antiretroviral therapy in Dar es Salaam, Tanzania: The Mitra Plus study. J Acquir Immune Defic Syndr 2009;52(3):406-416. [http://dx.doi.org/10.1097/ QAI.0b013e3181b323ff] 11. Thior I, Lockman S, Smearton LM, et al. Breastfeeding plus infant zidovudine prophylaxis for 6 months v. formula feeding plus zidovudine for 1 month to reduce mother-to-child transmission in Botswana: A randomized trial: The Mashi study.

JAMA 2006;296(7):794-805. [http://dx.doi.org/10.1001/jama.296.7.794]

7

SAJCH

12. Wortley PM, Lindegren ML, Fleming PL. Successful implementation of perinatal HIV prevention guidelines. A multistate surveillance evaluation.

MMWR Recommendation Rep 2001;50(RR-6):17-28.

13. Siegfried N, van der Merwe L, Brocklehurst P, Sint TT. Antiretrovirals for reducing the risk of mother-to-child transmission of HIV infection. Cochrane Database Syst Rev 2011;(7):CD003510. [http://dx.doi.org/10.1002/14651858.CD003510.pub3] 14. National Department of Health, South Africa. Guidelines for the Management of HIV in Children. Pretoria, South Africa: National Department of Health, 2010. http://familymedicine.ukzn.ac.za/Libraries/Guidelines_Protocols/2010_ Paediatric_Guidelines.sflb.ashx (accessed 17 February 2014).

15. Lillian RR, Kalk E, Technau KG, Sherman GG. Birth diagnosis of HIV infection in infants to reduce infant mortality and monitor for elimination of motherto-child transmission. Paediatric Infect Dis J 2013;32(10):1080-1085. [http:// dx.doi.org/10.1097/INF.0b013e318290622e] 16. Hsiao NY, Stinson K, Myer L. Linkage of HIV-infected infants from diagnosis to antiretroviral therapy services across the Western Cape, South Africa. PLoS One; 8(2):e55308. [http://dx.doi.org/10.1371/journal.pone.0055308. Pub 2013 Feb 6] 17. National Department of Health, South Africa. National Strategy Plan on HIV, STIs and TB 2012-2016. 2011. http://www.hst.org.za (accessed 17 February 2014).

18. Draper B, Abdullah F. A review of the prevention of mother-to-child transmission programme of the Western Cape provincial government, 20032004. S Afr Med J 2008;98(6):431-434.

19. Kumbi S, Bedri A, Abashawl A, Isehak A, Coberly JS, Ruff AJ. Reasons for refusal of HIV testing in two Ethiopian antenatal clinics. Abstract (No. 4970:4970), International AIDS Society, Poster Exhibition: The XIV International AIDS Conference.

20. Rosa H, Goldani MZ, Scanlon T, et al. Barriers for HIV testing during pregnancy in Southern Brazil. Rev Saúde Pública 2006;40(2):220-225.

21. Pool R, Nyanzi S, Whitworth JA. Attitude of pregnant women towards VCT in rural south-west Uganda. AIDS Care 2001;13(5):605-615. [http://dx.doi. org/10.1080/09540120120063232] 22. Moodley D, Srikewal J, Msweli L, Maharaj NR. A bird’s eye view of PMTCT coverage at two regional hospitals and their referral clinics in a resource-limited setting. S Afr Med J 2011;101(2):122-125.

23. Khan RBI, Robertson BA, Railton J, Gani S, Tsolo F. An audit of the HIV information documented on the road-to-health cards of children attending primary health care clinics in Capricorn district of Limpopo Province. S Afr Paediatr Rev 2012;9(1):4-8.

24. Ramoraswi MM, Lekhuleni ME, Mbambo-Kekane NP, Mpolokeng M. Knowledge of clients regarding voluntary counselling and testing at Mankweng primary health care facilities at Capricorn District, Limpopo Province, South Africa. Afr J Phys Health Educ Recreation Dance 2013;19(4):62-74.

25. Eisenhut M, Kawsar M, Connan M, Balachandran T. Why are HIV-positive mothers refusing to have their children screened for vertically transmitted HIV infection?

Int J STD AIDS 2009;20(7):506-507. [http://dx.doi.org/10.1258/ijsa.2008008508] 26. Andrews S, Handyside R, Carpenter L, Price A, Majewska W, Prime K. Testing children of mothers with HIV infection: Experience in three south-west London HIV clinics. HIV Med 2012;13(2):138-140. [http://dx.doi.org/10.1111/ j.1468-1293.2011.00948.x] 27. Hussain R, McMaster P. An audit on HIV testing of children born to HIV positive mothers. Arch Dis Child 2011;95(7):569. [http://dx.doi.org/10.1136/ abc.2009.175778] 28. Auvinen J, Kylmä J, Suominen T. Male involvement and prevention of motherto-child transmission of HIV in sub-Saharan Africa: An integrative review.

Curr HIV Res 2013;11(2):169-177.

29. The South African Medical Association. Human Rights and the Ethical Guidelines on HIV and AIDS 2006: A Manual for Medical Practitioners. https://www.samedical.org (accessed 21 February 2014).

30. McQuoid-Mason D. Routine testing for HIV ethical and legal implications. S Afr Med J 2007:97(6):416.

MARCH 2016 Vol. 10 No. 1

SHORT REPORT

A comparison of clinician and caregiver assessment of functioning in patients attending a child and adolescent mental health clinic in Nigeria A O Okewole,1 MBBS, FWACP; M U Dada,2 MB ChB, FWACP; M Bello-Mojeed,3 MBBS, FWACP; O C Ogun,3 MBBS, FWACP Department of Clinical Services, Neuropsychiatric Hospital, Aro Abeokuta, Nigeria Department of Psychiatry, Ekiti State University Teaching Hospital, Ekiti, Nigeria 3 Clinical Services Department, Federal Neuropsychiatric Hospital, Yaba Lagos, Nigeria 1 2

Corresponding author: A O Okewole (niranokewole@gmail.com) Objective. To compare clinician and caregiver assessments of functioning and impairment among children attending a specialist mental health facility.

Methods. Caregivers of patients attending the Harvey Road Child and Adolescent Centre (N=155) were consecutively recruited over a 1-month period. The caregivers were requested to fill in a sociodemographic questionnaire, the Columbia Impairment Scale (CIS), the 12- item version of the General Health Questionnaire (GHQ-12) and the Zarit Burden Interview. Scoring on the Children’s Global Assessment Scale (CGAS) was done by clinicians.

Results. The mean (standard deviation (SD)) age of patients and caregivers was 12.3 (4.9) years and 41.5 (8.9), respectively. Mean (SD) scores on the CGAS and CIS were 55.3 (22.8) and 16.4 (14.3), respectively, while mean scores on the Zarit Burden Interview and the GHQ were 27.9 (17.2) and 2.5 (2.6), respectively. Poorer clinician-rated functioning was associated with lower educational level of the child, a main diagnosis other than seizure disorder, longer duration of illness, and presence of comorbidity. Worse caregiver rating of impairment was equally associated with lower child education, and main diagnosis other than seizure disorder. Both CGAS and CIS scores showed significant correlation with Zarit and GHQ scores. Finally, a significant correlation was found between CGAS and CIS scores (r=−0.388, p<0.001).

Conclusion. Children attending specialist neuropsychiatric services have functional impairment which may be associated with lower child education, and which is linked with increased caregiver burden and psychiatric morbidity.

S Afr J Child Health 2016;10(1):8-11. DOI:10.7196/SAJCH.2016.v10i1.966

Previously, psychopathology has been viewed either from a categorical or a dimensional perspective. These have evolved into the multidimensional approach now favoured by the major classification systems in psychiatry.[1] A key component of the multidimensional classification is the assessment of impairment, functioning and disability. Functional impairment, often a consequence of psychiatric symptoms, represents a reduced level of social, educational and/or occupational domains of life.[2] It is argued that functional impairment should be viewed as a separate entity from the basic psychopathology.[3] The World Health Organization (WHO) in 2001 adopted an International Classification of Functioning, Disability and Health (ICF), which describes human functioning in terms of body structure, function, activities and participation. These are aspects of human function which influence, and are influenced by, state of health and environmental and personal factors.[4,5] The ICF was adopted for use in children and adolescents in 2007.[5,6] Functioning and impairment have become central to our clinical decision-making. However, there is relatively little research on how well their impact can be measured.[2] Rating scales for assessment of impairment are few in number and rarely used in clinical practice.[7] One commonly used tool is the Children’s Global Assessment Scale (CGAS)[8-10] which is a clinician-administered scale. In a society where expertise in child and adolescent mental health is rare, it is often necessary to rely on the caregiver for ratings of patient illness, response to treatment, and outcome. The caregiver, being more often in contact with the patient, can provide extremely valuable information on the patient’s level of functioning. The extent to which the caregiver’s report agrees with the clinician’s rating is however often unclear.

In developing countries like Nigeria, there is a dearth of data on the level of functional impairment in children requiring services. [8] This study therefore aimed to compare clinicians’ and caregivers’ 8

SAJCH

assessments of functioning among children and adolescents attending a specialist psychiatric clinic.

Methods

A convenience sample of caregivers of children attending the Harvey Road Child and Adolescent Centre, Lagos (N=155) were recruited over a 1-month period. This sample was taken from a pool of an average of 400 patients who visit the centre monthly: two clinics are run per week, with about 50 children seen at each clinic. The children seen at the clinic have a variety of diagnoses including autism spectrum disorders (ASD), attention deficit and hyperactivity disorder (ADHD), psychotic disorders, and intellectual disability. However, a large proportion of children seen have seizure disorder either as the main diagnosis or as a comorbidity with another disorder. Despite there being at least two paediatric neurology clinics in the Lagos metropolis, caregivers often prefer to use the Child and Adolescent Clinic because of the circumscribed nature of the service.

To meet this demand, trainee psychiatrists are required to do neurology postings as part of their training, and in certain cases referrals are advised.

Instruments

A sociodemographic questionnaire was completed by the caregiver, which contained information about the child and caregiver. Some additional clinical details were obtained from the hospital records and filled in by the clinician.

The CGAS[11] is a clinician rating of functioning in the child and adolescent population. It has been found to have good reliability and validity[12] and has been used in a previous study among Nigerian children.[8] The CGAS rates children on a scale of 1 - 100, and is divided into ten categories. Higher scores on the CGAS are associated with better functioning.

MARCH 2016 Vol. 10 No. 1

SHORT REPORT The Columbia Impairment Scale (CIS) [13] is a 13-item self-rated questionnaire for assessing the caregiver’s subjective impression of im­ pair­ ment in an ill child. Higher scores on the CIS are associated with poorer func­tion­ ing. The items assess four key areas of func­ tioning: interpersonal relations; broad psycho­ pathological domains (e.g. anxiety, depression or behaviour problems); functioning in job or school; and use of leisure time.[1,13] While it has proven validity in Caucasian populations, no available studies of its validity were found in this environment.

The burden of caring for the patients was assessed among caregivers using the Zarit Burden Interview.[14] Psychiatric morbidity among the caregivers was assessed using the 12-item General Health Questionnaire (GHQ-12).[15] The Lagos metropolis has residents from several of the ethnic groups which make up Nigeria. Translation of the study instruments into any particular dialect was not considered beneficial. The instruments were therefore administered in English.

Procedure

Subjects were recruited during routine clinic visits. Informed consent was duly sought, with verbal assent from the children where appropriate. Scoring on the CGAS was done by senior residents in psychiatry. The CGAS is part of routine patient assessment and no formal training was required. The caregivers were requested to fill in a sociodemographic questionnaire, the GHQ-12, the Zarit Burden interview and the CIS. Subjects who did not understand any particular item had such items explained to them by the investigators, and in some cases the items were read out to them.

Ethical considerations

Ethical approval was obtained from the Ethical Committee of the Federal Neuro­ psych­iatric Hospital, Yaba. Informed consent was obtained from all subjects after the study protocol had been explained to them. Verbal assent was also obtained from children who were considered old enough.

Statistical analysis

Results were calculated as frequencies. Group comparisons were done using χ2 tests, correlations, independent sample t-tests and analysis of variance where appropriate. Tests were two-tailed, and the level of significance was set at p<0.05.

Results

A total of 155 caregivers of children with neuro­ psychiatric disorders participated in this study.

The caregivers’ ages ranged from 19 to 65 years, mean (standard deviation (SD)) age being 41.5 (8.9) years. The mean age of the children was 12.3 (4.98) years. Other sociodemographic

and clinical variables of the children and their caregivers are presented in Table 1. Among the children, there was a slight male predominance, and more than half were in primary school or at a lower level (nursery school, kindergarten, or no formal education). Nearly half had been ill for more than 5 years, and in over half the main diagnosis was seizure disorder, while more than a third had a comorbid condition.

The majority of the caregivers were female (71% being mothers), employed, married and of the Christian religion.

The mean score on the CGAS completed by clinicians was 55.3 (22.8). The mean scores on the CIS, the Zarit Burden interview and the GHQ-12, which were completed by the caregivers, were 16.4 (14.3), 27.9 (17.2) and 2.5 (2.6), respectively. Table 2 presents the associations between CGAS and CIS scores, respectively with various patient and caregiver variables. Lower CGAS scores (implying poorer clinician-rated functioning) were associated with lower educational level of the child, a main diagnosis other than seizure disorder, longer duration of illness, and presence of comorbidity.

Higher CIS scores (implying worse caregiver rating of impairment) were equally associated with lower child education, and main diagnosis other than seizure disorder. Mothers also rated the children worse than other caregivers. Both CGAS and CIS scores showed significant correlation with Zarit and GHQ scores. Finally, a significant correlation was found between CGAS and CIS scores (r=−0.388, p<0.001).

Discussion

This study provides comparative information on functional impairment among children with neuropsychiatric disorders attending a specialist facility in Lagos, Nigeria. A sig­ nificant correlation was found between caregiver-rated level of functioning and clinician-rated level of impairment in the children. Concurrent validity of the CGAS and CIS has been reported in a Caucasian sample,[1] and in Nigeria, between the CGAS and another caregiver-rated measure, the Child Behaviour Questionnaire.[8] Factors such as duration of illness and presence of comorbidity were associated with the clinician’s assessment of functioning but not with that of the caregivers. From a clinical standpoint, it appears intuitive that these factors will contribute to, and be indices of, the functional capacity of the child over time. Longer duration of illness and psychiatric diagnosis have been associated with linguistic impairment among children with epilepsy. [16] According to Hamiwka and Wirrell,[17] comorbid conditions in children with epilepsy include cognitive impairment and neuropsychiatric problems, which are often more disabling to the children than the seizures themselves, and are a source of increased stress and burden for families.

9

SAJCH

MARCH 2016 Vol. 10 No. 1

Table 1. Sociodemographic profile of children and their caregivers Variable

n (%)

Child’s gender Male

82 (52.9)

Female

73 (47.1)

Child’s level of education Primary or lower

89 (57.4)

Vocational or special

26 (16.8)

Secondary

34 (21.9)

Unspecified

6 (3.9)

Duration of illness (years) <1

18 (11.6)

1-5

59 (38.1)

>5

72 (46.5)

Unspecified

6 (3.9)

Main diagnosis Seizure disorder

87 (56.1)

ADHD/ASD

36 (23.2)

Mood/psychotic disorders

28 (18.1)

Intellectual disability

4 (2.6)

Comorbidity Present

58 (37.4)

Absent

43 (27.7)

Unspecified

54 (34.8)

Gender of caregiver Female

124 (80)

Male

31 (20)

Relationship of caregiver to child Mother

110 (71.0)

Father

19 (12.3)

Others

26 (16.7)

Caregiver’s employment status Employed

129 (83.2)

Unemployed

22 (14.2)

Unspecified

4 (2.6)

Caregiver’s marital status Married

125 (80.6)

Not married

30 (19.4)

Caregiver’s religion Christianity

110 (71.0)

Islam

41 (26.5)

Other/unspecified

4 (2.6)

However, in a study among children with temporal lobe epilepsy, Pereira and Valente[18] found no association between global functioning (as rated by clinicians using the CGAS) and seizure variables such as duration

SHORT REPORT Table 2. Associations of CGAS and CIS scores with various child and caregiver variables Variables

CGAS scores, mean (SD)

Association

CIS scores, mean (SD)

Association

55.3 (22.8)

r=0.142, p=0.106

16.4 (14.3)

r=−0.065, p=0.425

Male

52.9 (22.6)

t=–1.303, p=0.195

16.9 (15.9)

t=0.360, p=0.719

Female

58.2 (22.5)

Patient variables Age Sex 16.0 (11.7)

Level of education F=12.276, p<0.001*

Primary or lower

54.1 (21.8)

18.7 (15.4)

Vocational or special

40.2 (17.5)

16.0 (12.3)

Secondary

68.9 (20.1)

10.6 (11.2)

F=3.971, p=0.021*

Duration of illness (years) F=7.794, p=0.001*

<1

70.3 (16.0)

1-5

59.0 (21.7)

12.0 (11.9) 17.4 (15.6)

>5

48.4 (21.9)

16.0 (13.5)

F=1.014, p=0.365

Main diagnosis F=7.336, p<0.001*

Seizure disorder

63.3 (20.4)

ADHD/ASD

38.6 (13.8)

12.8 (12.0) 19.6 (14.1)

Psychotic disorders

31.7 (20.8)

13.0 (15.1)

Intellectual disability

51.4 (22.9)

23.7 (17.2)

F=2.562, p=0.041*

Presence of comorbidity Yes

43.7 (19.2)

No

68.0 (18.4)

t=–6.164, p<0.001*

16.7 (13.1)

t=1.722, p=0.079

12.0 (12.9)

Presence of psychosis Yes

52.3 (25.5)

No

56.0 (21.4)

t=–0.594, p=0.554

12.8 (14.2)

t=–0.731, p=0.466

15.4 (13.0)

Caregiver variables 55.3 (22.8)

r=–0.082, p=0.057

16.4 (14.3)

r=0.057, p=0.517

Male

56.0 (22.5)

t=0.745, p=0.458

15.7 (14.4)

t=–1.25, p=0.213

Female

52.3 (24.4)

Age Sex

19.3 (13.9)

Relationship to patient Mother

55.6 (23.1)

Other

54.3 (23.2)

t=0.264, p=0.793

20.7 (14.9)

t=–2.109, p=0.037*

15.0 (13.9)

Employment status Employed

57.1 (22.2)

Unemployed

46.3 (24.7)

t=1.747, p=0.083

16.4 (14.3)

t=–0.214, p=0.831

17.1 (14.8)

Marital status Married

55.7 (22.8)

Not married

49.2 (22.6)

t=1.461, p=0.146

15.9 (14.1)

t=–0.836, p=0.405

57.2 (22.1)

Religion t=0.637, p=0.525

16.2 (15.2)

t=0.328, p=0.743

Christianity

57.2 (22.1)

Islam

54.4 (23.2)

Zarit score

55.3 (22.8)

r=–0.47, p<0.001*

16.4 (14.3)

r=0.550, p<0.001*

GHQ score

55.3 (22.8)

r=–0.237, p=0.006*

16.4 (14.3)

r=0.506, p<0.001*

17.1 (12.1)

*Significant.

10

SAJCH

MARCH 2016 Vol. 10 No. 1

SHORT REPORT of illness. Nevertheless, the fact that no significant association with these variables was found with the caregivers’ assessment may be a limitation of the CIS as an instrument, rather than conclusive evidence that longer duration of illness and presence of comorbidity do not factor into a mother’s assessment of the child’s level of impairment.

For both clinicians and caregivers, a lower level of educational attainment was associated with poorer functional capacity. Functional impairment and poor educational outcomes have been reported in ADHD and autism,[19,20] as well as in epilepsy.[21] A search of the literature, however, yielded no study providing empirical justification for the agreement between clinicians and caregivers.

Further, for both clinicians and caregivers, seizure disorder was associated with relatively better functioning, possibly because in between seizure episodes most children are essentially normal, and in many cases seizure control is attainable. Disorders such as ADHD and intellectual disability, which run a stable course, take a persistent toll on the caregiver, and have been shown to be associated with more impairment.[8] The agreement between clinicians and caregivers on the relative better functioning of children with epilepsy, however, requires further study.

Caregiver-rated level of functioning was significantly associated with the relationship with the child, with mothers likely to rate the child as more impaired. This may be because mothers have more contact time with the child, and the burden of caring for an ill child falls more often on the mother.[22] Being a mother has been reported to be associated with more psychiatric morbidity among caregivers of children with epilepsy,[23] and mothers of functionally impaired children are more likely to have mental ill health,[8] which has been postulated to influence their assessment of the child.[24] Both the CGAS and the CIS showed a significant association with the burden experienced by the caregiver and the psychiatric morbidity experienced by the caregiver. Higher impairment, or poorer functioning, results in more requirements being made of the caregiver, with a consequent heavier burden and a greater likelihood of psychiatric morbidity in the caregiver.[25,26] The study was limited in that for certain variables, incomplete in­ formation reduced the number of units available for analysis, in what was at best a modest sample size. Another limitation of the study, besides the absence of prior validation of the CIS, is the fact that raters on the CGAS were not blind to patient diagnosis. This has the possibility of introducing bias. Furthermore, although all raters had undergone similar residency training and their use of the CGAS could be considered equivalent, no objective assessment of inter-rater reliability was done. The study was also conducted in a single clinic population, limiting generalisability. Nevertheless, this study gives vital information from a less-researched child population.

Although this study was not designed as a validation study, the CIS appears to be a useful tool in the assessment of functional impairment in children and adolescents. Further studies will, however, need to be done to assess its psychometric properties. Studies will also need to utilise more robust tools such as the WHO Disability Assessment Schedule.

Conclusion

The study showed that children attending specialist neuropsychiatric services have notable levels of functional impairment which may be associated with lower child education and which are linked with increased caregiver burden and psychiatric morbidity. This highlights the clinical importance of routine assessment of functional impairment in children with neuropsychiatric disorders. An additional implication of this finding, in a resource-constrained setting, is that policy initiatives with creative strategies are required for a holistic approach to neuropsychiatric disorders. This has to be multisectoral, involving in particular educational and social welfare services. Mother-andchild healthcare platforms may also need to be integrated to ensure that the burden and morbidity associated with caring for children with neuropsychiatric disorders are not unattended.

11

SAJCH

Acknowledgement. This article was presented at the World Psychiatric Association Regional Conference, Abuja, Nigeria, 22 - 24 October 2009.

References 1. Steinhausen HC, Metzke CW. Global measures of impairment in children and adolescents: Results from a Swiss community survey. Aust N Z J Psychiatry 2001;35(3):282-286. [http://dx.doi.org/10.1046/j.1440-1614.2001.00901.x] 2. Stringaris A, Goodman R. The value of measuring impact alongside symptoms in children and adolescents: A longitudinal assessment in a community sample.

J Abnorm Child Psychol 2003;41(7):1109-1120. [http://dx.doi.org/10.1007/ s10802-013-9744-x] 3. Sartorius N. Disability and mental illness are different entities and should be assessed separately. World Psychiatry 2009;8(2):86. [http://dx.doi. org/10.1002/j.2051-5545.2009.tb00220.x] 4. World Health Organization (WHO). International Classification of Functioning, Disability and Health (ICF). Geneva: WHO, 2001.

5. Faulks D, Norderyd J, Molina G, et al. Using the International Classification of Functioning, Disability and Health (ICF) to describe children referred to special care or paediatric dental services. PLoS ONE 2013;8(4):e61993. [http:// dx.doi.org/10.1371/journal.pone.0061993] 6. WHO. International Classification of Functioning, Disability and Health Children and Youth Version (ICF-CY). Geneva: WHO, 2007.

7. Skounti M, Bitzaraki A. The clinical impairment of children with attention deficit hyperactivity disorder: Problematic of diagnostic criteria. Psychiatriki 2011;22(2):158-163.

8. Tunde-Ayinmode M, Adegunloye O, Ayinmode B, Abiodun O. Psychiatric disorders in children attending a Nigerian primary care unit: Functional impairment and risk factors. Child Adolesc Psychiatry Ment Health 2012;6(1):28. [http://dx.doi.org/ 10.1186/1753-2000-6-28] 9. Al-Ansari AM. Measurement of impairment among children with attention deficit hyperactivity disorder as part of evaluating treatment outcome. Sultan Qaboos Univ Med J 2013;13(2):296-300. [http://dx.doi. org/10.12816/0003237] 10. Wigman JTW, Devlin N, Kelleher I, et al. Psychotic symptoms, functioning and coping in adolescents with mental illness. BMC Psychiatry 2014;14:97. [http:// dx.doi.org/10.1186/1471-244X-14-97] 11. Schaffer D, Gould MS, Brasic J, et al. A Children’s Global Assessment Scale (CGAS). Arch Gen Psychiatry 1983;40(11):1228-1231.

12. Rey JM, Starling J, Wever C, Dossetor DR, Plapp JM. Inter-rater reliability of global assessment of functioning in a clinical setting. J Child Psychol Psychiatry 1995;36(5):787-795. [http://dx.doi.org/10.1111/j.1469-7610.1995.tb01329.x] 13. Bird HR, Shaffer D, Fisher P, et al. The Columbia Impairment Scale (CIS): Pilot findings on a measure of global impairment for children and adolescents. Int J Methods Psychiatr Res 1993;3(3);167-176.

14. Zarit SH, Orr NK, Zarit JM.The Hidden Victims of Alzheimer’s Disease: Families Under Stress. New York: New York University Press, 1985.

15. Goldberg D, Gater R, Sartorius N, et al. The validity of two versions of the General Health Questionnaire in the WHO study of mental illness in general health care. Psychol Med 1997;27(1):191-197.

16. Caplan R, Siddarth P, Vona P, et al. Language in pediatric epilepsy. Epilepsia 2009;50(11):2397-2407.

17. Hamiwka LD, Wirrell EC. Comorbidities in pediatric epilepsy: Beyond ‘just’ treating the seizures. J Clin Neurol 2009;24(6):734-742. [http://dx.doi. org/10.1177/0883073808329527] 18. Pereira A, Valente KD. Severity of depressive symptomatology and functional impairment in children and adolescents with temporal lobe epilepsy. Seizure 2013;22(9):708-712. [htpp://dx.doi.org/10.1016/j.seizure.2013.05.008] 19. Bussing BR, Mason DM, Bell L, Porter P, Garvan C. Adolescent outcomes of childhood attention deficit/hyperactivity disorder in a diverse community sample. J Am Acad Child Adolesc Psychiatry 2010;49(6):595-605. [http:// dx.doi.org/10.1016/j.jaac.2010.03.006] 20. Shattuck PT, Narendorf SC, Cooper B, Sterzing PR, Wagner W, Taylor JL. Postsecondary education and employment among youth with an autism spectrum disorder. Pediatrics 2012;129(6):1042-1049. [http://dx.doi. org/10.1542/peds.2011-2864] 21. Reilly C, Atkinson P, Das KB, et al. Academic achievement in schoolaged children with active epilepsy: A population-based study. Epilepsia 2014;55(12):1910-1917. [http://dx.doi.org/10.1111/epi.12826] 22. Shore CP, Austin JK, Dunn DW. Maternal adaptation to a child’s epilepsy.

Epilepsy and Behaviour 2004;5(4):557-568. [http://dx.doi.org/10.1016/j. yebeh.2004.04.015] 23. Babalola EO, Adebowale TO, Onifade PO, Adelufosi AO. Prevalence and correlates of generalized anxiety disorder and depression among caregivers of children and adolescents with seizure disorders. J Behav Health 2014;3(2):122127. [http://dx.doi.org/10.5455/jbh.20140526121601] 24. Ordway MR. Depressed mothers as informants on child behavior: Methodological issues. Res Nurs Health 2011;34(6):520-532. [http://dx.doi. org/10.1002/nur.20463] 25. Okewole A, Dada MU, Ogun O, Bello-Mojeed M. Prevalence and correlates of psychiatric morbidity among caregivers of children and adolescents with neuropsychiatric disorders in Nigeria. Afr J Psychiatry 2011;14(4):306-309. [http://dx.doi.org/10.4314/ajpsy.v14i4.8] 26. Dada MU, Okewole NO, Ogun CO, Bello-Mojeed M. Factors associated with caregiver burden in a child and adolescent psychiatric facility in Lagos, Nigeria: A descriptive cross sectional study. BMC Pediatrics 2011;11(1):110. [http:// dx.doi.org/10.1186/1471-2431-11-110]

MARCH 2016 Vol. 10 No. 1

RESEARCH

Prevalence of positive coeliac serology in a cohort of South African children with type 1 diabetes mellitus S Tayob,1,2 MB ChB, DCH, FCP (Paed), Cert Paed Endo; K Pillay,2,3 MB ChB, DCH, FCP (Paed), Cert Paed Endo; B Tlou,4 MSc; Y Ganie,1,2 MB ChB, DCH, FCP (Paed), Cert Paed Endo epartment of Paediatrics and Child Health, Nelson R Mandela School of Medicine, Faculty of Health Sciences, University of KwaZulu-Natal, D Durban, South Africa 2 Division of Paediatric Endocrinology, Inkosi Albert Luthuli Central Hospital, Durban, South Africa 3 Centre for Paediatric Endocrinology and Diabetes, Westville Hospital, Durban, South Africa 4 Discipline of Public Health Medicine, School of Nursing and Public Health, University of KwaZulu-Natal, Durban, South Africa 1

Corresponding author: S Tayob (shafeekat@gmail.com)

Background. Coeliac disease (CD) is characterised by immune-mediated damage to the mucosa of the small intestine. Both CD and type 1 diabetes (T1D) have common auto-immune origins. Many patients with CD and T1D are asymptomatic or present with only mild symptoms; hence early diagnosis may only be facilitated by serological screening. Distal duodenal biopsy remains the gold standard for confirming the diagnosis.

Objective. To describe the prevalence of CD in T1D patients presenting to the paediatric endocrine service at Inkosi Albert Luthuli Central hospital (IALCH) in Durban and document the relationship between positive coeliac serology and small-bowel biopsy results.

Methods. A retrospective chart review was done at IALCH, the paediatric tertiary referral centre for KwaZulu-Natal (KZN) Province. The study sample included all patients with newly diagnosed T1D diagnosed between January 2008 and December 2011.

Results. A total of 120 newly diagnosed T1D patients were included in the study, of whom 49 (40.8%) were coeliac serology positive and 61 (50.8%) serology negative. There was no significant difference between the two groups regarding mean age of presentation with diabetes, race, sex, urban v. rural origin and baseline anthropometric measurements. Of patients in the serology-positive group, 97.6% had no symptoms suggestive of CD. Of the 49 patients who were coeliac serology positive, 8 (16%) were biopsied: 3 (37.5%) were positive, 1 (12.5%) had intra-epithelial lymphocytes and 4 (50%) were negative. There was a strong positive correlation between biopsy results and titres of endomysial antibody results (p=0.047).

Conclusion. There is a high prevalence of coeliac serology positivity in newly diagnosed T1D patients in KZN. This study provides evidence for screening of children with T1D for CD, and also confirms the low prevalence of symptoms.

S Afr J Child Health 2016;10(1):12-15. DOI:10.7196/SAJCH.2016.v106i1.835

Coeliac disease (CD) is characterised by immune-mediated damage to the mucosa of the small intestine. These changes are triggered by ingestion of gluten and related substances found in cereal grains. CD affects both developing and developed nations.

CD and type 1 diabetes (T1D) have common auto-immune origins. Both are associated with major histocompatibility complex class II antigen DQ2 encoded by alleles DQA1*501 and DQB1*201, thus providing a common genetic basis for the disease expression.[1] Recent work has also revealed seven shared non-human leukocyte antigen loci associated with CD and T1D.[1] This shared genetic basis is strongly suggestive of a common aetiology for both conditions.[1] Most estimates put the prevalence of CD at close to 1% of the general population.[1] The prevalence of CD in T1D has been reported to be five to seven times greater than in the general population, with increased prevalence rates among most ethnic groups.[1] Early epidemiological studies showed that CD was largely a disorder of white populations; however, recent studies reveal a similar prevalence in other race groups.

Although sampling rates and diagnostic criteria differ among studies, rates of biopsy-proven CD in paediatric T1D range from as low as 2.4% in Finland to 16.4% in Algeria.[1] The prevalence of biopsy-proven CD in the paediatric T1D population was 10.3% in a Libyan study, 4% in an Egyptian study and 5.3% in a Tunisian study.[1] The prevalence of CD in sub-Saharan Africa (which includes different ethnic populations to other parts of Africa) is as yet undescribed.

12

SAJCH

The classic presentation of CD includes symptoms related to gastro­ intestinal malabsorption, such as malnutrition, failure to thrive, diarrhoea, anorexia, constipation, vomiting, abdominal distension and pain. Non-gastrointestinal symptoms of CD include short stature, pubertal delay, fatigue, vitamin deficiencies and iron-deficiency anaemia.[1] Many patients with T1D who have CD are either asymptomatic (silent disease) or present with only mild symptoms.[1] Hence early diagnosis may only be facilitated by serological screening.

The main reason for screening asymptomatic individuals is to institute early treatment, thereby decreasing the risk of longterm CD-related complications, including malignancy of the gastrointestinal tract and osteoporosis, iron-deficiency anaemia and growth failure secondary to malabsorption.[1-3] A second advantage of screening asymptomatic patients with diabetes is the potential for improved diabetes control; undiagnosed CD has been associated with increased frequency of hypoglycaemic episodes.[3] Longstanding CD may be associated with an increased risk of retinopathy, and non-adherence to a gluten-free diet may increase the risk of microalbuminuria.[3] Intestinal biopsy is considered the gold standard in the diagnosis of CD. The staging of mucosal changes suggested by Marsh is now widely used.[4] The introduction of serological testing has facilitated screening of populations at risk for CD. Screening tests for IgA endomysial anti­­ bodies (EMA) and tissue transglutaminase (tTG) IgA have been reported to be the most sensitive and specific.[1] Both tests demonstrate

MARCH 2016 Vol. 10 No. 1

RESEARCH

Methods

The study design was a retrospective chart review. The study population included all children with newly diagnosed diabetes referred to the Paediatric Endocrine Unit at IALCH between January 2008 and December 2011. Children and adolescents with type 2 diabetes (T2D) or neonatal diabetes or known T1D patients with CD were excluded.

The study sample was divided into two groups based on their coeliac serology results and then compared. TTG IgA and IgG, EMA and antigliadin antibody (AG) IgA and IgG were measured in all patients. The Bio-Rad Laboratories (South Africa) kit was used for the tTG and AG enzyme-linked immunosorbent assay and the Diagnostic Technical Services (South Africa) kit was utilised for the EMA indirect immunofluorescence. A positive tTG IgA or IgG (>15 U/mL), EMA or AG IgA or IgG (>15 U/mL) were included in the analysis.

Serological tests and biopsies were carried out while patients were still on a glutencontaining diet. Total IgA was not routinely measured but requested only in those with an initial negative serological result in whom there was a strong suspicion of disease.

Symptoms were captured as either present or absent using a standardised questionnaire.

Specific symptoms included were chronic diarrhoea, constipation, abdominal pain, nausea, vomiting, flatulence, bloating, loss of weight and fatigue.

Data were collated on an Excel (Microsoft, USA) spreadsheet and analysed using SPSS version 21 (IBM, USA). Categorical variables were assessed using the Pearson’s χ2 test and quantitative variables using the Student’s t-test. A p-value of <0.05 was considered statistically significant.

Results

There were 139 newly diagnosed diabetes patients referred to the paediatric endocrine service between January 2008 and December 2011. Of these, 120 (86.3%) had T1D, 8.6% had T2D and 5% had neonatal diabetes.

Of the 120 T1D patients, 49 (40.8%) were coeliac serology positive, 61 (50.8%) were coeliac serology negative, and the serological result was unknown in 10 patients. Hence the prevalence of positive coeliac serology in our cohort of T1D patients was 44.5%.

Of the 49 patients with a positive serological result, 17 were tTG IgA positive and all of the 8 patients who underwent biopsy were tTG IgA positive.

The coeliac serology-positive and -negative groups were compared in terms of their demo­ graphic features, as summarised in Table 1.

The mean age at presentation was 7.32 years in the serology-positive group and 6.84 years in the serology-negative group (p=0.464).

The majority of patients in both groups were black Africans (71.4% in the serologypositive group and 70.5% in the serologynegative group). In the serology-positive group 22.4% of the patients were Indian, 2% were white and 4.1% were coloured.

In the serology-negative group 21.3% of the patients were Indian, 4.9% were white and 3.3% were coloured. Ethnicity had no significant effect on the prevalence of positive coeliac serology (p=0.876).

There was a female predominance overall which was more marked in the coeliac 50

90 37

80

serology-positive group, in which 57.1% were female and 42.9% were male; however, this was not statistically significant (p=0.509).

The majority of the patients in the serology-positive group were of rural origin (53.1%), while in the serology-negative group 60.7% of patients were of urban origin. This difference was not significant (p=0.151).

There was no significant difference observed in the baseline anthropometric measurements between the two groups.

It was interesting to note that in the serology-positive group 97.6% (n=41) of patients had no symptoms suggestive of CD, while only 2.4% (n=1) had symptoms of CD. The auto-immune markers of T1D as well as the thyroid antibodies in both the coeliac serology-positive and -negative groups were compared (Fig. 1). Of patients in the serologypositive group 77.1% (n=37) had positive glutamic acid decarboxylase (GAD) or insulin auto-antibodies 2 (IA2), while 85% (n=50) of patients in the serology-negative group had positive GAD or IA2 (p=0.312). Of the patients, 76% (n=35) in the serology-positive group and 85% (n=45) in the serology-negative group had positive islet cell antibodies (ICA) (p=0.266), while 29% (n=13) of patients in the serology-positive group and 26% (n=15) in the serology-negative group had positive thyroid antibodies (p=0.772).

Of the 49 coeliac serology-positive patients, eight (16%) underwent biopsy.

This is a very small number and is one of the limitations of this study. Three of the 8 patients (37.5%) had a positive biopsy result, 1 (12.5%) had intra-epithelial lymphocytosis and 4 (50%) had negative biopsies. Of the 3 positive biopsies, 1 was classified as Marsh stage 1, and 2 as Marsh stage 3a. All of the eight patients who were biopsied were tTG IgA positive. Two of the three positive biopsies were also

45

Coeliac serology positive

35

Coeliac serology negative

70 60 60 %

sensitivity and specificity of >90%.[3,5,6] According to the American Gastroenterology Association,[7] in the primary care setting the tTG IgA is the gold standard serological test for the detection of CD. CD and autoimmune thyroid disorders share a common genetic predisposition, namely the DQ2 allele, which accounts for the higher incidence of thyroid autoimmune disorders in CD than in the general pop­ ulation.[8] Despite the advent of sensitive and specific serological testing, routine screening for CD in diabetic populations may not be universal practice. The International Society for Pediatric and Adolescent Diabetes (ISPAD) and the European Society for Pediatric Gastroenterology and Hepatology (ESPGHAN) both recommend screening of T1D patients for CD.[3,9] ISPAD recommends screening for CD at the time of diabetes diagnosis, and every 1 - 2 years thereafter.

From 2008 onwards all new T1D patients attending the Inkosi Albert Luthuli Central Hospital (IALCH) diabetes clinic in Durban have been screened for CD at diagnosis. Analysis of these data to assess the prevalence of CD in this T1D population will provide information on implications for future screening practices.

50

13

40

15

30 10 0 GAD/IA2 positive p=0.312

ICA positive p=0.266

Thyroid antibody positive p=0.772

Fig. 1. Auto-immune markers of T1D and thyroid disease.

13

SAJCH

MARCH 2016 Vol. 10 No. 1

RESEARCH EMA positive. Three of the five biopsynegative patients were positive for all three antibodies, one was EMA and tTG IgA positive (AG IgA negative) and one was only tTG IgA positive.

Although the number of patients biopsied was inadequate, the results showed a positive correlation (p=0.047) between biopsy and the presence of EMA. A correlation could not be assessed for tTG IgA as all of the eight patients who underwent biopsy had a positive tTG IgA. There was no significant correlation between biopsy and tTG IgG or AG.

Discussion

The prevalence of positive coeliac serology in our cohort of T1D patients was 44.5%.

If the TTG IgA alone had been used, the prevalence in this study would have decreased to 15.5%. This is a high prevalence compared with previously described rates for other countries. However, our prevalence rates were for coeliac serology positivity, while the others were for biopsy-proven CD. A limitation and contributing factor to the high prevalence of positive coeliac serology in this study is that patients with positive AG IgA were also included in the analysis. Combining several tests for CD in lieu of tTG IgA alone may marginally increase sensitivity but reduces specificity and is therefore not recommended. The specificity of coeliac antibody testing in black South Africans is not known. A control group of non-diabetic black South Africans was not used to compare the serological results. Causes of false-positive serology results, such as inflammatory bowel disease, irritable bowel syndrome, cirrhosis and viral infections, were excluded.

There may have been a racial bias as the majority of patients were of black African

origin in both the coeliac serology-positive (71.4%) and -negative (70.5%) groups; how­ ever, this reflects the population statistics in South Africa (SA). The research was conducted in a government hospital, which may also result in possible bias as privatesector patients, who represent a different demographic and socioeconomic group, were not included.

The mean age of biopsy-proven CD in T1D in an Italian study of children in 2008 was 8.1 (4.3) years.[1] The mean age of the serology-positive patients was similar in this study at 7.32 years. This is reflective of the shift toward a younger age of onset of T1D worldwide.

This study found higher coeliac serology positivity among females (57.1%) than males (42.9%), which is consistent with international data.

Of patients in the serology-positive group, 97.6% had no symptoms of CD. This was a retrospective study, and it is likely that there was inadequate information in the patient records, which may explain the high percentage of asymptomatic patients compared with other studies. Our result was higher than that reported in a North American study performed at a CD clinic, which revealed that 71.4% of children reported no gastrointestinal symptoms at the time of a positive screening test.[1] A possible reason for our results is that the predominance of gastrointestinal symptoms is most common in children <3 years,[1] and the mean age of our serology-positive group was 7.32 years. In asymptomatic individuals at high risk (e.g. those with T1D) CD should always be diagnosed using duodenal biopsies, as this population more often have false-positive tTG results.[9] Biopsy may not be required in those patients with signs or

Table 1. Comparison of demographics of the CD serology-positive and CD serologynegative T1D patients

Age (years)

CD serology positive (N=49), % (n)

CD serology negative (N=61), % (n)

p-value

7.32

6.84

0.464

Race African

71.4 (35)

70.5 (43)

Indian

22.4 (11)

21.3 (13)

White

2.0 (1)

4.9 (3)

Coloured

4.1 (2)

3.3 (2)

Male

42.9 (21)

49.2 (30)

Female

57.1 (28)

50.8 (31)

0.876

Sex

Urban

46.9 (23)

60.7 (37)

53.1 (26)

39.3 (24)

14

SAJCH

Study limitations

This study took the form of a retrospective chart review. Patients with a positive AG IgA result were included in the analysis, resulting in a higher than expected prevalence of positive coeliac serology. Specific symptom variables were not individually captured, and this may explain the low prevalence of symptoms in patients. An important

Urban v. rural Rural

symptoms suggestive of CD and tTG levels >10 times the upper limit of normal.[9] There was no significant difference between the two groups in terms of the auto-immune markers of T1D or thyroid antibodies. This is in keeping with a study by Sumnik et al., cited

by Fasano,[10] who performed a multicentre retrospective case-control study comparing data from 84 diabetic children with CD with 167 diabetic children without CD, and concluded that the occurrence of thyroid autoimmunity in diabetic children is not related to coexisting CD. However, Velluzi et al., cited by Kumar et al.[8] found that CD patients had a three- to fourfold increase in the incidence of thyroid auto-immunity.

Of the 49 coeliac serology-positive patients, 8 (16%) underwent biopsy. This is a small number and a significant limitation of the study. There is no qualified paediatric gastro­ enterologist at IALCH and it was a challenge to obtain biopsies. Of the 8 patients biopsied, 3 (37.5%) had a positive result (1 Marsh stage 1 and 2 Marsh stage 3a). This is a lower rate than expected. A possible explanation for this result is that the biopsy technique was not standardised. The biopsies were not performed by the same doctor (paediatric surgeon or adult gastroenterologist), and the number of specimens obtained and biopsy sites differed. At least four biopsies should be taken to increase sensitivity and specificity as the typical coeliac lesions can be patchy. [4,8] Correct histological interpretation by the pathologist is another important issue.

Specimens were not assessed by the same pathologist; however, all specimens were subsequently reviewed by the head of the Department of Pathology.

Although the number of patients biopsied was small, the positive correlation (p=0.047) between biopsy and EMA results found in this study was in keeping with a review article which examined 32 studies and found a mean specificity and sensitivity of 99% and 95% respectively for this serological marker.[1] The same article found the mean specificity and sensitivity of tTG IgA in 27 studies to be 95% and 87% respectively.[1] However, a correlation for tTG IgA could not be assessed in this cohort as all patients who underwent biopsy had a positive tTG IgA result. Numerous studies have documented the poor accuracy of the AG IgA tests, making them unsuitable for screening purposes.

MARCH 2016 Vol. 10 No. 1

RESEARCH limitation is that only a small number of patients were biopsied and that the biopsy procedure was not standardised.

Conclusion

There is a high prevalence of coeliac serology positivity in newly diagnosed T1D patients in KwaZulu-Natal Province, SA. Although all patients did not undergo biopsy, this study provides evidence for screening of children with T1D for CD. There was no significant difference with respect to GAD/IA2 and ICA or thyroid antibodies between the two groups. The relationship between CD and glycaemic control in these patients remains to be established. A prospective follow-up study is required where patients are commenced on a gluten-free diet and clinical and glycaemic outcomes are assessed.

Recommendations

ISPAD recommends screening for CD using the TTG IgA and/or the EMA in all T1D patients at diagnosis and every 1 - 2 years thereafter. The ESPGHAN recommends the use of TTG IgA as the initial screening test for CD.[9] All three coeliac antibodies should not be routinely measured for screening purposes. We support the ISPAD recommendations to screen all newly diagnosed T1DM for CD in SA. Patients with a positive screening test should be referred to a paediatric gastroenterologist for a duodenal biopsy to confirm the diagnosis of CD.

15

SAJCH

References 1. Sud S, Marcon M, Assor E, Palmert MR. Coeliac disease and paediatric type 1 diabetes: Diagnostic and treatment dilemmas. Int J Paediatr Endocrinol 2010;2010:161285. [http://dx.doi.org/10.1155/2010/161285] 2. Cronin CC, Shanahan F. Insulin dependent diabetes mellitus and coeliac disease.

Lancet 1997;349(9058):9058. [http://dx.doi.org/10.1016/S0140-6736(96)09153-2] 3. Kordonouri O, Klingensmith G, Knip M, et al. Other complications and diabetes-associated conditions in children and adolescents. Pediatr Diabetes 2014;15(Suppl. 20):S270-278. [http://dx.doi.org/10.1111/pedi.12183] 4. Sollid LM, Lundin KEA. Diagnosis and treatment of coeliac disease. Mucosal Immunol 2009;2(1):3-7. [http://dx.doi.org/10.1038/mi.2008.74] 5. Lewis NR, Scott BB. Systematic review: The use of serology to exclude or diagnose coeliac disease (a comparison of endomysial and tissue transglutaminase antibody tests). Alimentary Pharmacol Ther 2006;24(1):4754. [http://dx.doi.org/10.1111/j.1365-2036.2006.02967.x] 6. Hill ID. What are the sensitivity and specificity of serologic tests for celiac disease? Do sensitivity and specificity vary in different populations?

Gastroenterology 2005;128(4 Suppl 1):S25-32.

7. American Gastroenterology Association (AGA). AGA institute medical position statement on the diagnosis and management of coeliac disease.

Gastroenterology 2006;131(6):1977-1980.

8. Kumar V, Rajadhyaksha M, Wortsman J. Coeliac disease-associated autoimmune endocrinopathies. Clin Diagn Lab Immunol 2001;8(4):678-685.

9. Husby S, Koletzko S, Korponay-Szabo IR. European Society for Paediatric Gastroenterology, Hepatology, and Nutrition Guidelines for the management of coeliac disease. J Pediatr Gastroenterol Nutr 2012;54(1):136-160. [http:// dx.doi.org/10.1097/MPG.0b013e31821a23d0] 10. Fasano A. Systemic auto-immune disease in coeliac disease. Curr Opin Gastroenterol 2006;22(6):674-679. [http://dx.doi.org/10.1097/01.mog.0000245543.72537.9e]

MARCH 2016 Vol. 10 No. 1

RESEARCH

Effect of community integrated management of childhood illness on mothers’ healthcare-seeking behaviour and home management of childhood illness in Ile-Ife, South-West Nigeria: A household survey O A Ogundele,1 MB ChB, MPH, FWACP; T Ogundele,2 MB ChB, MPH; O S Olajide,1 MB ChB, MPH, FWACP; O I Agunbiade,3 BNSc Department of Community Health, Obafemi Awolowo University Teaching Hospital Complex, Ile-Ife, Osun State, Nigeria Department of Paediatrics, Obafemi Awolowo University Teaching Hospital Complex, Ile-Ife, Osun State, Nigeria 3 Intensive Care Unit, Obafemi Awolowo University Teaching Hospital Complex, Ile-Ife, Osun State, Nigeria 1 2

Corresponding author: O A Ogundele (femiaoo@yahoo.com) Background. Care-seeking interventions, as part of community integrated management of childhood illness (CIMCI), have the potential to substantially reduce child mortality in countries where common childhood illnesses are a major problem. Prompt and appropriate careseeking practices are important to avoid many deaths attributed to delays in or not seeking care, particularly in developing countries such as Nigeria.

Objective. To assess the effect of community-level intervention on mothers’ care-seeking behaviour for common childhood illnesses and related influencing factors.

Methods. The study had a comparative cross-sectional design and was conducted in two local government areas (LGAs) of Osun State, South-West Nigeria. A total of 722 mothers of index children aged <5 years were selected through a multistage cluster sampling technique.

Data were collected and analysed using SPSS version 16.0. Descriptive, bivariate and multivariate analyses were performed.

Results. Care-seeking for children who reported illness was higher in the CIMCI-implementing LGA (90.2%) compared with 74.8% in the non-implementing LGA (p=0.002). Care was sought within the first 48 hours of perceived onset of illness for 83.2% and 57.9% of sick children in the CIMCI-implementing and non-implementing LGAs, respectively. Residing in a CIMCI-implementing area (odds ratio (OR) 2.54, 95% confidence interval (CI) 1.24 - 5.45) and maternal education level (OR 1.50, 95% CI 1.06 - 3.03) were identified as predictors of healthcare-seeking practices among mothers.

Conclusion. The study concluded that a high level of care-seeking behaviour exists where community-level intervention was operating.

Therefore the CIMCI programme should be strengthened further and also scaled up to include non-implementing communities.

S Afr J Child Health 2016;10(1):16-19. DOI:10.7196/SAJCH.2016.v106i1.912

Untreated infections are a major cause of childhood mortality and morbidity in developing countries.[1] While uncomplicated illness can be treated at home, severe illness should be treated by a health professional.

Most child deaths occurring in low- and middle-income countries such as Nigeria are attributable to a handful of diseases and are avoidable through existing interventions[2] such as the integrated management of childhood illness (IMCI) strategy developed by the World Health Organization (WHO), the United Nations Children’s Fund and other technical partners.[3-5] In Nigeria currently 3 out of 20 children die before their fifth birthday, and 70% of these deaths are due to acute respiratory infections (mostly pneumonia), diarrhoea, measles, malaria or malnutrition – and often a combination of these.[6,7] The WHO has estimated that seeking prompt and appropriate care could reduce child deaths due to acute respiratory infections by 20%.[8] A high number of children die without ever being taken to a health facility, indicating that the prevalence of appropriate care-seeking is low. Even for those who are taken to a health facility, visits are often not timely enough. Appropriate care-seeking means that the need to take the child for treatment outside the home is recognised, that the care is not delayed, and that the child is taken to an appropriate health facility or provider. Interventions to improve care-seeking behaviour include health education of caregivers through training of community health workers.[9] Community IMCI (CIMCI) is an integrated childcare approach that aims at improving household and community practices that are likely 16

SAJCH

to have the greatest effect on child survival, growth and development.

These key practices are growth promotion and development, disease prevention, home management, care-seeking and compliance. CIMCI has become important because most children die at home in spite of availability of health services.[4] CIMCI was implemented in Ife Central local government area (LGA) in 2005. Community resource persons (CORPs) were trained to provide information and mobilise caregivers on child survival at household and community level. CORPs provide information on appropriate care-seeking and encourage as much as possible that caregivers adopt these practices. The CIMCI strategy has now operated for more than 6 years in communities and households to promote improved preventive and curative health behaviours.

The objective of this study was to assess the effect of community-level intervention on mothers’ care-seeking behaviour for common childhood illnesses, and the factors that influenced the mothers’ behaviour.

Methods

Study sites

The study was of a comparative cross-sectional design, and was conducted between October and December 2012 in two LGAs of Osun State in South-West Nigeria. This state has a population of about 3.4 million[10] people. Ife Central Local Government, the study site, implements CIMCI. Ilesa East Local Government, the comparison site, does not implement the CIMCI programme. A tertiary health facility exists in both LGAs. There are 11 primary healthcare centre

MARCH 2016 Vol. 10 No. 1

RESEARCH facilities and one comprehensive healthcare centre in the 11 wards in each of the selected LGAs. The community settlement is organised into streets in both of these LGAs, which comprise both rural and urban communities.

starting point in the selected street, and systematic random selection of households was carried out. Interviews began in the first eligible household, moving systematically until the targeted number of interviews had been obtained.

Study population and sample size

Stage 5 The basic sampling unit was the household, and one mother/caregiver with an eligible child was interviewed in every household that was selected. In situations where the household had more than one eligible subject, the child with a birthday nearest to the interview period was chosen.

The study population was mothers of children 0 - 59 months old, and their index children (the child selected for the study).

Excluded from the study were mothers or caregivers who did not give consent or whose children were ill during the study period.

The sample size was determined using a formula for comparing independent pro­­ portions. This gave a sample size of 722 mothers or caregivers and their index children aged <5 years.

Sampling technique

A multistage cluster sampling technique was used to select 361 respondents each from the two study areas.

Stage 1 Ife Central LGA was purposively selected from the 30 LGAs in the state as the only LGA that implements CIMCI. Ilesa East LGA, the comparison LGA, was randomly selected from among the non-CIMCI-implementing LGAs in the state.

Stage 2 Wards 4 and 5 in Ife Central LGA were purposively selected, being the only two wards in the LGA that were implementing CIMCI, while wards 2 and 8 were randomly selected by balloting system from among the 11 wards in Ilesa East LGA. Stage 3 The enumeration areas (EAs) of the two implementing wards (wards 4 and 5) and those of the two randomly selected wards (wards 2 and 8) were made the primary sampling units. One-fifth of the EAs in each LGA were selected using simple random selection. At each site the sample size of 361 was divided by the selected EAs to determine the number of respondents to be recruited from each EA. Ife Central had 180 EAs, from which 36 EAs were selected, while Ilesa East had 150 EAs, from which 30 EAs were selected. An average of 10 respondents were recruited from selected EAs in wards 4 and 5 in Ife Central and 12 respondents from the selected EAs in wards 2 and 8 in Ilesa East until the sample size of 361 was obtained in each LGA. Stage 4 Household listings for streets in the EAs were done in both LGAs. One street was randomly selected from each EA by the ballot system.

A household was randomly selected as the

Data collection instrument

A structured, pretested interviewer-adminis­ tered questionnaire was used for data collec­ tion. The questionnaire was adapted from that previously used and validated by the WHO Nigeria.[11] The instrument covered four key areas of household practices: growth promotion, disease prevention,

home management, and care seeking. Five appropriately trained research assistants were recruited to assist in data collection. The interviewers were given an interview guide that was developed before the training.

Ethical clearance

Ethical clearance was obtained from the Ethics and Research Committee of Obafemi Awolowo University Teaching Hospitals Complex. Written informed consent was obtained from the study participants.

Data analysis

The data were analysed using SPSS version 16.0 (SPSS Inc., Chicago). Appropriate univariate, bivariate and multivariate analyses were done. Factors that were significantly associated (p<0.05) with healthcare-seeking behaviour were then used in logistic regress­ ion models to estimate associations of each covariate with this behaviour. A wealth index was constructed by assessing the presence or absence of durable assets in the household.

Table 1. Sociodemographic characteristics of respondents and index child by LGA (N=361) Characteristics

CIMCIimplementing LGA, n (%)

Non-CIMCI LGA, n (%)

≤30

178 (49.3)

170 (47.1)

>31

183 (50.7)

191 (52.9) χ2=2.43, df=1, p=0.119

Marital status Married

339 (93.9)

348 (96.4)

Not married

22 (6.1)

13 (3.6) χ2=23.20, df=1, p<0.001

Maternal education status Below secondary education

87 (24.1)

38 (10.5)

Secondary and above

274 (75.9)

323 (89.5) χ2=8.400, df=1, p=0.004

Maternal occupation Housewife

18 (5.0)

39 (10.8)

Not housewife

343 (95.0)

322 (89.2) χ2=1.10, df=1, p=0.290

Sex Male

209 (57.9)

195 (54.0)

Female

152 (42.1)

166 (46.0) χ2=24.78, df=1, p<0.001

Index children age (months) ≤11

76 (21.1)

37 (38.0)

>12 - 59

285 (78.9)

224 (62.0) χ2=0.894, df=1, p=0.344

Wealth index First and second quintile

234 (64.8)

246 (68.1)

Third and above

127 (35.2)

115 (31.9) χ2=16.41, df=1, p<0.001

Family size <5

248 (68.7)

295 (81.7)

>6

133 (31.3)

66 (18.3)

df = degrees of freedom. *A significance level of p<0.05 was used.

17

Statistical indices* χ2=0.355, df=1, p=0.552

Maternal age (years)

SAJCH

MARCH 2016 Vol. 10 No. 1

RESEARCH Presence was scored as 1 and absence as 0, and the mean assets score was recategorised into five different wealth quintiles of equal proportions (lowest, second, third, fourth and highest wealth quintiles).

Table 2. Mothers’ healthcare-seeking practices for children with reported illness 2 weeks prior to study, by LGA (N=361) CIMCIimplementing LGA, n (%)

Results

Statistical indices χ2=1.41, df=1, p=0.236

Index child sick 2 weeks prior to survey Yes

112 (31.0)

127 (35.2)

No

249 (69.0)

234 (64.8)

Total

361 (100)

361 (100) χ2=12.63, df=1, p<0.001

Symptoms/signs of illness* Fever, cough and diarrhoea

99 (88.4)

126 (99.2)

Others

13 (11.6)

1 (0.8)

Total

112 (100)

127 (100) χ2=9.49, df=1, p=0.002

Children taken to health facility* Yes

101 (90.2)

95 (74.8)

No

11 (9.8)

32 (25.2)

Total

112 (100)

127 (100) χ2=15.16, df=1, p<0.001

Time to care-seeking after onset of illness 1st and 2nd day

84 (83.2)

55 (57.9)

3rd day and after

17 (16.8)

40 (42.1)

Total

101 (100)

95 (100)

* Included only those that reported illness 2 weeks prior to survey.

40 35 Percentage of children (%)

A total of 722 households with children aged 0 - 59 months were enrolled in the study (Table 1). Almost 90% of mothers in the non-implementing LGA had secondary school education and above, compared with 75.9% in the CIMCI-implementing LGA (p<0.001). Only 5% of mothers in the implementing LGA and 10.8% in the non-implementing LGA were housewives.

More than half (57.9% and 54%) of the children from the implementing areas and non-implementing areas respectively were males. About 65% of study respondents in the CIMCI-implementing LGA were in the first and second wealth quintiles, compared with 68.1% in the non-implementing LGA; this difference was not statistically significant (p=0.344).

Within 2 weeks of the survey, 31% and 35.2% of index children in the CIMCI-implementing LGA and the nonimplementing LGA were reported to have been ill (Table 2). Of these children, a high­ er proportion of those from the CIMCIimplementing LGA (90.2%) were taken to health facilities compared to 74.8% from the non-implementing LGA (p=0.002).

Fever, cough and diarrhoea were common symptoms in both LGAs, but were more frequent in the non-implementing LGA (88.4% in the implementing and 99.2% in the non-implementing LGA), this difference being statistically significant (p<0.001). Of the total treated episodes of illnesses, care was sought more frequently within the first 2 days (48 hours) of perceived onset of illness in the implementing than in the non-implementing LGA (83.2% and 57.9% of sick children, respectively; p<0.001).

In the rest of the cases – 16.8% in the implementing LGA and 42.1% in the nonimplementing LGA – care was sought from the third day onward.

As regards home treatment (Fig. 1), 25% of caregivers in the implementing LGA did not give any form of initial treatment before seeking care, compared with 8.7% in the non-implementing LGA. About 35.5% of caregivers in the non-implementing LGA gave paracetamol as an initial treatment, while tepid sponging was a very common practice in the implementing LGA (18.8%), compared with 4.7% in the nonimplementing LGA. At the bivariate analyses level, residing in CIMCI-implementing LGA (p<0.001) and mothers’ education level (p=0.001) were significantly associated with care-seeking,

Non-CIMCI LGA, n (%)

35.7

CIMCI

35.5

Non-CIMCI

30

28.3

25

25

22.8 18.8

20 15

12.5

10 4.7

5 0

8.7

8

Antimalarial/ antibiotics

Tepid sponge

Paracetamol

Herbal treatment

Nothing

Fig. 1. Home treatment of children who reported illness 2 weeks prior to study.

while age of mothers, age of the child, family size, marital status, occupation and wealth index were not statistically significant.

A multivariate logistic regression analy­ sis was done to determine which selected sociodemographic characteristics were asso­ ciated with healthcare-seeking behaviour among mothers (Table 3). The result revealed that mothers who reside in the CIMCIimplementing area (odds ratio (OR) 2.54, 95% confidence interval (CI) 1.24 -5.45) and those who had secondary education and above (OR 1.50, 95% CI 1.06 - 3.03) were more likely to seek care from the health facilities than the others.

18

SAJCH

MARCH 2016 Vol. 10 No. 1

Discussion

Prompt and appropriate care-seeking practice is very important to avoid many deaths attributed to delays and not seeking care, particularly in developing countries.

In this study care was sought from health facilities for nine in ten sick children in the CIMCI-implementing LGAs compared with seven in ten in the non-implementing LGAs; the difference was statistically significant. Adegboyega et al.[12] in a study in Lagos, Nigeria, reported the prevalence of care-seeking as less than that observed in this study. The difference observed in this study may therefore not be unconnected

RESEARCH Conclusion Table 3. Factors associated with healthcare-seeking behaviours of mothers for common childhood illnesses in Ile Ife, Nigeria (N=722) Variable

OR (95% CI)

p-value

Maternal age (Ref = ≤30 yr) ≥31 yr

0.18 (0.12 - 1.28)

0.655

Mother’s education (Ref = < secondary) ≥ Secondary

1.50 (1.06 - 3.03)

0.001*

Marital status (Ref = not married) Married

1.12 (0.35 - 4.42)

0.130

Family size (Ref = n≥6) ≤5

1.50 (0.69 - 3.12)

0.143

Occupation (Ref = housewives) Not housewives

0.52 (0.33 - 1.47)

0.780

Residence (Ref = non-CIMCI) CIMCI

2.54 (1.24 - 5.45)

0.001*

Wealth index (Ref = 1st and 2nd) 3rd quintile and above

0.46 (0.30 - 1.30)

0.216

Age of the child (Ref = <11 months) >12 months

0.86 (0.51 - 1.45)

0.677

Ref = reference. * Significant at the level of p<0.05.

to the effect of CIMCI implementation in the study area. Here CORPs were trained at the inception of the programme to provide information on appropriate care seeking to caregivers at household and community level, and to motivate them in adopting the practices as much as possible (findings from focus group discussion not reported here). Poor care-seeking practices have, however, been reported from Ethiopia and other developing countries,[13,14] unlike in this study, and this could be due to differences in access to health facilities, educational backgrounds, cultural factors and socioeconomic status.

For most mothers and caregivers in the CIMCI-implementing LGA in this study, care-seeking for childhood illness begins within the initial 48 hours, compared with the non-implementing LGA, where four in ten mothers and caregivers seek care on the third day or later. This is in keeping with findings by Tinuade et al.,[15] who observed that care-seeking is often delayed beyond 24 hours, and that most mothers sought initial care at home. Possible reasons for delay could be to try home care including traditional treatment, a lack of money, inadequate access to health facilities and poor illness perception.

Home treatment patterns also differed significantly between CIMCI-imple­ menting and non-implementing communities; ~19%

of mothers tepid sponge for fever in the implementing LGA, compared with ~5% in the non-implementing LGA. The preference for tepid sponging in the CIMCI-implementing LGA is to ensure that the fever does not get too high before mothers seek care, to make sure that the pattern of the fever is not marred before presentation, and to discourage inappropriate drug use. Most mothers gave some form of treatment, ranging from paracetamol, antimalarials and antibiotics, which was consistent with findings in other studies.[14,16] In this study the mother’s level of education and residence in a CIMCI-implementing area were predictors of care-seeking. This is in keeping with existing studies[13] that have shown the same, as well as other factors such as economic status, mothers’ age, ethnicity and distance from the health facility.

Study limitations

A limitation of this study was the crosssectional study design, which measured the exposure and outcome at the same time, and cannot measure the cause-and-effect relationship. There was also no measure of an adverse effect on the child of delay in seeking healthcare. Another issue is that respondents may not reveal their actual practice for fear of unknown action. This was, however, minimised by interviewers first gaining the confidence of the respondents.

19

SAJCH

MARCH 2016 Vol. 10 No. 1

The findings of this study showed the positive effect of community-level intervention in improving healthcare-seeking behaviour.

The CIMCI programme should therefore be strengthened and scaled-up to nonimplementing communities.

References 1. Aguilar AM, Alvarado R, Cordero D, et al.

Mortality Survey in Bolivia: The Final Report.

Investigating and Identifying the Causes of Death for Children Under Five. Arlington, USA: Basic Support for Institutionalizing Child Survival (BASICS) Project, 1998.

2. World Health Organization. World Health Report.

Geneva, Switzerland: World Health Organization, 2004.

3. United Nations Children’s Fund. A Working Paper on Home and Community Health Care to Enhance Child Survival, Growth and Development. New York: UNICEF, 1999.

4. World Health Organization. Improving family and community practices. A component of the IMCI strategy. Bull WHO 1998;73(15):119-128.

5. World Health Organization/Child and Adolescent Health. Community and Household Moving Forward Programme – Report. Geneva: World Health Organization, 2000.

6. National Population Commission. Nigeria Demographic and Health Survey 2008. Calverton, USA: ICF Macro, 2009.

7. Policy Project Nigeria. Child Survival in Nigeria: Situation, Response, and Prospects. October 2002. http://www.policyproject.com/pubs/countryreports/ nig_csrevised.pdf (accessed 28 January 2016).

8. Victora C, Jennifer B, Olivier F, Roeland M. Reducing deaths from diarrhoea through oral rehydration therapy. Bull WHO 2000;1(78):10.

9. Hill Z. Recognized childhood illness and their traditional explanation: Exploring options for careseeking interventions in the context of the IMCI strategy. Trop Med Int Health 2003;8(7):668-676. [http://dx.doi.org/10.1046/j.1365-3156.2003.01058.x] 10. Ministry of Health Osogbo. Osun State Health Facilities Inventory. Osogbo, Nigeria: Ministry of Health Osogbo, 2009:7-8.

11. United Nations Children’s Fund/Federal Ministry of health. IMCI in the Hands of Families: Nigeria Country Report of Baseline Studies on Key Family and Community Practices in IMCI-implemented LGAs. Abuja, Nigeria: Federal Ministry of Health, 2005.

12. Adegboyega AA, Onayade AA, Salawu O. Care-seeking behaviour of caregivers for common childhood illnesses in Lagos Island Local Government Area, Nigeria. Niger J Med 2005;14(4):461.

13. Assefa T, Belachew T, Tegegn A, Deribew A. Mothers’ health care seeking behavior for childhood illnesses in Derra district, Northshoa zone, Oromia regional state, Ethiopia. Ethiop J Health Sci 2008;18(3):87-94.

14. Tessema F, Makonnen A, Fekadu A. Mothers health services utilization and health care seeking behavior during infant rearing: West Ethiopia.

Ethiop J Health Dev 2002;16(Special Issue):51-58.

15. Tinuade O, Iyabo RA, Durotoye O. Health-careseeking behaviour for childhood illnesses in a resource-poor setting. J Paediatr Child Health 2010:46(5):238-242. [http://dx.doi.org/10.1111/ j.1440-1754.2009.01677.x] 16. Olaogun AA, Adebayo AA, Ayandiran OE, Olasode OA. Effects of mothers’ socio-economic status on the management of febrile conditions in their under five children in a resource limited setting. BMC Int Health Hum Rights 2006;6:1. [http://dx.doi.org/ 10.1186/1472-698X-6-1]

RESEARCH

Audiological findings in a group of neurologically compromised children: A retrospective study K Baillieu,1 MScMed (Child Health Neurodevelopment); K Khoza-Shangase,1 PhD (Audiology); L Jacklin,2 FCP (Paed) (SA), MMed (Paed) epartment of Speech Pathology and Audiology, Faculty of Humanities, School of Human and Community Development, D University of the Witwatersrand, Johannesburg, South Africa 2 Department of Paediatrics, Faculty of Health Sciences, University of the Witwatersrand, Johannesburg, South Africa 1

Corresponding author: K Baillieu (kbaillieu@gmail.com) Background. Hearing loss is more prevalent in developing countries. Later diagnosis of hearing loss will result in delayed access to rehabilitation.

It is typically more difficult to obtain subjective information required in a hearing test from neurocompromised children, causing audiologists to frequently turn to objective measures such as the auditory brainstem response (ABR) measure to obtain this information.

Objective. To describe the ABR results in a group of neurologically compromised children and to establish a relationship between ABR findings and behavioural audiometry results, where these existed.

Methods. A retrospective review was conducted on 40 ABR patient records of neurologically compromised participants aged 5 months to 10 years.

Behavioural audiometry results were sought where these existed. Hearing status was described per ear for objective and behavioural results, and descriptive statistics were conducted.

Results. Behavioural audiometry results were obtained in 72.5% of ears. Results correlated between ABR and behavioural audiometry for only 7.5% of ears, which were all diagnosed with normal hearing. About12.5% of ears were misdiagnosed with behavioural audiometry.

Premature infants were most likely to cope with behavioural audiometry. Hearing loss was highest in participants with cerebral palsy, Down syndrome, prematurity and retroviral disease.

Conclusions. Behavioural audiometry appears to be a largely unreliable method of testing the hearing of children diagnosed with neurological disorders, as results were obtained in only 27.5% of the study sample; however, it remains the gold standard in paediatric hearing testing to evaluate the entire auditory system and provides information on how a child processes sound. Hearing thresholds should be established via objective testing. Conditioning should continue for a behavioural audiological test battery, with adaptations for the child’s developmental ability.

S Afr J Child Health 2016;10(1):20-24. DOI:10.7196/SAJCH.2016.v10i1.936

Background

Early detection of hearing loss

Hearing loss is a common sensory disability, with 800 000 babies born with or acquiring a hearing loss worldwide each year.[1] Ninety percent of children diagnosed each year with a permanent bilateral hearing loss of ≥40 dB reside in developing countries such as South Africa (SA).[1] In SA, the prevalence rate of hearing loss is estimated at 3/1 000 in the private healthcare setting and 6/1 000 in public healthcare.[2] These high numbers highlight the importance of early hearing detection and intervention (EHDI) in developing countries. Documented evidence of EHDI benefits includes better expressive speech and language outcomes,[1] thus decreasing the burden of permanent hearing loss and the limitation of educational opportunities, and ultimately improving long-term outcomes.[2] The Health Professions Council of South Africa (HPCSA) identi­ fied populations at risk of developing hearing loss, who should be screened at birth if they present with recognised conditions.[3] Acquired, late-onset and progressive hearing losses are not identified at birth; therefore, the HPCSA recommends that children with certain conditions associated with hearing loss receive hearing screening between the ages of 29 days and 2 years.[3] The implementation of the HPCSA’s position statement on EHDI has not been successful for a number of reasons. In SA, 85% of the population access public healthcare.[2] Only 7.5% of public hospitals offer infant hearing screening, and only one hospital has been reported to offer universal hearing screening, i.e. <10% of babies born in SA are afforded the opportunity of newborn hearing screening.[1] Certain factors, such as equipment, personnel 20

SAJCH

constraints and burden of disease, which prioritise life-threatening conditions such as HIV/AIDS, are barriers to EHDI.

Early detection of hearing loss and intervention

Many medical conditions in children fall in the category ‘neuro­disability’ – some conditions place affected individuals at higher risk for hearing impairment than the general population, e.g. Down syndrome (DS).[4] Other neurological disorders may result in a speech and language delay.

It is therefore vital that the hearing status of these children be established early so that the benefits of EHDI can be realised.

Hearing function and neurological disorders

Hearing loss is higher in the HIV-infected paediatric population than in the general paediatric population.[5] Hearing loss has been found to be largely conductive as a result of chronic suppurative otitis media,[6] although sensorineural hearing loss (SNHL) has also been documented in higher-resourced areas.[5] Hearing loss is more common in children with DS than in those with other developmental disabilities and healthy children[4] because the former have narrow ear canals[7] and eustachian tube dysfunction.[7] All children with autistic spectrum disorder (ASD) will manifest auditory-related dysfunction of some level.[8] These dysfunctions can include hyperacusis (increased sensitivity to sound), delayed cortical responses to low-frequency sounds, disrupted encoding of simple sounds, difficulty listening in noisy environments, and deafness.[8] Children with cerebral palsy (CP) can have any amount of accompanying hearing loss.[9] Of the infants with birth asphyxia worldwide, approximately 3% have mild hearing loss, and a further

MARCH 2016 Vol. 10 No. 1

RESEARCH 3% severe hearing loss.[10] Similarly, 5 - 10% of premature infants globally are diagnosed with some degree of hearing loss.[11] With regard to children with hydro­ cephalus, 81.5% have some degree of hearing loss,[12] which can improve with surgery to decrease cranial pressure.[12] Among a cohort of meningitic children in Kenya, 43.4% presented with SNHL post meningitis, of whom 61% had mild to moderate hearing loss and 38.7% severe to profound loss.[13]

Assessing hearing function in participants

When testing children’s hearing, it is preferable to use the cross-check principle, where a variety of tests are conducted and the results compared.[14] Behavioural audiometry allows threshold estimation across the entire frequency range for both air and bone conduction, which is essential information for speech and language deve­ lop­ment.[15] Behavioural audiometry also provides an estimate of the entire auditory system from the outer ear to the auditory cortex, thus providing information on cognitive development.[15] Behavioural au­ dio­metry alone is not enough, as the results obtained may be misleading and result in misdiagnosis.[14] A test battery approach is reportedly 20% better to diagnose hearing loss than a single test.[14]

Assessing hearing in special needs populations

Owing to the neurological challenges influen­ cing the ability of children with neurological disorders to be tested behaviourally, objective testing such as the auditory brainstem response (ABR) measure is recommended.[15] The developmental age of the child has an effect on the success of the behavioural audiometry attempt.[16] For children with special needs, certain adaptations can be made to the test battery and tasks to accommodate their needs.[17] Positioning, test stimuli and the required response can be altered, based on motor and play skills of the child.[17] According to the American Speech and Hearing Association (ASHA), for children with developmental disabilities, the diagnosis of hearing loss should rely primarily on objective testing, such as with the ABR.[18] However, immittance audiometry, otoacoustic emissions, a comprehensive case history, behavioural observation, and functional hearing measures should all be conducted or at least attempted to supplement the ABR results.[18] The ABR has been shown to provide a reliable estimate of hearing thresholds for the paediatric population.[18] To optimise the efficiency of the test, different testing protocols may be used so that maximum information regarding the child’s hearing status is obtained in the shortest possible time.[19]

Studies have shown that children with neurological disorders have ABRs that differ from those of children with normal neurological function.[20] Children with ASD generally present with prolonged latencies of wave V and interpeak latencies.[20] The ABR of children with DS has been shown to present with shorter latencies and lower amplitudes.[21] In severe perinatal asphyxia, amplitudes of all waves – particularly wave V – were reduced in the first month of life.[22] In a study in children with hydrocephalus, 88% of the cohort displayed a wide range of abnormalities on ABR.[23] Similarly, children with CP presented with prolonged latencies on ABR testing.[24] Children with meningitis were found to have increased interpeak latencies for waves I - V and I - III, and 14.8% had no visible waveforms at the maximum output of the equipment.[25] These differences should therefore be taken into account when analysing the results in neurologically compromised children; hence the current retrospective review of audiological findings in a group of neurologically com­ promised children.

Objectives

The primary objective of this study was to describe the hearing test results in a group of neurologically compromised participants.

The objectives were to: • describe the audiological ABR findings to determine hearing function in this group • compare audiological ABR and behavioural audiometry findings where these existed • describe hearing results per pathological condition.

Methods

After obtaining the relevant permission, a descriptive retrospective review of patient records was conducted. Permission to review

records was obtained from the relevant autho­ rities, and data collection began only after permission from the pertinent review and ethics committees (protocol No. M120217).

The first 40 audiological records of participants between the ages of 5 months and 10 years at a tertiary hospital in Johannesburg, SA were selected. All testing conducted for patient records was performed by qualified audiologists registered with the HPCSA. Participants were tested during natural sleep, i.e. without sedation, at the tertiary hospital, with one or more of the following disorders: • CP • developmental delay • HIV/AIDS • hydrocephalus • DS • birth asphyxia • ASD • meningitis.

The hearing testing protocol for paediatric patients is depicted in Fig. 1. Participants in the study had their hearing tested according to this protocol.

All hearing results for each participant were documented, including ABR and behavioural audiological testing if available. In cases of multiple attempts at behavioural audiometry, the most recent results were used. Each participant’s results were des­cribed per ear, based on the click and tone-burst ABR thresholds documented on the records. The classification by Gelfand[26] (Table 1) was used to describe the degree of hearing loss, based on the ABR threshold and behavioural results, using headphones, where available. Correction factors used in ABR testing are depicted in Table 2.

Normal hearing in at least one cochlea was

Targeted high-risk screening

Hospital file screened in identified high-risk wards High-risk factors present

No risk factors present No screening indicated

AABR Refer Repeat AABR with tympanogram Pass Refer Refer with type B with type B tympanogram tympanogram Diagnostic ABR Abnormal

Repeat tympanogram in 3/12

Pass Discharge with pamphlet on signs of hearing loss Pass with type A tympanogram Discharge with pamphlet on signs of hearing loss

Normal Discharge

Referral to medical doctor and ABBR in 1/12

Repeat diagnostic ABR or behavioural audiometry Abnormal

Normal

Enrolment in aural habilitation

Follow-up ABR in 1 year

Fig. 1. Protocol for the paediatric hearing test (AABR = automated auditory brainstem response).

21

SAJCH

MARCH 2016 Vol. 10 No. 1

RESEARCH

Behavioural results were obtained in only 22 of the 80 ears tested. As four of the partici­ pants were aged between 5 and 6 months, behaviour­al audiometry was not attempted owing to their developmental capabilities.

For six participants behavioural audiometry was aborted, and for a further six testing occurred in free field – therefore no earspecific results could be obtained.

Half of the ears for which there were behav­ ioural results, were diagnosed with a hearing loss v. only 44% on ABR testing.

In the sample of 80 ears tested, only 7.5% were diagnosed with normal hearing on both behavioural and objective testing.

For 58 ears, no behavioural results were obtained, and for the remaining 16 ears there was a discrepancy between behavioural and ABR results. Of these, 68.8% presented with worse hearing behaviourally than on objective testing. Therefore, 92.5% of the results differed according to behavioural and objective testing, and those that did correlate were in normal hearing ears only.

Occurrence, n (%)

Test method

25 (62.5)

22 ABR only, 5 both, 3 behavioural only

Bilateral symmetrical hearing loss

12 (30)

10 ABR only, 2 behavioural only

Bilateral asymmetrical hearing loss

8 (20)

7 ABR only, 1 behavioural only

Unilateral hearing loss

1 (2.5)

1 ABR only, 0 behavioural only

Profound hearing loss

Children, n

4 2

2

2 0

10

5

nd

ro m

pa lsy ra l

sy

0

15

n

–5

Do w

4 kHz

Ce re b

Click 500 Hz 1 kHz 2 kHz

e

0

Table 2. Correction values for ABR using earphones[27]

1

2

1

2 1

2 0

e

≥90

5

as

Severe hearing loss

5 4

ise

71 - 90

6

s

Moderately severe hearing loss

7 6

ra ld

56 - 70

7

gi ti

Moderate hearing loss

in

41 - 55

8

en

Mild hearing loss

Re tro vi

26 - 40

Total presenting with the pathological condition

M

Slight hearing loss

Behavioural

10

10

lu s

16 - 25

ABR

11

ha

Normal hearing

12

y

<15

The responsiveness of a developmentally delayed participant is asso­ciated with their

Normal hearing

r it

Degree of hearing loss

Diagnosis of hearing status

Hearing status

Table 1. Classification of degree of hearing loss[26] Pure-tone average (dB)

Discussion

Table 3. Hearing status in the sample (N=40)

Hy dr oc ep

The records of 21 males and 19 females were drawn, the average age being 28 months (standard deviation (SD) 24 (5 - 115) months).

As depicted in Table 3, of the records drawn for 40 participants, 62.5% presented with bilateral normal hearing, 30% with bilateral symmetrical hearing loss, 20% with an asymmetrical hearing loss, and one participant was diagnosed with a unilateral hearing loss. In all participants hearing loss was diag­nosed with ABR testing, and some results of patients with normal hearing were corroborated on behavioural testing. One participant presented with a bilateral profound hearing loss when tested behaviourally; however, ABR testing indicated a lesserdegree bilateral asymmetrical hearing loss.

Analysis per ear indicated that for 58 of 80 ears (72.5%), no behavioural audiological results had been obtained. Therefore, for 60% of the participants in this study, no behavioural results were obtained in either ear, and in 25% unilaterally. Half of the 16 participants for whom behavioural results were available

The two most common diagnoses were prematurity (27.5%) and CP (25%). Hydro­ce­ phalus (17.5%), DS (12.5%), meningitis (12.5%) and birth asphyxia (10%) were also common, yielding a combined 52.5%.

More hearing losses were diagnosed on ABR testing, with a total of 22 unilateral or bilateral hearing losses being diagnosed, as depicted in Fig. 2. Only six losses were diagnosed on the same sample with behavioural audiometry. A majority of participants with CP, DS, prematurity and retroviral disease were diagnosed with a hearing loss, as 19 of a combined total of 28 with these conditions presented with a hearing loss objectively. Additionally, 18 participants were diagnosed with normal hearing on ABR testing v. 7 on behavioural audiometry.

The majority of participants with hydrocephalus (6 of 7) and menin­gitis (3 of 5) were diagnosed with normal hearing on ABR testing. For the six main conditions, the participants with DS and HIV could not be tested behaviourally.

Seven of 10 participants with CP, 3 of 5 with meningitis, 4 of 7 with hydrocephalus and 6 of 11 with prematurity could not undergo behavioural audiometry reliably.

ABR compared with behavioural results

at u

Results

Results per neurological condition

presented with normal hearing. Three of these presented with normal hearing bilaterally using headphones; five were tested in free field and therefore had normal hearing in at least one cochlea.

Pr em

classified as ≤25 dB hearing loss on freefield testing. For participants tested in free field, overall hearing status was described and one ear was marked ‘no results’.

Descriptive statistics were conducted per ear for all 40 participants in terms of the incidence and degree of hearing loss, which were compared with the ABR and behavioural results, where available. The frequency of the different neurological diagnoses was established and the hearing results of the six most frequent conditions were analysed according to the degree of hearing loss.

Fig. 2. Number of hearing losses diagnosed on ABR and behavioural audiometry for the six main conditions.

22

SAJCH

MARCH 2016 Vol. 10 No. 1

RESEARCH developmental level, leading to both high false-positive and high false-negative results in behavioural audiometry.[17] In our study, more hearing losses were diagnosed on ABR testing than with behavioural audiometry, indicating a difference in results of the two methods. One participant presented with a profound hearing loss bilaterally on behavioural audiometry and to a lesser degree on ABR testing. If the diagnosis of profound hearing loss was accepted, the participant’s hearing would have been inappropriately amplified with hearing aids, risking further damage to the auditory system. This discrepancy between behavioural and objective findings reiterates that objective testing is particularly important in diagnosing hearing loss in a population with neurological disorders.

Likewise, in only 6 of the 80 ears tested was there a correlation between behavioural and ABR results and all of these ears were of normal hearing. These results suggest that for this study population behavioural audiometry was less sensitive at diagnosing both hearing loss and normal hearing than ABR testing.

Behavioural audiometry

By estimating the participant’s developmental age, behavioural audiometry could be conducted in the same manner as for normaldeveloping participants of the same developmental age.[16] Current findings indicate that no behavioural results could be obtained for a majority in this sample; hence the importance of planning testing batteries based on developmental rather than chronological ages.

It has been well established that behavioural audiometry remains the gold standard in hearing testing, as it provides an estimate of the entire auditory system from the outer ear to the auditory cortex.[15] In half of this study population behavioural audiometry was not attempted.

For a child with a neurological disorder to have the best chance of completing the task required for behavioural audiometry, adap­ tations to the test battery would enhance acquisition of more reliable results.[16] Interpretation of the results should be holistic, ensuring that these are cross-checked.

Difficulty in obtaining behavioural responses

The results of this study correspond to the finding that a child’s ability to attend to a stimulus and nature of the response are dependent on the level of neuromaturation.[15] The child's difficulty in responding and the tester not recognising the response would lead to hearing thresholds appearing worse than the child’s actual hearing abilities.

A greater number of ears with a profound hearing loss were diagnosed on ABR testing than on behavioural audiometry. None of these participants objectively diagnosed with a profound hearing loss had any behavioural results. This could be the result of incorrect conditioning below the ear threshold; therefore, the child will not learn the conditioned response.[16] If a child does not appear to condition under headphones at loud intensities, vibrotactile conditioning with a bone oscillator should be attempted to ascertain whether this lack of response is a potential hearing loss or a result of developmental delay.[16] No participant in this study objectively diagnosed with a profound hearing loss could be tested behaviourally, which emphasises the importance of correct conditioning.[16] Audiology departments at state hospitals service large populations.

Time pressure may therefore lead to audiologists aborting the test and sending the child for an ABR. The audiologist may not include the adaptations and correct conditioning techniques that allow a neurologically disordered child the best opportunity to respond in a behavioural hearing test.

Hearing loss with regard to diagnosis

It is important to know with which condition the child has been diagnosed, as this can assist the audiologist with regard to whether a hearing loss should be expected and what type of hearing loss to 23

SAJCH

anticipate, should it correspond to the diagnosis.[16] The audiologist will also be able to gauge what to expect with speech, language and auditory behaviours, and to estimate the expected cognitive and developmental age of the child.[17] Objective results can also be analysed, taking these factors into account.

In this study sample almost half of the participants were diagnosed with a hearing loss on ABR testing, which is higher than the incidence of hearing loss of 6/100 live births in the general paediatric population.

In this study sample of 5 participants with DS, 4 presented with hearing loss, which is higher than the average of 46.1%.[9] This could be a result of some participants being diagnosed with normal hearing on screening measures in a diagnostic hearing evaluation and therefore being discharged without the requirement for ABR testing.

Participants obtaining ‘refer’ results from the hearing screening at the research site are booked for an ABR and will therefore have a higher chance of presenting with a hearing loss, as they have already been referred on hearing screening.

Of the 11 participants born prematurely, 6 were diagnosed with a hearing loss on ABR testing, which is higher than the global prevalence of 5 - 10%.[11] The higher occurrence of hearing loss in this subgroup emphasises the need for newborn hearing screening.

Universal newborn screening is the gold standard; however, a lack of resources has prevented the instigation of this screening in SA.[28] High-risk or targeted screening should, based on this study, include all premature infants.

Prematurity yielded the smallest number of no responses in behavioural audiometry compared with the five other common conditions. This indicates that these participants are generally able to undergo behavioural audiometry, given extra time to allow for brain matu­ration. However, with the higher occurrence of hearing loss in this premature population than in the general paediatric population, hearing loss would then be diagnosed at a later stage if more time was allowed for maturation. Audiological habilitation for those infants later diagnosed with hearing loss would be less effective.

More than half of the study population with meningitis presented with hearing loss, which reiterates the need to test this population once a diagnosis of meningitis has been made.

The three conditions yielding the least behavioural results are those for which the highest percentage of hearing loss was diagnosed, i.e. CP, retroviral disease and DS. This therefore emphasises the importance of these participants having their hearing tested, and the use of objective testing in obtaining results.

HPCSA risk factors for hearing loss

The HPCSA have identified participants who are at risk of developing a hearing loss and therefore should receive a hearing screen.[2] For this study sample, of the 40 participants selected, only 14 would have been at risk according to the HPCSA criteria. Of the 26 participants with conditions not recognised by the at-risk register, 9 were found to have a hearing loss. Therefore, 22.5% of participants with a hearing loss diagnosed in this study sample would not have received a hearing screen. These were participants with CP, prematurity and birth asphyxia. This indicates the need to revise the criteria for targeted screening at the research site and the HPCSA position statement, so that it is in line with the Joint Committee on Infant Hearing guidelines,[29] which include follow-up screening for all children admitted to the neonatal intensive care unit for >5 days.

Early intervention and neurological disorder

Current findings have shown that hearing loss is more prevalent in populations with neurological disorders than in the general paediatric population. The findings also prove that most of these participants are diagnosed from objective measures, as behavioural audiometry is unreliable in such participants.

MARCH 2016 Vol. 10 No. 1

RESEARCH According to the HPCSA’s position statement,[3] the guidelines for early intervention of hearing loss services in SA for hospital-based settings recommend confirmation diagnostic testing by 3 months of age before enrolment in an early intervention programme before 6 months of age.

In this study, not one participant met these criteria. For many of the participants behavioural audiometry was attempted two or three times before a referral to the ABR clinic. This would delay the confirmation of hearing loss by a number of months.

Conclusion

It is suggested that all patients with neurological disorders receive a hearing screen as soon as a diagnosis of the neurological disorder is made. Those who do not pass hearing screening should be referred directly for objective testing in order to start habilitation. Behavioural audiometry should then be attempted, as this is the only measure that gives an indication of the functioning of the entire auditory system up to the auditory cortex.[15] The child should be conditioned during a number of appointments with the abovementioned adaptations to the conventional behavioural audiological test to meet their individual developmental needs. However, this should be supplementary to enrolment in an early intervention programme to afford the child the best possible language and communication development. Once behavioural results are obtained, amplification devices should be reprogrammed based on the results.

References 1. Friderichs N, Swanepoel D, Hall JW. Efficacy of a community-based infant hearing screening programme utilizing existing clinic personnel in Western Cape, South Africa. Int J Pediatr Otorhinolaryngol 2012;76:552-559.

2. Meyer ME, Swanepoel D, le Roux T, van der Linde M. Early detection of infant hearing loss in the private health care sector of South Africa. Int J Pediatr Otorhinolaryngol 2012;76:698-703.

3. Swanepoel D. Health Professions Council of South Africa Professional Board for Speech, Language and Hearing Professions: Early Hearing Detection and Intervention Programmes in South Africa – Position Statement. Pretoria: Health Professions Council of South Africa, 2007. http://www.hpcsa.co.za/ downloads/speech_education/early_hearing_detection_statement.pdf. (accessed 10 August 2013).

4. Abou-Elhamd KA, ElToukhy HM, Al-Wadaani FA. Syndromes of hearing loss associated with visual loss. Eur Arch Otorhinolaryngol 2014;271(4):635-636. [http://dx.doi.org/10.1007/s00405-013-2514-0] 5. Laughton B, Cornell M, Boivin M, van Rie A. Neurodevelopment in perinatally HIV-infected participants: A concern for adolescence. J Int AIDS Soc 2013;16(18603):1-11. [http://dx.doi.org/10.7448/IAS.16.1.18603] 6. Malt EA, Dahl RC, Haugsand TM, et al. Health and disease in adults with Down syndrome. Tidsskr Nor Laegeforen 2013;133(3):290-294.

7. Austeng ME, Akre H, Øverland B, Abdelnoor M, Falkenberg E, Kværner KJ. Otitis media with effusion in participants with Down’s syndrome. Int J Pediatr Otorhinolaryngol 2013;77:1329-1332. [http://dx.doi.org/10.1016/j. ijporl.2013.05.027]

24

SAJCH

8. Kulesza RJ, Lukose R, Stevens LV. Malformation of the human superior olive in autistic spectrum disorders. Brain Res 2010;1367:360-371. [http://dx.doi. org/10.1016/j.brainres.2010.10.015] 9. Rosenbaum P, Paneth N, Leviton A, Goldstein M, Bax M. A report: The definition and classification of cerebral palsy. Dev Med Child Neurol 2007;49:8-14.

10. Berke J. Birth asphyxia and hearing loss. 2011. http://deafness.about.com/ od/medicalcauses/a/Birth_Asphyxia_And_Hearing_Loss.htm (accessed 2 February 2016).

11. Nour NM. Premature delivery and the millennium development goal. Rev Obstet Gynecol 2012;5(2):100-105.

12. Dixon JF, Jones RO. Hydrocephalus-associated hearing loss and resolution after ventriculostomy. Otolaryngol Head Neck Surg 2012;146:1037-1039. [http:// dx.doi.org/10.1177/0194599811431234].

13. Karanja BW, Oburra HO, Masinde P, Wamalwa D. Risk factors for hearing loss in participants following bacterial meningitis in a tertiary referral hospital. Int J Otolaryngol 2013;2013:354725. [http://dx.doi.org/10.1155/2013/354725] 14. Diefendorf AO, Wynne MK. Paediatric audiology: A test battery approach. In: Kent RD, ed. The MIT Encyclopedia of Communication Disorders. Cambridge, MA: Massachusetts Institute of Technology, 2004:520-522.

15. Swanepoel D, Ebrahim S. Auditory steady-state response and auditory brainstem response thresholds in participants. Eur Arch Otorhinolaryngol 2009;266(2):213-219.

16. Northern JL, Downs MP. Hearing in Participants. 5th ed. Philadelphia, USA: Lippincott Williams and Wilkins, 2002.

17. Madell JR. Evaluation of hearing in participants with special needs. In: Madell JR, Flexer C, eds. Paediatric Audiology. Diagnosis, Technology and Management. New York: Thieme, 2008:82-89.

18. Purdy SC, Kelly AS. Auditory evoked response testing in infants and participants. In: Madell JR, Flexer C, eds. Paediatric Audiology: Diagnosis, Technology and Management. New York: Thieme, 2008:132-145.

19. Newman CW, Sandridge SA. Diagnostic audiology. In: Hughs GB, Pensak ML, eds. Clinical Otology. 3rd ed. New York: Thieme 2007:109-121.

20. Kwon S, Kim J, Choe B-H, Ko C, Park S. Electrophysiologic assessment of central auditory processing by auditory brainstem responses in participants with autism spectrum disorders. J Korean Med Sci 2007;22(4):656-659.

21. Kręcicki T, Zalesska-Kręcicka M, Kubiak K, Gawron W. Brain auditory potentials in children with Down syndrome. Int J Paediatr Otorhinolaryngol 2005;69(5):615-620.

22. Jiang ZD, Brosi DM, Shao XM, Wilkinson AR. Sustained depression of brainstem auditory electrophysiology during the first months in term infants after perinatal asphyxia. Clin Neurophysiol 2008;119:1496-1505.

23. Kraus N, Özdamar Ö, Heydemann PT, Stein L, Reed NL. Auditory brain-stem responses in hydrocephalic patients. Electroen Clin Neuro 1984;59:310-317.

24. Zhu Q, Wang T, Liang J. The evaluation of auditory brainstem response and auditory steady-state response in children with cerebral palsy. J Clin Otorhinolaryngol 2006;20(22):1018-1019.

25. Topcu I, Cüreoqlu S, Yaramiş A, et al. Evaluation of brainstem auditory evoked response audiometry findings in children with tuberculous meningitis at admission. Auris Nasus Larynx 2002;29(1):11-14.

26. Gelfand SA. Pure tone audiometry. In: Gelfand SA, ed. Essentials of Audiology.

3rd ed. New York: Thieme, 2009:127-157.

27. Stevens J, ed. Guidelines for Early Audiological Assessment and Management of Babies Referred from the Newborn Hearing Screening Programme (version 2.5). London, UK: NHS Newborn Hearing Screening Programme, 2011.

28. Delaroche M, Gavilan-Cellie I, Maurice-Tison S, Kpozehouen A, Dauman R. Is behavioural audiometry achievable in infants younger than 6 months of age?

Int J Pediatr Otorhinolaryngol 2011;75:1502-1509.

29. Joint Committee on Infant Hearing. Position statement: Principles and guidelines for early hearing detection and intervention programmes. Pediatrics 2007;120(4):898-921.

MARCH 2016 Vol. 10 No. 1

RESEARCH

Risks for communication delays and disorders in infants in an urban primary healthcare clinic D Claassen, BComm Path; J Pieterse, BComm Path; J van der Linde, MComm Path; E Kruger, MComm Path; B Vinck, PhD Department of Speech-Language Pathology and Audiology, University of Pretoria, South Africa Corresponding author: J van der Linde (jeannie.vanderlinde@up.ac.za) Background. Many risk factors may result in a communication delay. Reliable identification methods are essential to identify infants at risk of communication difficulties in the primary healthcare context. Literature on identifying both at-risk and established risk factors associated with communication disorders in South Africa is limited.

Objective. To identify and describe risk factors for communication delays in infants 0 - 12 months of age at Daspoort Polyclinic in Gauteng.

Methods. A structured interview schedule was utilised to conduct an interview with the caregiver participants. Convenience sampling was used to select 96 caregiver participants.

Results. The results obtained indicated that all infant participants presented with exposure to one or more risk factors that may possibly impact communication development. High frequencies of risk factors included colds and/or flu during pregnancy, previous miscarriages, maternal smoking, low educational levels and unemployment.

Conclusion. The high frequency of at-risk conditions within the Daspoort population justifies the importance of implementation of early communication intervention services in primary healthcare.

S Afr J Child Health 2016;10(1):25-28. DOI:10.7196/SAJCH.2016.v106i1.944

Developmental disorders in infants and toddlers burden the family and society by causing illiteracy, unemployment and low in­ come.[1] Communication delays and disorders can be prevented if risks are identified early.[2,3] Established risk factors include genetic, neurological and sensory disorders and severe toxic exposure that may result in conditions such as fetal alcohol spectrum disorder (FASD).[4,5] South Africa (SA) has a higher prevalence of FASD and HIV/AIDS than other countries in the world.[6,7] The identification of established risk factors for communication delays at birth can ensure that at-risk infants are monitored and integrated into effective early communication intervention programmes.[8] However, infants with established risks within the primary healthcare (PHC) context of SA are not identified and referred for timeous assessment and early intervention.[3] In SA 85% of the population relies on public healthcare services, in which early communication intervention has not been effectively implemented nationally.[9] Identifying at-risk infants early is of importance within the PHC context, as the ultimate goal of early communication intervention is prevention of communication delays and disorders.[3] Environmental risk factors may lead to developmental difficulties, such as communication delays or disorders, later in life.[1] Environ­ mental factors include alcohol and drug abuse,[10] parental neglect and abuse and lack of parent-child interaction.[11] Parents with low education levels, poor mental or physical health and poor coping strategies contribute to the number of at-risk infants.[2] Therefore a lack of parental knowledge and stimulation due to poor knowledge may lead to communication delays and disorders in young children. [2] Other factors that place young children at risk include adolescent mothers, single parents and low-income households.[12] In SA 29.8% of the population is unemployed, resulting in many children being raised in low-income households.[13] Poor living conditions hamper the quality and quantity of prenatal care, placing unborn infants at risk of both preterm birth and low birthweight, which are themselves risk factors for communication delay.[10] The presence of multiple risk factors for communication delays in an infant population warrants the implementation of early 25

SAJCH

identification services in underserved communities in SA. To reduce long-term financial burdens as a result of communication delays, which snowball into social and academic difficulties, at-risk infants must be identified quickly and receive early intervention services. [8] The use of a risk assessment is proposed, in addition to the use of a developmental screening tool, to facilitate early identification of at-risk infants in the PHC context. This will enable professionals from various specialties to identify risk factors associated with communication disorders.

Ideally all children at risk of experiencing possible communi­ cation difficulties should receive an intensive, high-quality early communication assessment and, if deemed necessary, early inter­ vention within the first 3 years of life.[1] Being able to identify risk factors in a reliable and effective manner is one step closer to realising the ideal of implementing early communication inter­ vention in the PHC system.

Objective

The objective of this study was to describe risk factors for communication delays or disorders in infants aged 0 - 12 months in an urban PHC clinic.

Methods

Setting and participants

Daspoort is a suburb situated in Pretoria-West, Gauteng Province, SA. The community consists of 6 355 individuals and 1 582 households. The three official SA languages predominantly spoken in the area are Afrikaans (83.19%), English (6.44%) and Sepedi (1.66%).[13] All the caregivers of infants aged from birth to 12 months who visited Daspoort Polyclinic on Fridays during March and April 2014 were asked to participate in the study. A total of 96 partici­ pants with an equal sex distribution among the infants (54 males and 42 females) were included in the study. The mean age of the infant participants was 4.75 months (standard deviation 3.70) (Table 1).

MARCH 2016 Vol. 10 No. 1

RESEARCH Data collection

A risk assessment tool compiled as part of a master’s study[14] and revised using literature was used to collect data. The tool consisted of five sections including general information, prenatal history, environmental risk factors, perinatal history and established risk con­ ditions. Only risk factors for communi­cation delays/disorders with supporting evidence were included in the tool.[1,8,10,11,15] Data were collected through a structured face-to-face interview with each caregiver. The interview schedule was administered by two final-year Speech-Language Pathology and Audiology students registered with the Health Professions Council of South Africa. Referrals were made to allied healthcare professionals when infants presented with established risk conditions, or when three or more conditions putting them at risk were identified.

Preventive strategies such as developmental screening and parent education on child development and appropriate stimulation of infants were provided upon return after referral. Comprehensive assessment was provided when deemed necessary by the allied healthcare professionals. Infants with one or two at-risk conditions were followed up during health-related or immunisation visits by means of informal general developmental surveillance in addition to the Road to Health booklet screening.

Data analysis

SPSS version 22 (IBM, USA) was used to analyse the data. Descriptive statistics were used to describe the results. Relationships between variables were determined with Pearson product moment correlations.

Correlations between risk factors of 0.3 and ≤0.5 were classified as moderate, and correlations of ≥0.5 were deemed strong.

Many participants (n=42) reported prev­ious miscarriages. A moderate inverse relationship was found between the mother’s age and previous miscarriages (r=–0.306; p=0.002).

The younger the mother, the more miscarriages were experienced.

Of the 32 mothers who indicated that they smoked during pregnancy, an average of 6.6 cigarettes was smoked daily. Also, 14 mothers consumed alcohol during their pregnancies. A moderate correlation between alcohol and drug usage and premature rupture of membranes during pregnancy was found (r=0.338; p=0.001). A family history of hearing loss was reported by 14 of the caregiver participants.

Half of the mothers presented with colds, flu or both during pregnancy. A total of 14 mothers were hospitalised during preg­ nancy as a result of vaginal bleeding, compli­ cations associated with a multiple pregnancy, pneumonia and asthma, round ligament disorder and thrombosis as well as placental, umbilical cord and uterine complications. Preterm birth was reported in 16 pregnancies.

Perinatal and established risk factors

The majority (77%) of the participants re­ ported one or more perinatal risk factor (Table 3). Only two of the infants presented with craniofacial abnormalities: a cleft palate in one and deformational plagiocephaly in another.

Table 1. Characteristics of participants (n=96) Characteristic

n (%)

Sex of baby

Ethics approval

Prior to data collection ethical clearance was obtained from the Research Committee of the Faculty of Humanities at the University of Pretoria. The researchers visited Daspoort Polyclinic one morning per week for 8 weeks.

Caregivers were informed about what the study entailed and voluntary participation was requested. Verbal and written informed consent was obtained from the caregivers.

Male

54 (56.3)

Female

42 (43.8)

Race of the baby Black

36 (37.5)

White

55 (57.3)

Asian

1 (1.0)

Coloured

4 (4.2)

Age of baby (months)

Results

Of the total population (n=96), 13 infants were exposed to ≥10 risk factors, while a further 35 infants had 6 - 9 risks. More than a third (37 infants) presented with 3 - 5 risk factors, whereas 11 infants only had 1 or 2 risks.

0-4

58 (60.4)

5-8

13 (13.5)

9 - 12

25 (26.0)

Relationship to the baby

Prenatal risk factors

Eleven mothers were older than 35 years of age at the birth of their infants (Table 2).

26

Change of hospitals was reported by 14 par­ ticipants; reasons included depleted medi­cal aid, cleft palate, maternal high blood pressure, extreme vaginal bleeding and meconium aspiration. A change in hospitals was also assoc­ iated with infants receiving phototherapy for hyperbilirubinaemia (r=0.452; p<0.001) or for those who had facial abnormalities (r=0.335; p=0.001). Phototherapy was provided to 7 of the infants, for an average of 39 hours, while 3 received exchange blood transfusions for treatment of hyperbilirubinaemia.

Low birthweight was described in 28 of the infants. After birth, 12 of the infants received oxygen for an average of 9 hours. Meningitis had been diagnosed in 3 infants. Other infections reported included laryngitis, Table 2. Prenatal risk factors (N=96) Risk factor

n (%)

Colds and/or flu during pregnancy

48 (50)

Previous miscarriages/stillbirths

42 (44)

Maternal smoking during pregnancy

32 (33)

Pregnancy duration (weeks) >42

17 (18)

32 - 37

15(16)

28 - 32

1 (1)

Maternal alcohol use during pregnancy

14 (15)

Family history of hearing loss

14 (15)

Hospitalisation during pregnancy

14 (15)

No pregnancy education

14 (15)

Mothers ≥35 years at infant’s birth

11 (12)

Maternal diabetes

7 (7)

HIV/AIDS infection during pregnancy

7 (7)

Pre-eclampsia

7 (7)

Premature rupture of membranes

7 (7)

Threatened abortion

6 (6)

Multiple pregnancies (twins)

5 (5)

Mothers <18 years at infant’s birth

4 (4)

Placental complications

3 (3)

Disorders in siblings Congenital disorders

3 (3)

Neurological disorders

2 (2)

Mental disability

1 (1)

Negative rhesus factor

2 (2)

Mother

89 (92.7)

Father

2 (2.1)

Grandmother

3 (3.1)

<2 doctor/clinic visits during pregnancy

1 (1)

Aunt

1 (1.0)

In vitro fertilisation

1 (1)

Caregiver

1 (1.0)

Home births

1 (1)

SAJCH

MARCH 2016 Vol. 10 No. 1

RESEARCH respiratory syncytial virus and otitis media.

The majority of mothers (n=51) preferred breastfeeding, whereas a third (n=32) formula-fed their infants.

A strong association was found between instruments (e.g. forceps and vacuums) used during birth and meconium aspiration (r=0.520; p<0.001). Moderate correlations were found between meconium aspiration and change of hospitals (r=0.302; p=0.003), and meconium aspiration and Apgar score at 1 minute (r=0.308; p=0.002).

Table 3. Perinatal risk factors (N=96)

Environmental risk factors

Most of the mothers (n=75) were unemployed and were the primary caregivers responsible for the daily care of the infant. Half (n=50) of the caregivers reported their highest level of education being Grade 10, while six had tertiary qualifications (Table 4).

The majority of participants (n=58) used public transport or reached the clinic as pedestrians. A third of the participants (n=31) lived in informal housing or with others and had three or more children.

Discussion

Birthweight (1 500 - 2 500 g)

28 (29)

Caesarean section as birth method

24 (25)

Infant received antibiotics after birth

15 (16)

Change of hospitals after birth

14 (15)

Infant received oxygen after birth

12 (13)

Umbilical cord around infant’s neck

12 (13)

Bronchopulmonary dysplasia in an infant

10 (10)

Incubator/warm table

10 (10)

Apnoea and bradycardia in infant

8 (8)

Birth position of infant (breech)

8 (8)

Phototherapy for hyperbilirubinaema

7 (7)

All participants presented with one or more risk factors, while 50% had six or more risks for communication delays or disorders. The higher the number of risk factors, the greater the impact on infant development.[16] The high prevalence of risks confirms the necessity of a reliable screening tool to identify possible communication problems timeously.

Half the participants reported that they had had colds and/or flu during their pregnancies. Pregnant women are at high risk of experiencing infections like influenza (flu).[17] This is of concern as infections can reach the fetus through the feto-placental barrier, which can affect the developing fetus’ health later in life.[18] A mother’s behaviour and general health during a pregnancy has a long-term impact on the health of a child. [19] Mothers should be encouraged to modify their own behaviour during pregnancy in order to ameliorate possible risks.

A disturbing finding was that one-third of the participants had smoked during pregnancy. Smoking during pregnancy is associated with attention-deficit hyper­ activity disorder,[5] preterm birth, low birthweight and infant morbidity and mortality.[20] Also, 15% indicated alcohol

Instruments used during birth

6 (6)

Table 4. Environmental risk factors (N=96)

Infant small for gestational age

6 (6)

Meconium aspiration

5 (5)

Risk factor

n (%)

Neonatal intensive care unit admission

5 (5)

Unemployed mothers

75 (78) 58 (60)

Respiratory distress syndrome

5 (5)

Pedestrian or use of public transport

Apgar score <7 (1 min)

4 (4)

Patent ductus arteriosus

4 (4)

Apgar score <7 (5 min)

3 (3)

Blood transfusion(s)

3 (3)

Meningitis in infant Other infections in infant

Risk factor

n (%)

Feeding method after birth Breastfeeding

51 (53)

Formula feeding

32 (33)

Bottle and breast

12 (13)

Nasogastric tube

1 (1)

Highest qualification ≤ Grade 10

50 (52)

Grade 11 or 12

40 (42)

≥3 other children

31 (32)

3 (3)

Poor/average general health of the mother

24 (25)

3 (3)

Employed mother

21 (22)

Necrotising enterocolitis

2 (2)

Financial support

Infant placed on ventilation

2 (2)

Grandparent/s

19 (20)

Septicaemia in infant

2 (2)

Government grant

8 (8)

Hydrocephalus in infant

1 (1)

Other family members

2 (2)

27

SAJCH

MARCH 2016 Vol. 10 No. 1

consumption during pregnancy, which could have resulted in FASD. Alcohol exposure during pregnancy may have devastating effects on neurological development and can lead to difficulties in a many areas including language, memory, learning, attention, motor coordination, problem-solving skills and abstract thinking.[4,5] A moderate correlation was also found between alcohol and drug usage during pregnancy and premature rupture of membranes.

Creating awareness in mothers on the impact of their health, alcohol use and smoking on their unborn children may help to eliminate these risks. Preterm birth (16%) and low birthweight (29%) were prevalent amongst participants in this study. Infants who present with perinatal risk conditions have a tendency to present with a general developmental delay as well as delayed language and learning development.[21] Previous research revealed that preterm and low birthweight infants present with language delays by the age of 3 years.[21] A disturbingly high percentage of young mothers who had had miscarriages (44%) was noted. Prior miscarriages are associated with preterm delivery and low birthweight.[22] Studies found that women who experienced a previous spontaneous abortion were 2.8 times more likely to have low birthweight babies and 1.7 times more likely to give birth preterm.[22] Thus the higher the number of previous miscarriages, the greater the risk of having a preterm, lowbirthweight infant.[22] Assisted vaginal delivery (AVD) is often associated with fetal distress or incorrect fetal head position, requiring the use of forceps or vacuum extraction.[23] Distressed fetuses may aspirate meconium, but possibly also vernix, blood and pus. Distressed fetuses are prone to aspiration, and AVD is often implemented when fetal distress occurs. This may explain the correlation between instrumentation use during birth and meconium aspiration.

Approximately one-third of the mothers were found to be using formula feeding after birth. The introduction of formula milk before the age of 6 months increases the risk of otitis media,[24] which may result in conductive hearing loss if left untreated. Conductive hearing loss causes infants to miss out on early language learning opportunities and may therefore negatively impact an infant’s communication development.[14] In contrast, breastmilk provides the infant with antibodies to protect against infections, enhances the function of the intestinal barrier and also protects the infant from inflammation.

Low maternal education and higher birth order have been identified as risk factors predictive of potential communication dis­­ orders.[14] Half of the mothers received

RESEARCH schooling only up to Grade 10, this possibly contributing to the high unemployment rate (78%). Unemployment is associated with a lack of sufficient financial resources, requiring supplementary support such as income from grandparents (20%), government grants or other family members. Poverty as such is an indirect cause of communication difficulties, as poverty-associated conditions rather than poverty itself cause communication delays or disorders.[14] Living in poor socioeconomic circumstances with increased environmental stressors has an adverse impact on a child’s development and specifically on language abilities.[14] There was a moderate correlation between change of hospitals after birth and phototherapy for hyperbilirubinaemia. This raises concerns with regard to secondary healthcare/regional hospitals’ ability to render these basic services, as a change in hospital is regarded as a risk in itself. In a developing SA the majority of patients rely on public healthcare services, which are often conservative.[9] Early intervention services in PHC are lacking as a result of limited human and financial resources as well as a lack of equipment, materials and an effective referral framework.[3] These results confirmed a need for early communication inter­ vention in an underserved community. The services required include a risk assessment, communication screening and developmental surveillance and, if deemed necessary, comprehensive assessment and early intervention. It is recommended that future research should compare the outcome of a risk assessment against the outcome of a diagnostic tool.

Study limitations

A possible limitation of the current study involved the use of a structured interview schedule, implying that the researchers had to rely on the caregivers’ report. However, it is widely accepted that parents hold a key position in the early identification and diagnosis of communication disorders within their children.[8,14]

Conclusion

Within this population, where all infants are at risk, regular communication screening and developmental surveillance should occur in order to monitor development. However, all children experiencing a possibility of a communication delay or disorder should receive early communication intervention within the first 3 years of life.[1] The high frequency of at-risk conditions within the Daspoort population justifies the need to implement early communication intervention services in PHC. Finally, preventive strategies such as parental training, communication screening, general developmental surveillance and a risk assessment should be implemented in underserved communities in SA, to support families burdened by communication delays or disorders.

References 1.

Guralnick MJ. Developmental science and preventative intervention for children at environmental risk. Infants Young Child 2013;26(4):270-285. [http://dx.doi.org/10.1097/iyc.0b013e3182a6832f] 2. Harrison LJ, McLeod S. Risk and protective factors associated with speech and language impairment in a nationally representative sample of 4- to 5-yearold children. J Speech Lang Hear Res 2010;53(2):508-529. [http://dx.doi. org/10.1044/1092-4388(2009/08-0086)]

28

SAJCH

3. Van der Linde J, Kritzinger A, Redelinghuys A. The identification process in early communication intervention (ECI) by primary healthcare personnel in Ditsobotla sub-district. South Afr J Commun Disord 2009;56:48-59.

4. Premji S, Benzies K, Serrett K, Hayden KA. Research-based interventions for children and youth with a Fetal Alcohol Spectrum Disorder: Revealing the gap.

Child Care Health Dev 2007;33(4):389-400. [http://dx.doi.org/10.1111/j.13652214.2006.00692.x] 5. Gillberg C, Soderstrom H. Learning disability. Lancet 2003;362(9386):811-821. [http://dx.doi.org/10.1016/S0140-6736(03)14275-4] 6. Rehle TM, Hallett TB, Shisana O, et al. A decline in new HIV infections in South Africa: Estimating HIV incidence from three national HIV surveys in 2002, 2005 and 2008. PLoS One 2010;5(6):e11094. [http://dx.doi.org/10.1371/ journal.pone.0011094 7. Viljoen DL, Gossage PJ, Brooke L, et al. Fetal alcohol syndrome epidemiology in a South African community: A second study of a very high prevalence area. J Stud Alcohol 2005;66(5):593-604. [http://dx.doi.org/10.15288/ jsa.2005.66.593] 8. Rossetti L. Communication Intervention Birth to Three. Vancouver: Singular Thomson Learning, 2001.

9. Swanepoel D, Störbeck C, Friedland P. Early hearing detection and intervention in South Africa. Int J Pediatr Otorhinolaryngol 2009;73(6):783-786. [http:// dx.doi.org/ 10.1016/j.ijporl.2009.01.007] 10. Cone-Wesson B. Prenatal alcohol and cocaine exposure: Influences on cognition, speech, language, and hearing. J Commun Disord 2005;38(4):279302. [http://dx.doi.org/10.1016/j.jcomdis.2005.02.004] 11. Barwick MA, Cohen NJ, Horodezky NB. Infant communication and the mother-infant relationship: The importance of level of risk and construct measurement. Infant Ment Health J 2004;25(3):240-266. [http://dx.doi. org/10.1002/imhj.20000] 12. Qi CH, Kaiser AP, Milan S, Hancock T. Language performance of lowincome African American and European American preschool children on the PPVT–III. Lang Speech Hear Serv Sch 2006;37(1):5-16. [http://dx.doi. org/10.1044/0161-1461(2006/002)] 13. Statistics South Africa. Census 2011. https://www.statssa.gov.za/Census2011/ default.asp (accessed 20 February 2014).

14. Kritzinger AM. Vroeë kommunikasie-ontwikkeling van biologiese risikobabas.

Pretoria: University of Pretoria, 1994.

15. Beitchman JH, Jiang H, Koyama E, et al. Models and determinants of vocabulary growth from kindergarten to adulthood. J Child Psychol Psychiatry 2008;49(6):626-634. [http://dx.doi.org/10.1111/j.1469-7610.2008.01878.x] 16. Paul R, Roth FP. Characterizing and predicting outcomes of communication delays in infants and toddlers: Implications for clinical practice. Lang Speech Hear Serv Sch 2010;42(3):331-340. [http://dx.doi.org/10.1044/01611461(2010/09-0067)] 17. Zerbo O, Iosif A-M, Walker C, Ozonoff S, Hansen RL, Hertz-Picciotto I. Is maternal influenza or fever during pregnancy associated with autism or developmental delays? Results from the CHARGE (Childhood Autism Risks from Genetics and Environment) Study. J Autism Dev Disord 2013;43(1):2533. [http://dx.doi.org/10.1007/s10803-012-1540-x] 18. Murphy VE, Mattes J, Powell H, Baines KJ, Gibson PG. Respiratory viral infections in pregnant women with asthma are associated with wheezing in the first 12 months of life. Pediatr Allergy Immunol 2014;25(2):151-158. [http:// dx.doi.org/10.1111/pai.12156] 19. Tomlinson M, O’Connor MJ, Le Roux IM, et al. Multiple risk factors during pregnancy in South Africa: The need for a horizontal approach to peri­ natal care. Prev Sci 2014;15(3):277-282. [http://dx.doi.org/10.1007/s11121013-0376-8] 20. Machaalani R, Ghazavi E, Hinton T, Waters KA, Hennessy A. Cigarette smoking during pregnancy regulates the expression of specific nicotinic acetylcholine receptor (nAChR) subunits in the human placenta. Toxicol Appl Pharmacol 2014;276(3):204-212. [http://dx.doi.org/10.1016/j.taap.2014.02.015] 21. Schirmer CR, Portuguez MW, Nunes ML. Clinical assessment of language development in children at age 3 years that were born preterm. Arq Neuropsiquiatr 2006;64(4):926-931. [http://dx.doi.org/10.1590/s0004-282x2006000600007] 22. Brown JS, Adera T, Masho SW. Previous abortion and the risk of low birth weight and preterm births. J Epidemiol Community Health 2008;62(1):16-22. [http://dx.doi.org/10.1136/jech.2006.050369] 23. Naz H, Sarosh M, Parveen S, Sultana A. Fetomaternal morbidity associated with vacuum versus forceps delivery. Pak J Surg 2012;28(2):126-129.

24. Abrahams SW, Labbok MH. Breastfeeding and otitis media: A review of recent evidence. Curr Allergy Asthma Rep 2011;11(6):508-512. [http://dx.doi.org/ 10.1007/s11882-011-0218-3]

MARCH 2016 Vol. 10 No. 1

RESEARCH

Haemophilia: A disease of women as well T Naicker,1 MB ChB, DCH, FC Paed; C Aldous,2 PhD; R Thejpal,1 MB ChB, FC Paed, Certificate in Paediatric Clinical Haematology 1 2

Department of Paediatrics, Inkosi Albert Luthuli Central Hospital, Durban, South Africa School of Clinical Medicine, University of KwaZulu-Natal, Durban, South Africa

Corresponding author: T Naicker (thironanaicker@gmail.com)

Background. While haemophilia is a disease phenotype in males only, it has an effect on females too. In South Africa, there is no documentation on the views and experiences of haemophilia carrier mothers regarding the disease, or about their response to carrier testing for daughters or other female members of their families. The burden of child care may fall entirely on the mother in some cultures, and having a son with haemophilia may make coping difficult. Knowing their carrier status would allow daughters of carrier mothers to be aware of the chances of themselves having a son with haemophilia. Knowing their own factor levels may also be useful in understanding possible excessive bleeding in themselves.

Objectives. To record the experiences of haemophilia carrier mothers in KwaZulu-Natal and their attitudes and opinions on carrier testing for female members of their families.

Methods. Forty mothers of haemophiliac sons were interviewed using a structured questionnaire specifically designed and tested for this study.

Results. From this cohort, there were 21 potential carrier daughters and 25 potential carrier sisters who would be eligible for further testing and counselling. All the study participants expressed their desire to have carrier testing available for female family members. They also expressed their concerns regarding raising a son with haemophilia, including some culture-specific issues.

Conclusion. A diagnosis of haemophilia carriership is seen by mothers of haemophiliac sons as important for female members of their families. A protocol for the care of haemophilia carrier women is therefore necessary.

S Afr J Child Health 2016;10(1):29-32. DOI:10.7196/SAJCH.2016.v10i1.961

Haemophilia A is the most common inherited bleeding disorder and affects all population groups. It is an X-linked recessive disease, with carrier mothers’ sons having a 50% risk of being affected and the mothers’ daughters a 50% risk of being carriers.[1] The worldwide prevalence of haemophilia A is estimated to be between 1:5 000 and 1:10 000 men.[2] In South Africa (SA), with a population of ~25 million males,[3] it is estimated that there are between 2 500 and 5 000 haemophiliac men. There are approximately five potential female carriers for each male with haemophilia.[4] There are therefore between 12 500 and 25 000 SA women who are carriers.

In carriers, a wide range in factor VIII levels is seen: from very low, resembling affected males, to the upper limit of normal.[5] This range has been attributed to the phenomenon of lyonisation or random X-chromosome inactivation, which takes place early in embryonic life.[6,7] Carrier women could benefit from knowledge of both their genetic (mutation present or not) and their phenotypic (level of plasma factor activity) status.[1,4] It is vital for a woman to know her carrier status for three important reasons: • The high incidence and intensity of bleeding symptoms of haemophilia A carriers, including bleeding after tooth extraction, easy bruising, postsurgical bleeding and menorrhagia.[8] • The higher risk for primary and secondary postpartum haemorrhage in carrier mothers.

• The risk in carrier mothers for haemorrhage in their newborn male infants who may not yet be diagnosed, especially scalp and intracranial bleeds.[9] Previously, pedigree analysis and clotting factor VIII levels were used to diagnose carriership in haemophilia.[7] However, since the publication of the sequence of the factor VIII gene in 1984, a large number of mutations that cause haemophilia A have been identified and have provided an avenue for confirmatory genetic testing.

During the last three decades, genetic counselling, carrier testing and prenatal diagnosis of haemophilia have become an integral part 29

SAJCH

of the comprehensive care of haemophilia.[1,4] Prenatal diagnosis of haemophilia is generally indicated in families with severe or moderate forms of haemophilia.[7,10] Prenatal testing can be done for carrier women if the mutation in the family has been identified.[4] The issue of carrier testing for an X-linked recessive genetic disease in minors is a contentious one, and is normally offered when a patient turns 18 years old.[11] However, with haemophilia, a carrier woman can show symptoms of bleeding and may be clinically affected during her childhood. It has been suggested that it would be in the best interest of the girl child to know her carrier status earlier, as it would be of medical benefit to her.[12] In SA, there is no documentation on the views and experiences of haemophilia carrier mothers regarding the disease or about their response to carrier testing for daughters or other female members of their families. While in some provinces such as Gauteng and the Western Cape carrier testing is provided, mothers at King Edward VIII Hospital, a busy tertiary hospital in KwaZulu-Natal (KZN), have not had access to this service. King Edward VIII Hospital is a public referral centre in the Durban metropolitan area providing regional and tertiary services to Region 1 of KZN and parts of the Eastern Cape. King Edward VIII is a 922-bed hospital with around 360 000 outpatients annually. It serves a mainly indigent and isiZuluspeaking population.

Methods

The authors designed a structured interview questionnaire con­ taining a total of 56 questions, which covered information about general demographics, the carrier mother’s affected child, personal information and the mother’s views on carrier testing for other women or girls who had not yet had children. The interview questionnaire was piloted at the clinic and amendments were made.

Mothers attending the haemophilia clinic for the first time were excluded because at this initial visit the child is assessed and counselling is conducted by the haemophilia clinic sister. This

MARCH 2016 Vol. 10 No. 1

RESEARCH first visit is time consuming and patients need the opportunity to assimilate the newly received information. From March 2012 until April 2014, 40 mothers who attended the haemophilia clinic at King Edward VIII Hospital were interviewed and a pedigree was drawn for each family.

Ethical approval for the study was given by the Biomedical Research and Ethics Committee at the University of KZN. Respondents’ informed written consent was obtained for participation in the study and interviews were conducted in private rooms at the study site.

Results

Demographics

The majority of mothers (n=27) in our study were of African descent, 10 were Indian, 2 coloured and 1 white. The mothers’ ages ranged from 23 to 66 years, with the average being 34 years. Only 18 mothers lived in family units together with the child’s biological father. The father remained a part of his son’s life in 29 cases.

The impact of the diagnosis on the family

Half of the cohort had a positive family history of haemophilia (Fig. 1). Interestingly, one mother learnt about the illness during a biology class at school and recognised similar symptoms in her brother. She convinced her mother to get her brother investigated, hence uncovering the haemophilia gene in her family.

were still unaware of their results. The majority of mothers (n=37) were informed that they were carriers based on pedigree and not formal molecular testing.

Mothers discovered at an average age of 26 years that they were carriers, but the median age was 30 years. Child-bearing decisions were affected based on this knowledge and 30 mothers stated that they did not intend having further children. Thirty-three women felt that a carrier mother should limit the number of children they bear.

The haemophilia clinic nurse counselled the mothers and dis­ cussed the inheritance with them. Despite understanding the gene­ tics of haemophilia, half of the mothers expressed feelings of guilt and felt that they were responsible for giving this illness to their child.

Twenty-seven mothers had first-pregnancy sons, of whom 23 were diagnosed with haemophilia. Eight mothers who had first-pregnancy haemophiliacs had no further children.

In eight families there was another son who was also diagnosed with haemophilia. Three of these boys had died from a haemophiliarelated problem. There was one mother who had three sons and all were haemophiliacs (Table 1).

Table 1. Information from family pedigrees Carrier (N=40) Age of haemophilia carrier mother (years), mean (range)

Brother (n=10) Father (n=8) Cousin (n=1) Nephew (n=1)

26 (23 - 66)

Age of mother when first child is diagnosed with haemophilia (years), mean

23

No further children after a haemophiliac son, n

8

More than one haemophiliac son, n

8

Potential carrier daughters, n

21

Potential carrier sisters, n

25

Obligate carrier sisters (haemophiliac fathers), n

5

Fig. 1. Family member affected by haemophilia.

At diagnosis, the impact of haemophilia on the family was interpreted as a major stressor in the majority (n=33) of cases. Reasons cited included lack of prior knowledge of haemophilia, that it was a lifelong and chronic illness, and that their child was restricted when playing.

With time, 31 mothers felt that living with haemophilia became more manageable and became less stressful on their family lives.

This was accounted for by counselling from a healthcare worker who assisted their understanding of the illness, their learning how to administer factor, and that with time their sons experienced fewer bleeds, which made the situation easier to cope with.

Only 11 mothers with a family history of haemophilia felt that witnessing a relative go through the trials of the disease made it easier to handle the situation with their own son. They felt they were more experienced and educated about haemophilia, that it was a familiar entity to them and that they were emotionally prepared.

Mothers who were not coping stated they feared the sight of blood and were concerned about death, and one mother thought that she needed antidepressants. Mothers reported that siblings felt they were afforded less attention when the ill child needed care.

Many mothers (n=34) experienced financial strain when visiting hospital owing to the high cost of travelling long distances from their home to the clinic.

The impact of haemophilia on the carrier mother

Menorrhagia was experienced by 22 of the mothers. Only two mothers had been investigated for abnormal bleeding; however, they 30

SAJCH

Most mothers (n=34) perceived their sons to be normal with a ‘blood disorder’. Six mothers commented: ‘He is delicate, he needs to be careful and he is restricted from doing things other normal children do.’

Most mothers (n=31) received emotional support from the child’s father following the diagnosis. Even though half of the mothers were aware of a family history of haemophilia, only two fathers were aware of the risk of having a haemophiliac son prior to pregnancy.

Six women in our cohort experienced negative comments from the child’s father and in-laws regarding assignment of blame, such as ‘This never existed in our family until you came along.’ Four mothers were blamed by their partners for bringing this illness into the family.

Two of these fathers subsequently abandoned their families as they could not cope with their sons’ illness.

Use of traditional medicine

Eleven families felt that the child was bewitched because he suffered nose bleeds or swollen joints. Eight families initially used traditional/ herbal medication to treat the bleeding.

The affected child

The average age of diagnosis of the haemophiliac sons was 24 (range 0 - 72) months. Of the four babies diagnosed in the neonatal period, three had positive family histories, which resulted in earlier detection; the other was detected post circumcision. Of the 29 haemophiliacs of school-going age, only 15 were in the correct grade for their age.

MARCH 2016 Vol. 10 No. 1

RESEARCH Joint and mouth bleeds, as well as extensive bruising, were the most common first presentations. Over the years, 12 boys had more than 10 hospital admissions each, often for joint-related bleeds. Four boys had never been admitted prior to the time of this study.

Mothers’ attitudes to carrier testing

All the mothers felt that carrier testing is of vital importance. Being equipped with the knowledge of one’s carrier status of haemophilia allows a carrier to make informed choices regarding future child bearing. When questioned about their advice to potential carrier daughters regarding starting a family, nine mothers would discourage their daughters from having children, and seven would advise them to limit the number of children to a maximum of two. The rest of the mothers felt child-bearing decisions should be their daughters’ own choice, after knowing their carrier status.

Thirty mothers felt other females in their family may want to be tested for carriership. Twenty-one mothers had potential carrier daughters and there were 25 possible carrier sisters. Only six respondents’ families had ever discussed carrier testing before and these women were open to carrier testing. Over half (n=25) of the mothers felt that a carrier should not be forced to take a test and that it should be a carrier’s right to choose if she wanted to be tested.

The age cited by just over half (n=21) the mothers as the best time to undergo carrier testing is during one’s teenage years, with 16 favouring early childhood followed by three who favoured adulthood. The overwhelming reason was that a carrier should know her status prior to any pregnancy.

Antenatal testing via amniocentesis was felt to be a positive intervention and almost all (n=37) mothers would advise relatives to have such testing done. Reasons given included knowing their son is affected prior to delivery would enable them to make decisions regarding continuation of the pregnancy, although the majority of mothers (n=33) would not consider termination of pregnancy for a severe haemophiliac fetus.

Discussion

The purpose of the study was to document the experiences of haemophilia carrier mothers at a busy tertiary hospital in KZN regarding carrier testing, prenatal diagnosis, raising a haemophiliac son, genetic counselling and the effect of being a carrier. We also obtained their opinion on potential carrier testing with particular regard to their daughters and other possible carriers.

Gender plays an important role in shaping the burden of care in the African context. Women are more likely than men to take on caregiving activities.[13] In SA, most mothers assume almost exclusive responsibility for child rearing and households. In our study, only 29 fathers were involved in their sons’ lives. During the interview process, only one father accompanied the mother to the Haemophilia Clinic for treatment or counselling. Support from partners is vital to coping and living with haemophilia. Parents’ quality of life can be shaped by perceived emotional strains and worries concerning the child’s future. Mothers who are not supported by their partners risk loss of quality of life.[14] The decision not to have more children draws attention to the emotional, psychological and physical burden of the disease on a carrier. This finding is similar to that of a Swedish study by Tedgård et al.,[15] where carriers who did not choose prenatal diagnosis often abstained from further pregnancies after the birth of a haemophiliac child, and they had significantly fewer children than the remainder of the carriers.[15] In SA, preconception genetic counselling is rare. Twenty mothers in our cohort had a family history of haemophilia; however, only two chose to disclose this information to their partners. One has to question whether the genetic counselling information was correctly relayed to members of the affected family or influenced by other 31

SAJCH

factors related to disclosure. Many mothers may not have considered themselves to be at risk of conceiving a child with haemophilia, despite knowing about the family history. The average age of finding out they were carriers was 26 years old, usually after the diagnosis in their son was made.

In addition to genetic counselling, psychological counselling by a psychologist or social worker needs to be offered to mothers and families who are grappling with the chronic nature of the illness in their sons.

Nine mothers felt that they were not coping with their boys’ haemophilia years after the initial diagnosis was made. Family counselling would also benefit siblings who feel marginalised, as less time was dedicated to their needs. Families, including siblings, need specific interventions to assist with the emotional, financial and physical stress that they experience when dealing with a chronically ill child.[16] Mothers need to equip themselves with knowledge about the illness and how it affects their families. In our clinic, emphasis was placed on treatment in the affected haemophiliac, and not on prevention or detection in future generations. A paradigm shift needs to occur among carrier mothers.

From the international literature, it has been cited that 89% of women considered carrier testing to be a positive thing and 49% felt that the best time to test would be in teenage years.[1] In our cohort, all women (100%) felt carrier testing to be of vital importance and 53% felt that the teenage years would be most suitable to have the test.

Conclusion

Being a haemophilia carrier mother affects many aspects of life, own health in terms of possible excessive bleeding, reproductive choices and perception of self as a responsible parent. A young carrier female should receive education and genetic counselling about the illness, reproductive choices and future prospects of raising a child with haemophilia.

Our study highlights the need for accessible carrier testing in KZN. We need to educate the community about the benefits of carrier testing and implement a strategy to improve carrier screening and detection in KZN.

Recommendations

As a result of the study we have developed the following recommen­ dations for a protocol for the care of potential carrier haemophilia women.

Once a diagnosis of haemophilia is made in her child, the mother should receive the following: 1. Counselling from trained medical staff regarding the condition, the inheritance pattern, how it affects her son and other family members 2. Blood tests for factor levels and molecular carrier detection in the mother (her genotype) 3. A pedigree/family tree 4. A contact and carrier tracing card which she can give to other potential carrier women in her family, enabling and empowering them to get tested 5.

Once her results are ready, counselling on the implications regarding bleeding, child bearing, etc.

6. R eferral to a local haemophilia support group 7. Access to prenatal counselling for haemophilia carriers and testing when required at a comprehensive haemophilia care clinic.

Acknowledgments. We acknowledge Prof. Miriam Adhikari for review­ ing the article.

References 1. Ranta S, Lehesjoki AE. Hemophilia A: Experiences and attitudes of mothers, sisters and daughters. Pedriatr Hematol Onc 1994;11(4):387-397.

MARCH 2016 Vol. 10 No. 1

RESEARCH 2. National Human Genome Research Institute. Learning About Haemophilia.

September 2011. http://www.genome.gov/20019697 (accessed 13 August 2014).

3. Statistics South Africa. Mid-year Population Estimates. July 2014. http://www. statssa.co.za (accessed 30 March 2015).§ 4. Street AM, Ljung R, Lavery SA. Management of carriers and babies with haemophilia. Haemophilia 2008;14(Suppl 3):181-187. [http://dx.doi.org/10.1111/ j.1365-2516.2008.01721.x] 5. Ay C, Thom K, Abu-Hamdeh F, et al. Determinants of factor VIII plasma levels in carriers of haemophilia A and in control women. Haemophilia 2010;16(1):111-117. [http://dx.doi.org/10.1111/j.1365-2516.2009.02108.x] 6. Renault NK, Dyack S, Dobson MJ, Costa T, Lam WL, Greer WL. Heritable skewed X-chromosome inactivation leads to haemophilia A expression in heterozygous females. Eur J Hum Genet 2007;15(6):628-637. [http://dx.doi. org/10.1038/sj.ejhg.5201799] 7. Husain N. Carrier analysis for hemophilia A: Ideal versus acceptable. Expert Rev Mol Diagn 2009;9(3):203-207. [http://dx.doi.org/0.1586/erm.09.3] 8. Miesbach W, Alesci S, Geisen C, Oldenburg J. Association between phenotype and genotype in carriers of haemophilia A. Haemophilia 2011;17(2):246-251. [http://dx.doi.org/10.1111/j.1365-2516.2010.02426.x] 9. Hooper WC, Miller CH, Key NS. Complications associated with carrier status among people with blood disorders: A commentary. Am J Prev Med 2010;38(Suppl 4):S456-S458. [http://dx.doi.org/10.1016/j.amepre.2010.01.009]

32

SAJCH

10. Ghosh K, Shetty S, Tulsiani M. Evolution of prenatal diagnostic techniques from phenotypic diagnosis to gene arrays: Its likely impact on prenatal diagnosis of hemophilia. Clin Appl Thromb Hemost 2009;15(3):277-282.

11. Borry P, Fryns JP, Schotsmans P, Dierickx K. Carrier testing in minors: A systematic review of guidelines and position papers. Eur J Hum Genet 2006;14(2):133-138. [http://dx.doi.org/10.1177/1076029607308870] 12. Borry P, Evers-Kiebooms G, Cornel MC, Clarke A, Dierickx K on behalf of the Public and Professional Policy Committee (PPPC) of the European Society of Human Genetics. Genetic testing in asymptomatic minors. Eur J Hum Genet 2009;17:711-719. [http://dx.doi.org/10.1038/ejhg.2009.25] 13. Schatz EJ. ‘Taking care of my own blood’: Older women’s relationships to their households in rural South Africa. Scand J Public Health 2007;69:147-154. [http://dx.doi.org/10.1080/14034950701355676] 14. Wiedebusch S, Pollmann H, Siegmund B, Muthny FA. Quality of life, psychosocial strains and coping in parents of children with haemo­ philia. Haemophilia 2008;14(5):1014-1022. [http://dx.doi.org/10.1111/j.13652516.2008.01803.x] 15. Tedgård U, Ljung R, McNeil TF. Reproductive choices of haemophilia carriers.

Br J Haematol 1999;106(2):421-426.

16. Anderson T, Davis C. Evidence-based practice with families of chronically ill children: A critical literature review. J Evid Based Soc Work 2011;8(4):416-425. [http://dx.doi.org/10.1080/15433710903269172]

MARCH 2016 Vol. 10 No. 1

RESEARCH

Serious bacterial infections in febrile young children: Lack of value of biomarkers M Karsas, MB ChB; P J Becker, PhD; R J Green, PhD, DSc Department of Paediatrics and Child Health, School of Medicine, Faculty of Health Sciences, University of Pretoria, South Africa Corresponding author: M Karsas (mkarsas@hotmail.com) Background. Serious infections in children are difficult to determine from symptoms and signs alone. Fever is both a marker of insignificant viral infection, as well as more serious bacterial sepsis. Therefore, seeking markers of invasive disease, as well as culture positivity for organisms, has been a goal of paediatricians for many years. In addition, the avoidance of unnecessary antibiotics is important in this time of emerging multiresistant micro-organisms.

Objective. To ascertain whether acute-phase reactant tests predict positive culture results.

Methods. A prospective, cross-sectional study over a 1-year period included all documented febrile children under the age of 5 years (with an axillary temperature ≥38°C) who presented to Steve Biko Academic Hospital, Pretoria, with signs and symptoms of pneumonia, meningitis and/or generalised sepsis. Every child had clinical signs, chest radiograph findings, urine culture, blood testing (full blood count, C-reactive protein, procalcitonin) and blood culture results recorded.

Results. A total of 63 patients were enrolled, all of whom had an axillary temperature ≥38°C. C-reactive protein, procalcitonin and white cell count did not predict the presence of positive blood culture or cerebrospinal fluid culture results, nor infiltrates on chest radiographs.

No statistically significant correlations were found between the duration of hospital stay and the degree of fever (p=0.123), white cell count (p=0.611), C-reactive protein (p=0.863) or procalcitonin (p=0.392).

Conclusion. Biomarkers do not seem to predict severity of infection, source of infection, or duration of hospitalisation in children presenting to hospital with fever. The sample size is however too small to definitively confirm this viewpoint. This study suggests that clinical suspicion of serious infection and appropriate action are as valuable as extensive testing.

S Afr J Child Health 2016;10(1):33-36. DOI:10.7196/SAJCH.2016.v10i1.980

Young children often attend primary care institutions and emergency departments with acute infections. Most of these children have selflimiting conditions; however, a small proportion have serious or even life-threatening infections. This may be a source of anxiety for parents and may present a challenge to attending clinicians.[1] There is reasonably good evidence for the diagnostic value of clinical features for certain conditions, namely pneumonia, and to a lesser degree, meningitis. However, little is known about the clinical features predictive of serious outcome for febrile children presenting with nonspecific symptoms and signs with no clear focus of infection.[1] Various clinical tools have been developed in order to score febrile children as a means to predict the severity of illness, but these scores have been found to be nonspecific and of limited use in clinical practice, especially when used as positive predictors of serious bacterial infection (SBI).[2] Fever in young children may represent both insignificant viral infections and SBI which is often not associated with a distinguish­ able source of infection.[3] Seeking markers of invasive disease, as well as culture positivity for organisms, has been a goal of paediatricians for many years.

The evidence thus far, for multiple site testing, as well as multiple testing methods to detect SBIs in febrile young children, is unclear.

In addition to selecting antibiotics for appropriate infections, the avoidance of unnecessary antibiotic use should also be considered important in this time of emerging multiresistant micro-organisms.

This is an important aspect of antibiotic stewardship.[4]

Methods

A prospective, cross-sectional study over a 1-year period (1 December 2013 - 30 November 2014) of children presenting to Steve Biko Academic Hospital (SBAH), Pretoria, was conducted.

33

SAJCH

The study sampling strategy included paediatric patients requiring admission who met inclusion criteria and presented to casualty, the outpatients department or directly to the wards.

Approval was obtained from the Department of Paediatrics and Superintendent of SBAH, as well as the MMed and Ethics committees of the University of Pretoria. Informed consent was obtained from the parents/guardians of each participant.

SBAH is a large academic tertiary hospital located in Gauteng. It has two paediatric wards, with a maximum capacity of 56 beds, and one 7-bed paediatric ICU. All patients require referral from a clinic, medical practitioner or hospital. Patients not requiring tertiary care are down-referred to Tshwane District Hospital. The number of under-5 patients admitted per year is approximately 1 400 - 1 500.

Three-quarters (~1 000) are admitted for subspecialist care and tend to be afebrile on admission; an eighth (~200) are neonates (excluded from the study), while 50% of the remaining 200 - 300 patients have low-grade fevers <38°C on admission (having been given prior antipyretics). Therefore the total number of children under 5 years of age expected for this cohort was 100 - 150 participants.

This study included all documented febrile patients (axillary temperature ≥38°C) who presented to SBAH between the ages of 1 month and 5 years, with signs and symptoms of pneumonia, meningitis and/or generalised sepsis. Exclusion criteria for this study were children over the age of 5 years, neonates (less than 1 month of age), admission temperature <38°C, and children with exclusion criteria for a lumbar puncture (LP), namely focal neurological signs, papilloedema, readily deteriorating consciousness or Glasgow Coma Scale (GCS) <8, signs of raised intracranial pressure (falling pulse, rising blood pressure, dilating or poorly reacting pupils), continuous seizure activity, bleeding diathesis and neural-tube defects.[3] Data collected included all the clinical, laboratory, radiological and microbiology findings. Data were collected on the day of

MARCH 2016 Vol. 10 No. 1

RESEARCH admission, and updated as laboratory and microbiology results became available. In order to attempt to overcome the difference in clinical experience between admitting clinicians, specific signs and symptoms were listed in the data collection table that clinicians were instructed to document if present or absent on the day of admission.

Patient age, gender, admission axillary temperature (°C) and urine dipstick findings were also included in the captured data.

A 3-day cut-off value for the duration of hospitalisation was chosen for the purpose of this study as the median duration of hospitalisation for children at SBAH is 3 days. We expect those patients with severe illness to have a longer-than-median hospitalisation time.

A urine dipstick test was deemed positive in this study in the presence of leucocytes and/or nitrites. White cell count (WCC) was deemed to be increased if the WCC value was >17 × 109/L in children aged 1 - 12 months, and if >15 × 109/L in children 13 months 5 years.[5] Leucopenia was defined as WCC <4 × 109/L. A C-reactive protein (CRP) value was deemed positive/predictive of bacterial infection if CRP >40 mg/L[6] and procalcitonin (PCT) was positive if >0.2 µg/L.[7] Cerebrospinal fluid (CSF) findings were considered positive based either on biochemistry suggestive of meningitis, or on a positive CSF culture or PCR.[8] Chest radiographs (CXRs) were interpreted by the lead investigator using the World Health Organization (WHO) CXR interpretation methodology.[9] A CXR was deemed positive for hyperinflation in the presence of >8 visible posterior ribs.[9] This study also noted some combinations of both pneumonic changes and hyperinflation.

Statistical analysis of data

The primary data analysis focused on the proportion of children under the age of 5 years who were admitted to hospital and remained hospitalised for a period of 3 days or longer. Secondary data analysis was the agreement between the clinical picture and individual biomarkers, as well as among the individual biomarkers themselves.

All parametric data were analysed using a t-test and nonparametric data were analysed by means of Wilcoxon rank sum test using Stata-13 (StataCorp, USA).

The study was adequately powered.

Results

Sixty-three patients were enrolled (age range 1 month - 4.5 years), and 33 (52.4%) were male.

All children whose HIV status was unknown or showed signs of being clinically immunocompromised were tested for HIV infection.

Eight (12.7%) patients were confirmed to be HIV-positive, in 5 (62.5%) a significant organism was identified and in 2 (25.0%) multiple significant organisms were cultured. Two (3.2%) of the 63 patients were assessed as having malnutrition. Both patients were moderately acutely malnourished, and no organisms were cultured in either patient.

Thirty-seven (58.7%) patients were hospitalised for ≥3 days. The median temperature for all patients on admission (irrespective of duration of hospitalisation) was 38.4°C (range 38 - 40°C). There was no statistical significance between temperature on admission and duration of stay (p=0.123).

An organism was cultured from the blood in 13 (25.5%) of the 51 patients in whom a blood culture was performed. All cultures were performed on admission. Commensal flora were isolated from 11 (84.6%) of the 13 positive blood cultures (10 coagu­lase-negative Staphylococcus spp., 1 Micrococcus sp.). The two significant blood cultures isolated extended-spectrum beta-lactamase-producing Klebsiella pneumoniae and Salmonella Group D. A significant organism was identified from any site in 25 (39.7%) of the 63 patients included in this analysis. Sampled sites included blood (8/51, 15.7%), stool (5/16, 31.3%), urine (7/34, 20.6%), CSF 34

SAJCH

(2/22, 9.1%), respiratory secretions (8/30, 26.7%) or tissue (3/3, 100%). Among these 25 patients, significant multiple organisms were identified in 9 (36.0%). Thus, significant multiple organisms were cultured in 14.3% of the total patients.

Nine subjects in total had an elevated WCC, and of these two (22.2%) had positive insignificant blood cultures. None of the children with a decreased WCC had a positive blood culture. There was no correlation between temperature, WCC, CRP or PCT on admission and blood culture results (p=0.521; p=0.919; p=0.186; p=0.833, respectively). There was no correlation between duration of hospitalisation and blood culture results (p=0.560).

Of the 63 enrolled patients, 54 (85.7%) had a CXR. Of these 54 CXRs, 42 (77.8%) revealed positive findings. Forty-six (73.0%) of the 63 enrolled patients had respiratory symptoms and 35 (76.1%) of them had positive CXR findings. In the 35 children with respiratory symptoms and positive CXR findings, 16 (45.7%) had pneumonic changes, 11 (31.4%) had hyperinflation on CXR, and pneumonic change with hyperinflation was found in 8 (22.9%).

Of the 54 children who had CXRs, 30 (55.6%) had pneumonic changes. Nasopharyngeal aspirates (NPAs) were collected from 10 (33.3%) of the 30 children with pneumonic changes on CXR, 3 (30.0%) of which were positive for respiratory syncytial virus (RSV) and 1 (10.0%) was positive for Pneumocystis jiroveci (on immunoflourescence). Ten (33.3%) of the 30 children with pneu­ monic changes on CXR had induced sputum specimens collected, one of which (in an HIV-positive child) was positive for acidalcohol-fast bacilli (AFB) and one positive for Candida albicans (HIV-negative child). Twenty (37.0%) of the 54 children who had a CXR had positive findings of hyperinflation. NPAs were conducted in 5 (25.0%) of the 20 children who had hyperinflated CXRs, and 2 of these were positive, 1 for RSV and 1 for parainfluenza virus type 3.

Seven (70.0%) of the 10 children who had an elevated WCC and CXR performed had positive CXR findings. Both patients with a decreased WCC had radiographic changes. The proportion of all CXRs with positive findings (n=42) with and without elevated or decreased WCC were compared with those with normal CXRs (n=12). There was no correlation between WCC on admission and CXR findings of pneumonia or hyperinflation (p=0.145 and p=0.669, respectively).

A positive WCC (elevated or decreased WCC) in combination with positive CSF findings was found in 1 of the 3 patients who had a positive WCC and an LP. These positive CSF findings were based on biochemistry suggestive of meningitis, even though they had a negative CSF culture and PCR. Twenty-two LPs were conducted in this cohort.

There was no correlation between WCC on admission and duration of hospital stay (p=0.471).

CRP was measured in 59 (93.7%) of the 63 patients enrolled in the study. A positive CRP (defined as a CRP >40 mg/L)[6] was found in 25 (42.4%) patients. Median (range) CRP was 31.0 (<1 to 336) mg/L. There was no correlation between CRP on admission and duration of stay (p=0.863). A positive CRP in combination with positive CXR findings was found in 15 (71.4%) of the 21 patients who had a positive CRP and had had a CXR. There was no correlation between CRP on admission and CXR findings of hyperinflation (p=0.087) or pneumonia (p=0.368).

A positive CRP in combination with a significant positive blood culture result was found in 2 (9.5%) of the 21 patients who had a positive CRP and in whom a blood culture was performed. A positive CRP in combination with positive CSF findings was found in 6 (60.0%) of 10 patients who had a positive CRP and in whom an LP was performed. Five out of the 6 positive CSF findings were based on biochemistry suggestive of meningitis. One was positive based on CSF PCR for enterovirus.

MARCH 2016 Vol. 10 No. 1

RESEARCH PCT was measured in 25 (39.7%) of the 63 enrolled patients. The clinical profile of these patients varied greatly, and 11 (44.0%) were critically ill, requiring paediatric intensive care unit (PICU) admission. PCT was found to be positive (>0.2 µg/L)[7] in 21 (84.0%) of the 25 patients, 10 (47.6%) of whom were admitted to ICU. Five (50.0%) of these 10 died. Median (range) PCT was 5.7 (0 - 728) μg/L. Sixteen (76.2%) of the 21 children with a positive PCT remained hospitalised for 5 days or longer; however, statistically there was no correlation between PCT on admission and duration of stay (p=0.392).

A positive PCT in combination with positive CXR findings was found in 16 (80.0%) of the 20 patients who had a positive PCT and in whom a CXR was performed. A positive PCT in combination with a significant positive blood culture result occurred in 1 (4.8%) of the 21 patients who had a positive PCT and in whom a blood culture was performed. A positive PCT in combination with positive CSF findings occurred in 4 (50.0%) of the 8 patients who had a positive PCT and in whom an LP was performed. Three (75.0%) of the 4 were positive based on biochemistry suggestive of meningitis; 1 (25.0%) was positive based on CSF PCR for enterovirus.

All of the children enrolled in the cohort had urine dipsticks evaluated at admission, and 12 (19%) had positive urine dipsticks for leucocytes or nitrites. All of the urine specimens from children with positive urine dipsticks were culture-negative. Two (16.7%) of the children with positive urine dipsticks had viral pathogens isolated on stool specimens (positive on stool ELISA): adenovirus (n=1) and adenovirus and rotavirus (n=1). In 8 (15.7%) of the children with negative urine dipsticks an organism in the urine was cultured.

Of the 63 total enrolled patients, 11 (17.5%) required PICU admission and 5 (7.9%) patients died. Four of the five deaths occurred after the first 3 days of hospitalisation, with the median time to death being 10 days (range 2 - 20 days).

Discussion

The value of biomarkers for determining SBI in febrile children has revealed conflicting results. Some studies suggest that individual tests perform better than others or than clinical judgement of bacterial v. viral infection, while other studies do not.[3,10-13] The insensitivity of routine microbiological methodologies in identifying bacterial infections, particularly bacteraemia, is well described, and molecular diagnostic techniques may in fact be superior to acute-phase reactants in detecting SBI in children.[14] In a study performed to understand the epidemiology of childhood bacterial diseases, including invasive pneumococcal disease (IPD), screening criteria were used to identify children aged less than 5 years of age who had signs and symptoms of SBI.[14] The study concluded that PCR and antigen testing increased the sensitivity of detection and provided a more precise estimation of the burden of invasive bacterial disease than bacterial culture.[14] This study does not however argue against the use of CRP, PCT and other inflammatory biomarkers.

CXRs and their correlating biomarkers have been shown to have good test sensitivity for pulmonary disease.[9] In our study, all patients with a decreased WCC, 80% of those with a positive PCT, 71.4% with a positive CRP and 70% with an elevated WCC had positive CXR changes. However, these biomarkers cannot be used to distinguish pneumonic changes from hyperinflation.

The results from the analysis of data collected from febrile young children presenting to SBAH reveal that fever or degree of fever does not predict severity of infection, nor source of infection, nor duration of hospitalisation. Degree of fever does not predict biomarkers for bacterial infection (WCC, CRP and PCT). Elevated biomarkers are not related to duration of hospitalisation nor do they predict a positive blood culture. However, it is important to keep in mind that the sensitivity of blood cultures is known to be low, and this 35

SAJCH

is further impacted by the inordinately high (11/51, 21.6%) culture contamination rate at the study facility. The patient’s clinical picture may be more valuable than CRP/WCC when deciding on choice of antibiotics and whether or not an organism is cultured. CRP/WCC cannot be used to predict SBIs in febrile young children and therefore cannot be used to decide on choice of antibiotics.

A low WCC is just as significant a marker for sepsis in children as a high WCC. Literature suggests that the risk of bacteraemia increases from 0.5% if the WCC is <15 × 109/L to more than 18% if WCC is >30 × 109/L.[15] Findings from the literature[15] and our study reveal that more children with a WCC abnormality secondary to sepsis present with a high WCC. Twelve (85.7%) of the 14 patients with WCC abnormalities had an increased WCC. There was a wide range of CRP findings and CRP was specifically unhelpful in predicting infection severity when using cut-off values suggested by the literature.[6] PCT is an expensive biomarker and therefore not usually per­ formed on febrile children presenting to casualty; it is usually reserved for patients hospitalised in the PICU. Fourteen (56%) of the 25 patients in whom a PCT was performed were stable with various illnesses while 11 (44%) were critically ill requiring ICU admission.

The limited number of PCTs done, because of the cost restraint, as well as the wide variability of the positive PCT values (>0.2; range 3 - 728 µg/L) obtained, makes it difficult to interpret data. As 50% of patients with positive CSF results also had a high PCT, it may possibly have a predictive role with regard to meningitis; however, the cohort is too small to confirm this finding.

The sterility of the blood culture techniques used in this study is questionable. Eleven (21.6%) of the 51 blood cultures that were taken cultured a commensal. Therefore, it can be concluded that one-fifth of blood culture specimens are not taken using the correct sterile procedures. This makes it difficult to interpret findings regarding true disease-causing pathogens v. commensal organisms. This problem with specimen collection may reflect, too, the significant number of negative tests.

The results of this study question the validity of urine dipsticks or the method of reading the test. All urine dipsticks are performed in the ward and read by nursing or medical staff; this may call into question the reliability of the staff ’s performance and interpretation of the test. Formal urine sample tests are superior to urine dipsticks. [16] This study revealed that a negative urine dipstick does not rule out a urinary tract infection (UTI) and a positive urine dipstick does not confirm a UTI. Blood tests such as WCC, CRP and PCT may be of more value in assessing response to treatment, rather than predicting the severity of sepsis. Relying on them to withhold or start antibiotic therapy is not prudent. The choice of whether to use antibiotic therapy or when to start treatment should rather be based on clinical judgement, as is invariably the case in the clinical environment in busy hospitals in South Africa, where empiric antibiotic use is most often implemented before laboratory test results have become available.

Routine CXRs are not justified unless a bacterial pneumonia is clinically suspected. This study is a poor example of when to use CXRs judiciously. CXRs were conducted in cases of generalised sepsis, upper respiratory tract infections, viral pneumonias and acute gastroenteritis, without supportive clinical signs of bacterial pneumonia.

If there is any concern for bacterial meningitis, then an LP is still mandatory. CSF Gram stain, bacterial antigen and culture are extremely useful markers to aid in diagnosis, whereas there is no consensus regarding CSF cell count and biochemistry in differentiating between viral and bacterial meningitis. The empiric use of antibiotics in this case should also be based on clinical judgment and not biomarker testing, until the results of the Gram stain, bacterial antigen and culture are available.

MARCH 2016 Vol. 10 No. 1

RESEARCH This study was not without limitations. The small, possibly poorly representative sample size, with no control group and erratic sampling methods (owing to clinical discretion and cost constraints), reflects the large number of febrile children presenting to hospital with undocumented or low-grade fever.

Conclusion

Fever or degree of fever does not predict severity of infection, nor source of infection, nor duration of hospitalisation. Elevated biomarkers (WCC, CRP, and PCT) are not related to duration of hospital stay; they do not predict positive culture results nor identification of significant organisms. Thus, WCC, CRP, and PCT were not shown to be effective in predicting SBIs in febrile children under 5 years of age. This study suggests that clinical suspicion of serious infection and appropriate action are as valuable as extensive testing.

References 1. Thompson MJ, van den Bruel A. Diagnosing serious bacterial infection in febrile children. BMJ 2010;340:2062. [http://dx.doi.org/10.1136/bmj.c2062] 2. Bang A, Chaturvedi P. Yale Observational scale for prediction of bacteremia in febrile children. Indian J Pediatr 2009;76(6):599-604. [http://dx.doi. org/10.1007/s12098-009-0065-6] 3. Powell KR. Fever without a focus. In: Kliegman RM, Behrman RE, Jenson HB, Stanton BF, eds. Nelson Textbook of Paediatrics. 18th ed. New York: Saunders, 2007:1087-1089. [http://dx.doi.org/10.1016/b978-1-4377-06437.00096-6] 4. Squire EN Jr, Reich HM, Merenstein GB, Favara BE, Todd JK. Criteria for the discontinuation of antibiotic therapy during presumptive treatment of suspected neonatal infection. Pediatr Infect Dis 1982;1(2):85-90. [http://dx.doi. org/10.1097/00006454-198203000-00004] 5. Crisp S, Rainbow J. Emergencies in Paediatrics and Neonatology. 2nd ed. London: Oxford University Press, 2013. [http://dx.doi.org/10.1093/ med/9780199605538.001.0001]

36

SAJCH

6. Korppi M, Kröger L. C-reactive protein in viral and bacterial respiratory infection in children. Scand J Infect Dis 1993;25(2):207-213. [http://dx.doi. org/10.3109/00365549309008486] 7. Page AL, de Rekeneire N, Sayadi S, et al. Diagnostic and prognostic value of procalcitonin and C-reactive protein in malnourished children. Pediatrics 2014;133(2):363-370. [http://dx.doi.org/10.1542/peds.2013-2112] 8. Boyels TH, Bamford C, Bateman K, et al. Guidelines for the management of acute meningitis in children and adults in South Africa. S Afr J Epidemiol Infect 2013;28(1):5-15. http://www.sajei.co.za/index.php/SAJEI/article/view/528 (accessed 10 February 2016).

9. Cherian T, Mulholland EK, Carlin JB, et al. Standardized interpretation of paediatric chest radiographs for the diagnosis of pneumonia in epidemiological studies. Bulletin of the World Health Organization 2005;83(5):353-359. http:// www.who.int/bulletin/volumes/83/5/353.pdf (accessed 10 February 2016).

10. Manzano S, Bailey B, Gervaix A, Cousineau J, Delvin E, Girodias JB. Markers for bacterial infection in children with fever without source. Arch Dis Child 2011;96(5):440-446. [http://dx.doi.org/10.1136/adc.2010.203760] 11. Andreola B, Bressan S, Callegaro S, Liverani A, Plebani M, Da Dalt L. Procalcitonin and C-reactive protein as diagnostic markers of severe bacterial infections in febrile infants and children in the emergency department. Pediatr Infect Dis J 2007;26(8):672-677. [http://dx.doi.org/10.1097/inf.0b013e31806215e3] 12. Thayyil S, Shenoy M, Hamaluba M, Gupta A, Frater J, Verber IG. Is procalcitonin useful in early diagnosis of serious bacterial infections in children? Acta Paediatr 2005;94(2):155-158. [http://dx.doi.org/10.1111/j.1651-2227.2005. tb01883.x] 13. Rudinsky SL, Carstairs KL, Reardon JM, Simon LV, Riffenburgh RH, Tanen DA. Serious bacterial infections in febrile infants in the post-pneumococcal conjugate vaccine era. Acad Emerg Med 2009;16(7):585-590. [http://dx.doi. org/10.1111/j.1553-2712.2009.00444.x] 14. Anh DD, Kilgore PE, Slack MP, et al. Surveillance of pneumococcal-associated disease among hospitalized children in Khanh Hoa Province, Vietnam. Clin Infect Dis 2009;48 Suppl 2:S57-64. [http://dx.doi.org/10.1086/596483] 15. Luszczak M. Evaluation and management of infants and young children with fever. Am Fam Phys 2001;64(7):1219-1226. http://www.aafp.org/ afp/2001/1001/p1219.html (accessed 10 February 2016).

16. Welsh A, ed. Urinary Tract Infection in Children: Diagnosis, Treatment and Longterm Management. NICE Clinical Guideline. National Collaborating Centre for Women’s and Children’s Health 2007:52-63. http://www.nice.org.uk/guidance/ cg54/evidence/full-guideline-196566877 (accessed 10 February 2016).

MARCH 2016 Vol. 10 No. 1

RESEARCH

Selected facets of nutrition during the first 1 000 days of life in vulnerable South African communities L M du Plessis,1 PhD; M G Herselman,1 PhD; M H McLachlan,1 PhD; J H Nel,2 PhD 1 2

Division of Human Nutrition, Faculty of Medicine and Health Sciences, Stellenbosch University, Cape Town, South Africa Department of Logistics, Faculty of Economic and Management Sciences, Stellenbosch University, Cape Town, South Africa

Corresponding author: L M du Plessis (lmdup@sun.ac.za)

Background. Optimal nutrition during the first 1 000 days of life can reap lasting benefits throughout life.

Objectives. To assess infant and young child-feeding (IYCF) practices and mother/caregiver-child anthropometry in two vulnerable Breede Valley communities, Western Cape.

Methods. Mothers of children aged 0 - 23 months (N=322) were interviewed to assess IYCF practices. Anthropometric measurements of mothers/caregivers and children were performed according to standard procedures.

Results. Mothers reported early breastfeeding (BF) initiation in 75.2% (242/322) of cases. Of infants <6 months old, 38.5% (45/117) were recorded as exclusively breastfed (EBF). Cross-checking this figure with other research from the area, however, suggests significant over-reporting of EBF. One in five infants <6 months were exclusively bottle fed (19.7%; 23/117) and 48.4% (156/322) aged 0 - 23 months had received bottle feeding in the preceding 24 hours. Eighty-four percent (36/43) of 6 - 8-month-old infants were receiving complementary foods. BF was continued in 32.5% (13/40) of children 12 - 15 months old. In children 6 - 23 months, 44.0% (90/205) received foods from four or more food groups, 71.0% (145/205) received complementary foods the recommended minimum number of times or more, and 44.4% (91/205) received a mini­mum acceptable diet.

The prevalence of stunting and overweight in children was 28.9% and 21.8%, respectively. The prevalence of overweight in mothers/caregivers was 28.9%, and 33.7% were obese, with a mean waist circumference of 88.6 cm.

Conclusion. Indicators showed sub-optimal IYCF practices with child under- and overnutrition coexisting with maternal/caregiver overnutrition. This profile signals a need for urgent and appropriate interventions focusing on the first 1 000 days of life.

S Afr J Child Health 2016;10(1):37-42. DOI:10.7196/SAJCH.2016.v10i1.984

Background

A growing body of evidence is intensifying the focus and interest around the importance of nutrition during the first 1 000 days of life, i.e. from pregnancy up to a child’s second birthday.[1-3] Achieving good nutrition and healthy growth during this well-delineated time period in the lifecycle has been shown to reap lasting benefits throughout life. [1-6] This statement is more profound in view of the findings published in the Lancet Series on Maternal and Child Nutrition, 2013,[7] which stated that maternal and child malnutrition is persistent and encompasses both undernutrition and a growing problem of overweight and obesity in low- and middle-income countries. Concerted efforts should therefore be made to curb malnutrition in women of childbearing age as well as newborns and young children.[7] It is very important to assess infant and young child feeding (IYCF) practices on a continuous basis in order to act on problem areas appropriately and timeously.[8] A World Health Organization (WHO) Working Group developed valid and reliable indicators to assess IYCF practices.[9] These indicators focus on selected food-related aspects of child feeding that could be measured using data from population surveys.[9,10] They were mainly designed for use in large-scale surveys or national programmes, but it was proposed that they could also be used in smaller local and regional programmes.[9] The United Nations Children’s Fund (UNICEF) Programming Guide (2011) recommends that these WHO indicators be used for the situation assessment in the core process of development, planning and implementation of a comprehensive approach to improving IYCF.[8] A review of the IYCF practices around the world, using data derived from 86 developing countries with trend data on key indicators, painted a dismal picture.[8] The global rate for early initiation of breastfeeding remains below 40% and the rate of exclusive breastfeeding (EBF) during the first 6 months of life has only increased slightly (from 33% to 38%) over the past decade. The global data reflect very slow 37

SAJCH

advancement in improving overall EBF, but countries that have shown strong commitment and have devoted serious attention to improving IYCF have shown significant progress.[8] There is a less clear global picture for complementary feeding, since the indicators to measure this were finalised after those for breastfeeding.[9] However, according to the State of the World’s Children (2010) report, in developing countries only 58% of breastfed children aged 6 - 9 months had received any complementary foods in the previous 24 hours.[11] When stunting figures (29% on a global level in 2010) are reviewed to inform this picture, it becomes evident that a large proportion of young children are not receiving a varied diet on a frequent basis.[11] South Africa (SA) is classified as one of the 36 high-burden countries for child malnutrition, with specific reference to a stunting prevalence of higher than 20%.[12] In the absence of trend data on IYCF in SA, other information and indices have to be used to inform the IYCF profile.

The anthropometric status of young children in SA (i.e. >20% stunting, ~10% underweight and >30% overweight and obesity combined) is coupled with the presence of micronutrient malnutrition, including deficiencies in vitamin A, iron and a range of other micronutrients. [13,14] National breastfeeding data for SA are scarce, but there are reports of 88% of mothers initiating breastfeeding after birth.[15] The very low rates of reported EBF of ~8% at 6 months of age are of concern.[15] The majority of infants are formula fed or mixed fed and more than 70% receive other fluids and foods before the age of 6 months.[15,16] Female overweight and obesity (25% and 40%, respectively) in SA, described in the South African National Health and Nutrition Examination Survey (SANHANES-1), are also of concern.[14] Coupled with a vitamin A deficiency (VAD) prevalence of 13%, which indicates a moderate public health problem and anaemia prevalence of 23%, SA women of childbearing age (15 years and older)[14] face a predicament in the context of maternal nutritional health and resultant birth outcomes.[7]

MARCH 2016 Vol. 10 No. 1

RESEARCH The Community-Based Nutrition Security Project (CNSP) of the Division of Human Nutrition, Stellenbosch University, investigated the community food security situation in two vulnerable communities (Avian Park and Zweletemba) in the Breede Valley sub-district, Western Cape Province. This baseline survey provided an opportunity to investigate IYCF practices at household level in this area.

The main objective of this sub-section of the CNSP baseline survey was to assess the feeding practices of infants and young children aged 0 - 23 months with the core WHO validated indicators. The IYCF practices will be discussed within the context of selected anthropometric measurements of mother/caregiver-child pairs.

Methods

Study design and sample selection

A descriptive cross-sectional study was conducted. The sampling frame of the CNSP baseline survey consisted of all households within the selected communities. A simple random selection of households was performed. Households with young children (0 - 36 months) were the basic unit for selection and assessment in the CNSP study. A qualifying household was defined as ‘any household with at least one child (0 36 months) and mother/primary caregiver pair’. Age was verified with a birth certificate or Road-to-Health Booklet (SA tool to assess and monitor the health status of children younger than 5 years). In situations where more than one household lived at an address and both

households had a qualifying mother/primary caregiver-and-child pair, participation was determined through random selection. The same procedure was followed if more than one child of the same mother/ primary caregiver qualified for inclusion.

Households with infants or children between the ages of 0 and 36 months, who resided in Zweletemba or Avian Park for at least 27 weeks of the year, were eligible for inclusion.

Sample size

A power analysis on one-way analysis of variance (ANOVA) with a 5% significance design provided 90% power with a sample size of 170 per group (two groups in two areas). To allow for a response rate of 85%, the total sample size was calculated at 200 per group per area.

The final, slightly oversampled, sample size of the CNSP study was 443 mother/primary caregiver and child pairs. For the focus of this article, complete data were captured for 322 infants/young children in the age range 0 - 36 months (n=117 aged 0 - 6 months; n=205 aged 6 - 23 months), and a total of 312 for the mothers/caregivers.

Methodology

The complete methodology of the CNSP baseline survey is described in the parent protocol (Ethics Committee Ref No: N10/11/368).

The assessment of IYCF practices for children aged 0 - 23 months as well as a description of selected demographic and anthropometric data for these children and their mothers/caregivers is provided here.

Table 1. Definitions and calculation of WHO core indicators for assessing infant and young child feeding practices[9] Definition of indicator

Calculation

Early initiation of breastfeeding: Proportion of children born in the last 24 months who were put to the breast within 1 hour of birth

Children born in the last 24 months who were put to the breast within 1 hour of birth Children born in the last 24 months

Exclusive breastfeeding under 6 months: Proportion of infants 0 - 5 months of age who are fed exclusively with breastmilk

Infants 0 - 5 months of age who received only breastmilk during the previous day Infants 0 - 5 months of age

Continued breastfeeding at 1 year: Proportion of children 12 - 15 months of age who are fed breastmilk Introduction of solid, semi-solid or soft foods (complementary foods): Proportion of infants 6 - 8 months of age who receive solid, semi-solid or soft foods Minimum dietary diversity: Proportion of children 6 23 months of age who receive foods from 4 or more food groups

Children 12 - 15 months of age who received breastmilk during the previous day Children 12 - 15 months of age Infants 6 - 8 months of age who received solid, semi-solid or soft foods during the previous day Infants 6 - 8 months of age Children 6 - 23 months of age who received foods from ≥ 4 food groups during the previous day Children 6 - 23 months of age

Minimum meal frequency: Proportion of breastfed and nonbreastfed children 6 - 23 months of age who receive solid, semi-solid, or soft foods (but also including milk feeds for non-breastfed children) the minimum number of times or more

Breastfed children 6 - 23 months of age who received solid, semi-solid or soft foods the minimum number of times or more during the previous day Breastfed children 6 - 23 months of age and Non-breastfed children 6 - 23 months of age who received solid, semi-solid or soft foods or milk feeds the minimum number of times or more during the previous day Non-breastfed children 6 - 23 months of age

Minimum acceptable diet (summary infant and young child feeding indicator): Proportion of children 6 - 23 months of age who receive a mini­mum acceptable diet (apart from breastmilk)

Breastfed children 6 - 23 months of age who had at least the minimum dietary diversity and the minimum meal frequency during the previous day Breastfed children 6 - 23 months of age and Non-breastfed children 6 - 23 months of age who received at least 2 milk feedings and had at least the minimum dietary diversity not including milk feeds and the minimum meal frequency during the previous day Non-breastfed children 6 - 23 months of age

Consumption of iron-rich or iron-fortified foods: Proportion of children 6 - 23 months of age who receive an iron-rich food or iron-fortified food that is specially designed for infants and young children, or that is fortified in the home

Children 6 - 23 months of age who received an iron-rich food or a food that was specially designed for infants and young children and was fortified with iron, or a food that was fortified in the home with a product that included iron during the previous day Children 6 - 23 months of age

38

SAJCH

MARCH 2016 Vol. 10 No. 1

RESEARCH IYCF questionnaire

A one-page questionnaire consisting of ten questions was formulated based on the wording of the eight WHO-validated IYCF core indicators[9] (Table 1; column 1). It also in­cluded two questions on bottle-feeding from the optional indicators. The questionnaire had to be administered time-efficiently, since the anthropometric measurements and interviews (including sociodemographic questionnaire, household food insecurity access scale, food frequency questionnaire, hunger questionnaire and a dietary diversity questionnaire for mothers and children) of the broader CNSP baseline survey took approximately 2 hours to complete.

All questions were based on recall of the previous day, except for ‘breastfeeding initiation’ which was based on historic recall and ‘Does your baby receive any of the following (iron sources)?’ which was based on usual practice. The age range for the questions was 0 - 23 months, divided into specific monthly intervals, as appropriate to capture information for the different indicators. All questions could be answered either by ‘yes’ or ‘no’.

Training of fieldworkers, pilot study and data collection

Women were recruited from the two communi­ ties included in the study to act as fieldworkers.

These fieldworkers were trained and standard­ ised to administer the IYCF questionnaire, together with other questionnaires within the broader CNSP survey.

Height/length and weight measurements of children and the weight, height and waist

circumference (WC) measurements of mothers/ primary caregivers were taken by two dietitians aided by two trained assistants, according to standard procedures.[17,18] A pilot study was conducted in March 2011 over a period of 1 week in order to test the face and content validity of all the questionnaires. Data collection took place between March and July 2011. During data collection, questionnaires were checked for completeness and accuracy by CNSP research staff.

Data analysis

Data were captured in Microsoft Excel and analysed using SAS 9.3 (2002 - 2010) (SAS Institute Inc., USA).[19] IYCF indicators were calculated as prescribed in the ‘Indicators for assessing infant and young child feeding practices: Part I’ (Table 1; column 2) document.[9] Demographic data and indicators were described using means, standard deviations and percentages.

Children’s ages, weights and heights were used to calculate length/height-forage z-scores (HAZ), weight-for-age z-scores (WAZ), weight-for-length/height z-scores (WHZ) and body mass index (BMI)-for-age z-scores (BAZ) using WHO Anthro (version 3.2.2) software (StatSoft, 2013, USA).[20] Data were interpreted using WHO child growth standards and cut-off values[21] (Table 2).

Adult women’s BMI (weight (in kg)/ height (in m2)) was interpreted using the WHO Consultation on Obesity classification (1999), i.e. underweight (BMI <18.5 kg/ m2),

Table 2. Description and prevalence of nutritional disorders in children 0 - 23 months old Nutritional disorder

Children (N=312), n (%)†

Description*

Stunting

Length/height for age z-score (HAZ) of ≤–2 SD of the WHO child growth standard median

93 (28.81)

Underweight

Weight for age z-score (WAZ) ≤–2 SD of the WHO child growth standard median

15 (4.81)

Wasting

Weight for length/height z-score (WHZ) ≤–2 SD of the WHO child growth standard median

3 (0.96)

Overweight

Weight for length/height z-score (WHZ) ≥+2 SD of the WHO child growth standard median

68 (21.79)

[21]

*Source for description: De Onis et al. †

Number of children (N=312) differs from total number of data for 0 - 23 month old babies (N=322), because anthropometric measurements were not possible in 10 children.

Table 3. Dietary diversity scores of children, reported for five age categories (N=307) Age categories (months) 6 - 35.9

6 - 12.9

13 - 23.9

23.9 - 35.9

6 - 23.9

Children, n

307

87

108

112

195

Dietary diversity scores, mean (SD)

4.16 (1.61)

3.26 (1.35)

4.43 (1.53)

4.60 (1.62)

3.91 (1.56)

39

SAJCH

MARCH 2016 Vol. 10 No. 1

normal weight (BMI = 18.5 - 24.99 kg/m2), overweight (BMI = 25 - 29.99 kg/m2) or obese (BMI ≥30 kg/ m2) as well as waist circumferences with a cut-off point of 88 cm indicating a substantially increased risk of metabolic complications.[22] The cut-off used for maternal short stature is a height measurement of <1.45 m.[23]

Ethics and legal aspects

Ethics approval was granted for the CNSP baseline survey from the Health Research Ethics Committee, Faculty of Medicine and Health Sciences, Stellenbosch University (Ref No. N10/11/368). The consent form was explained and written informed consent was obtained from the mothers of the infants/ young children. This document was available in the three official languages of the Western Cape, i.e. English, Afrikaans and isiXhosa.

Participants were ensured of the anonymous nature of the interviews. Confidentiality was ensured by not recording any personal identification on records and anonymity was ensured by referring only to the group as a whole and not to individuals or individual findings.

Results

Selected demographic and anthropometric data of mothers and children

The average age (standard deviation (SD)) of the mothers/caregivers of children aged 0 23 months was 29.5 (9.67) years and that of the children was 9.85 (7.11) months. Of the mothers/caregivers, 89.1% (278/312) were the children’s biological mothers and 10.9% were caregivers (grandmothers/day-mothers).

The mean (SD) BMI of mothers/caregivers was 28.52 (7.98) kg/m2. The prevalence of underweight was 4.2%; normal weight was 33.3%; overweight was 28.9% and obesity 33.7%. The mean (SD) waist circumference was 88.6 (16.89) cm in mothers/caregivers of children aged 0 - 23 months. Mean (SD) maternal/caregiver height was 1.56 (0.073) m.

The prevalence of stunting, underweight, wasting and overweight (Table 2) in children aged 0 - 23 months was 28.9%, 4.8%, 0.96% and 21.8%, respectively.

Infant and young child feeding (IYCF) practices

Breastfeeding initiation within 1 hour after birth was calculated to occur in 75.2% (n=242/322) of the sample. Thirty-eight and a half percent (38.5%; n=45/117) of the infant population younger than 6 months was recorded as being exclusively breastfed (EBF). Breastfeeding was continued in 32.5% (n=13/40) of babies 12 - 15 months of age.

Nearly 20% of babies 0 - 6 months of age received exclusive bottle feeding (19.7%; 23/117), while 48.4% of babies aged 0 - 23

RESEARCH months received bottle feeding in the preceding 24-hour period (156/322).

Eighty-four percent (36/43) of infants 6 - 8 months of age received solid, semi-solid or soft foods. The proportion of children 6 - 23 months of age who received foods from four or more food groups was calculated at 44.0% (90/205). These food groups included: (i) grains, roots, tubers (e.g. bread, cereal, cooked porridge or potato); (ii) legumes and nuts; (iii) dairy products, excluding breastmilk (milk, yoghurt, cheese); (iv) meat, fish, poultry and liver/ organ meat; (v) eggs; (vi) vitamin A-rich fruit and vegetables (sweet potato, carrots, pumpkin, butternut, spinach, broccoli, apricot, peach, mango); and (vii) other fruits and vegetables.

Just over 70% of breastfed and non-breastfed children aged 6 - 23 months received solid, semi-solid, or soft foods (but also including milk feeds for non-breastfed children) the recommended minimum number of times or more per day (70.7%; 145/205). Fortyfour percent of children 6 - 23 months of age received a mini­mum acceptable diet (apart from breastmilk) (44.4%; 91/205). Eighty-nine percent (182/205) of children 6 - 23 months old received iron in a food or supplement form, with 55.6% (114/205) receiving iron-rich food/s, 42.4% (87/205) receiving a multivitamin containing iron and 53.2% (109/205) receiving baby cereal or foodstuffs manufactured for babies containing iron.

Discussion

In this sub-section of a larger research project, sub-optimal IYCF practices and poor anthropometric profiles were found in mother/ caregiver-child pairs in two vulnerable communities in the Breede Valley, Western Cape Province.

Early initiation of breastfeeding in newborn babies at 75% was lower than the national figure of 88%[15], but corresponds with data derived from a sub-study of the CNSP, where 77% of mothers with babies aged 0 - 6 months reported initiating breastfeeding early. [24] Early initiation of breastfeeding holds many documented benefits[25] and should be a supported practice at various points of contact with pregnant mothers, including in antenatal clinics and maternity wards.[26] When the very low average EBF rate for SA (~8%) is considered,[15] the reported EBF rate in this study, of more than a third of babies aged younger than 6 months, probably better reflects predominant and partial breastfeeding.[24] Similarly, when the data gathered from the IYCF questionnaire were checked for consistency against data derived from a dietary diversity questionnaire (DDQ), administered to mother/caregiver-child pairs within the broad CNSP baseline study, 35% of babies aged younger than 6 months were recorded as being EBF. However, a more in-depth assessment of breastfeeding practices in the younger age group (0 - 6 months) in the same communities suggests significant over-reporting of EBF with both the DDQ and the WHO indicator.[24] Previous work in SA has shown that the term ‘EBF’ was not well understood or practised.[27,28] In this study area it is possible that the term was either not well understood or participants reported on what they think should be done rather than their own practices. Furthermore, it is acknowledged that the EBF indicator lacks sensitivity (i.e. it commonly may classify children as EBF who may have received non-breastmilk liquids or foods prior to the survey) and therefore overestimates the proportion of EBF infants.[9] This overestimation of EBF rates by 1-day recall measures has been observed previously.[29-31] In-depth questioning around this specific aspect of infant feeding should therefore be stressed in field research for calculation of this indicator.

Only about a third of babies were reported as still being breastfed at 12 - 15 months of age, falling far short of the WHO recommendation of ‘continued breastfeeding up to two years of age and beyond’.[32] One in five babies received bottle feeding from birth and were never breastfed, and almost half of babies aged 0 - 23 months received bottle feeding during the previous 24 hours. Bottle feeding does not provide 40

SAJCH

a safe alternative to breastfeeding in SA, mainly because of poor caregiver knowledge and education, as well as a lack of resources that result in poor hygiene and suboptimal IYCF practices.[33] Furthermore, the development of undernutrition as well as overweight and obesity has also been linked to formula feeding, with practices of dilution and overconcentration/overfeeding, respectively.[34,35] Infants between 6 and 8 months of age should be receiving solid, semi-solid or soft foods; however, about a third of babies in this study did not receive complementary foods in the preceding 24 hours.

Almost half of the children 6 - 23 months of age received a diverse diet of minimum acceptability. The mean dietary diversity score (DDS) for children 6 - 35 months, calculated from the broader CNSP dataset was 4.16, indicating a diet of adequate diversity (Table 3).

However, in the age group 6 - 12 months the DDS was 3.26, and in the age group 6 - 23 months the DDS was 3.91. Both these values are below a score of 4, indicating insufficient dietary diversity.[36] This corresponds with the low percentage of children aged 6 - 23 months who received foods from four or more food groups, as calculated with the related WHO indicator.

Intake of iron-rich food and/or supplements in the studied communities seems acceptable. These data correspond with SANHANES-1 data, which has attributed this improvement at national level to the SA food fortification programme enacted in 2003.[14] However, the question in the IYCF questionnaire posed to mothers/caregivers in the Breede Valley communities specifically listed the following options for iron-rich food and/or supplements: (i) meat, fish, poultry and liver/organ meats; (ii) baby cereal or other special baby food enriched with iron; (iii) multivitamin syrup (containing iron) or iron drops, and did not include bread and maize meal, which are the staple foods fortified by law in SA.[40] Children in the studied communities seem to have adequate dietary iron intake from the mentioned sources.

In the Breede Valley sub-study, anthropometric indices indicated that maternal overnutrition and child undernutrition, particularly stunting, as well as child overnutrition coexisted. A high prevalence of overweight and obesity of mothers/primary caregivers, coupled with a high mean waist circumference, indicates an increased risk for non-communicable diseases. When the IYCF practices and child anthropometric profile are considered together with the maternal anthropometric profile, it is clear that nutritional practices in the first 1 000 days of life place the future development, growth and health of children in the Breede Valley in serious jeopardy.

Studies conducted in other provinces of SA also indicated the coexistence of stunting and overweight/obesity in children younger than 5 years of age.[35,41,42] Furthermore, the coexistence of overweight and obesity in mothers/caregivers and undernutrition, particularly stunting, in children has also been reported.[43-45] This presents evidence of a worrying double burden of malnutrition in SA communities undergoing a nutrition transition.[16] The National Department of Health’s Integrated Nutrition Programme (INP) for SA has focused on furthering the maternal and child nutrition agenda over the past two decades.[28] However, a recent, independently prepared report – Evaluation of Nutrition Interventions for Children from Conception to Age 5 – highlighted the limited progress SA has made in improving child nutrition over the past 20 years.[46] The report also states that, although the Departments of Health, Social Development (DSD), Agriculture, Forestry and Fisheries (DAFF) as well as the Department of Rural Development and Land Reform (DRDLR) each have sufficient policies, regulations, and strategies to guide their respective portfolios of nutrition interventions, evidence points to unequal commitment to nutrition across departments with varying levels of leadership, management, planning, budgeting, and staffing.[46] Various recommendations are made in the report to address the current situation, including elevating the status of the INP within the

MARCH 2016 Vol. 10 No. 1

RESEARCH national and provincial government structures with a well-developed nutrition plan that includes nutrition output in a delivery agreement across all sectors. It is also recommended that common indicators should be developed for tracking food and nutrition across all sectors with measurable targets over the short, medium, and long term, as well as a consolidated monitoring and evaluation framework for tracking delivery and the effects of nutrition interventions.[46] These recommendations support the collection of data relevant to the construction of the WHO IYCF indicators which should be included in community-based research projects, larger-scale population studies in SA, including the Demographic and Health Survey and follow-on SANHANES, as well as the country’s District Health Information System (DHIS).[28] It has been proposed in the updated SA IYCF policy (2013) that the following indicators should be calculated from information gathered in surveys and the 14-week data through the DHIS: (i) percentage of mothers initiating breastfeeding within 1 hour of birth; (ii) percentage of babies EBF at 14 weeks; and (iii) percentage of infants 0 - 6 months EBF.[47] The potential to monitor more indicators should be investigated and supported,[28] especially in the light of the recommendations made in the evaluation report.[46]

Study limitations

Although the WHO indicators used to assess IYCF are valid and reliable, the questionnaire used in this study was not validated. This limitation warrants a separate study.

Conclusion

It is of utmost importance to optimise infant nutrition and growth, especially in the first 2 years of life. Although indicators have limitations, they provide a good starting point for decision makers to implement appropriate interventions.[47] IYCF indicators applied in a household survey, among children aged 0 - 23 months, in the Breede Valley indicated sub-optimal IYCF practices. Anthropometric indices indicated that maternal overnutrition and child under- and overnutrition coexist. The combined anthropometric and IYCF practices profile points to poor nutrition during the first 1 000 days of life of infants and young children from these communities.

Valuable experience was gained working with the IYCF indicators at service delivery level. It is recommended that the National and Provincial Nutrition Directorates in collaboration with other government departments strengthen the use of at least some of the core set of WHO IYCF indicators in the DHIS community-based research projects, as well as larger-scale population studies in SA. Acknowledgements. The project manager, the management team and fieldworkers of the CNSP study, as well as the participants, are gratefully acknowledged for their contributions. Prof. S E Drimie is acknowledged for providing insightful comments on drafts of the paper.

Source of funding. The CNSP project received funding from Stellenbosch University (SU)’s Food Security Initiative (FSI), which forms part of the SU HOPE project. The first author also received funding from the Fund for Innovative Research in Rural Health (FIRRH) and the Stellenbosch University Rural Medical Education Partnerships Initiative (SURMEPI) (supported by the President’s Emergency Plan for AIDS relief (PEPFAR) through HRSA under the terms of T84HA21652) both administered by the SU Faculty of Medicine and Health Sciences.

References 1. Bhutta ZA, Ahmed T, Black RE, et al. What works? Interventions for maternal and child undernutrition and survival. Lancet 2008;371:417-440. [http//:dx.doi. org/10.1016/S0140-6736(07)61693-6] 2. Dewey KG, Huffman SL. Maternal, infant and young child nutrition: Combining efforts to maximize impacts on child growth and micronutrient status. Food and Nutrition Bulletin 2009;30(2 Suppl):S187-S189.

41

SAJCH

3. Victora CG, Adair LS, Fall C, et al. and the Maternal and Child Undernutrition Study Group. Maternal and child undernutrition: Consequences for adult health and human capital. Lancet 2008;371:340-357. [http//:dx.doi.org/10.1016/ S0140-6736(07)61692-4] 4. Victora CG, de Onis M, Hallal PC, et al. Worldwide timing of growth faltering: Revisiting implications for interventions. Pediatrics 2010;125:e473-e480. [http://dx.doi.org/10.1542/peds.2009-1519] 5. Martorell R, Horta BL, Adair LS, et al. Weight gain in the first two years of life is an important predictor of schooling outcomes in pooled analyses from five birth cohorts from low- and middle-income countries. J Nutr 2010;140(2):348354. [http://dx.doi.org/10.3945/jn.109.112300] 6. Fall CH, Borja JB, Osmond C, et al. COHORTS group. Infant-feeding patterns and cardiovascular risk factors in young adulthood: Data from five cohorts in low- and middle-income countries. Int J Epidemiol 2011;40(1):47-62. [http:// dx.doi.org/10.1093/ije/dyq155] 7. Black RE, Victora CG, Walker SP, et al. and the Maternal and Child Nutrition Study Group. Maternal and child undernutrition and overweight in low-income and middle-income countries. Lancet 2013 [http://dx.doi.org/10.1016/S01406736(13)60937-X] 8. The United Nations Children’s Fund. Programming guide: Infant and young child feeding. UNICEF [homepage on the Internet]. 2011. http://www.unicef. org/nutrition/files/Final_IYCF_programming_guide_2011.pdf (accessed 28 February 2015).

9. World Health Organization. Indicators for Assessing Infant and Young Child Feeding Practices: Part I. Geneva: World Health Organization, 2008. http://whqlibdoc.who.int/publications/2008/9789241596664_eng.pdf?ua=1 (accessed 28 February 2015).

10. Arimond M, Daelmans B, Dewey K. Indicators for feeding practices in children.

Lancet 2008;371(9612):541-542. [http://dx.doi.org/10.1016/S0140-6736(08)60250-0] 11. The United Nations Children’s Fund. State of the World’s Children 2010 report. New York: UNICEF. 2012. http://www.unicef.org/rightsite/sowc/pdfs/SOWC_Spec%20 Ed_CRC_Main%20Report_EN_090409.pdf (accessed 28 February 2015).

12. Black RE, Allen LH, Bhutta ZA, et al. Maternal and child undernutrition: Global and regional exposures and health consequences. Lancet 2008;371(9608):246260. [http://dx.doi.org/10.1016/S0140-6736(07)61690-0] 13. Labadarios D, Swart R, Maunder EMW, et al. Executive summary of the National Food Consumption Survey Fortification Baseline (NFCS-FB-I) South Africa, 2005. S Afr J Clin Nutr 2008;21(3):245-300.

14. Shisana O, Labadarios D, Rehle T, et al. South African National Health and Nutrition Examination Survey (NHANES-1). Cape Town: Human Sciences Research Council Press, 2013.

15. Mhlanga RE. Maternal, newborn and child health: 30 years on. In: Barron P, Roma-Reardon J, eds. South African Health Review. Durban: Health Systems Trust [homepage on the Internet]. 2008. http://www.hst.org.za/publications/841 (accessed 8 May 2015).

16. Du Plessis LM, Kruger S, Sweet L. Complementary feeding: critical window of opportunity from 6 months onwards. S Afr J Clin Nutr 2013;26(3):S129-S140.

17. Cogill B. Anthropometric Indicators Measurement Guide. Washington, DC: Food and Nutrition Technical Assistance (FANTA) Project, FHI 360, 2003. http://www.fantaproject.org/sites/default/files/resources/anthropometry-2003ENG.pdf (accessed 28 February 2015).

18. World Health Organization. Physical Status: The Use and Interpretation of Anthropometry. Report of a WHO Expert Committee. Technical Report Series No. 854 Geneva: World Health Organization, 1995. http://whqlibdoc.who.int/ trs/WHO_TRS_854.pdf?ua=1 (accessed 28 February 2015) 19. StatSoft, Inc. Electronic Statistics Textbook. Tulsa, OK: StatSoft. 2013. [Online].

Available: http://www.statsoft.com/textbook/ (accessed 28 February 2015).

20. World Health Organization. Anthro for personal computers. Version 3.2.2.

Software for assessing growth and development of the world’s children. Geneva: World Health Organization, 2011. http://www.who.int/childgrowth/software/ en/ (accessed 28 February 2015).

21. De Onis M, Onyango A, Borghi E, Siyam A, Pinol A, eds. WHO Child Growth Standards: Length/height-for-age, weight-for-age, weight-for-length, weightfor-height and body mass index-for-age: Methods and development. 2006. http://www.who.int/childgrowth/standards/Technical_report.pdf (accessed 28 February 2015).

22. World Health Organization. Obesity: Preventing and Managing the Global Epidemic: Report of a WHO Consultation. Technical Report Series No. 894.

Geneva: World Health Organization, 1999. http://whqlibdoc.who.int/trs/ WHO_TRS_894.pdf (accessed 28 February 2015).

23. Kruger S. Maternal anthropometry and pregnancy outcomes: A proposal for the monitoring of pregnancy weight gain in outpatient clinics in South Africa.

Curationis 2005;28(4):40-49. [http://dx.doi.org/10.4102/curationis.v28i4.1012] 24. Goosen C, McLachlan MH, Schubl C. Infant feeding practices during the first 6 months of life in a low-income area of the Western Cape Province. S Afr J Child Health 2014;8(2):50-54 [http://dx.doi.org/10.7196/sajch.675] 25. World Health Organization Collaborative Study Team on the Role of Breastfeeding on the Prevention of Infant Mortality. Effect of breastfeeding on infant and child mortality due to infectious diseases in less developed countries: A pooled analysis. Lancet 2000;355(9202):451-455. [http://dx.doi.org/10.1016/ s0140-6736(99)06260-1] 26. Perez-Escamilla R. Evidence Based Breast-Feeding Promotion: The BabyFriendly Hospital Initiative. J. Nutr 2007;137:484-487.

27. Doherty T, Chopra M, Nkonki L, et al. Effect of the HIV epidemic on infant feeding in South Africa: ‘When they see me coming with the tins they laugh at

MARCH 2016 Vol. 10 No. 1

RESEARCH me’. Bull World Health Organ 2006;84(2):90-96. http://www.who.int/bulletin/ volumes/84/2/90.pdf (accessed 28 February 2015).

28. Du Plessis LM, Pereira C. Commitment and capacity for the support for breastfeeding in South Africa. S Afr J Clin Nutr 2013;26(3):S120-S128.

29. Zohoori N, Popkin BM, Fernandez ME. Breastfeeding patterns in the Philippines: A prospective analysis. J Biosoc Sci 1993;25:127-138.

30. Piwoz EG, Creed De Kanashiro H, Lopez De Romana G, et al. Potential for misclassification of infants’ usual feeding practices using 24-hour dietary assessment methods. J Nutr 1995;125:57-65.

31. Arimond M, Ruel MT. Progress in Developing an Infant and Child Feeding Index: An Example Using the Ethiopia Demographic and Health Survey 2000.

Washington, DC: International Food Policy Research Institute, 2002.

32. World Health Organization. Global strategy for infant and young child feeding. Geneva: World Health Organization, 2003. http://whqlibdoc.who.int/ publications/2003/9241562218.pdf. (accessed 28 February 2015).

33. Du Plessis LM. Editorial: Infant and young child feeding in South Africa: Stop the crying beloved country. S Afr J Clin Nutr 2013;26(1):4-5.

34. Arenz A, Ruckerl R, Koletzkos B, von Kries R. Breast-feeding and childhood obesity: a systematic review. International Journal of Obesity 2004;28:1247– 1256. [http://dx.doi.org/10.1038/sj.ijo.0802758] 35. Mamabolo RL, Alberts M, Mbenyane GX, et al. Feeding practices and growth of infants from birth to 12 months in the central region of the Limpopo Province of South Africa. Nutrition 2004;20(3):327-333. [http://dx.doi.org/10.1016/j.nut.2003.11.011] 36. Food and Agriculture Organization. Guidelines for Measuring Household and Individual Dietary Diversity. Nutrition and Consumer Protection Division, Food and Agriculture Organization of the United Nations. Rome, Italy. 2010. http://www.fao.org/3/a-i1983e.pdf (accessed: 8 May 2015).

37. Goosen C, McLachlan MH, Schubl C. Factors impeding exclusive breastfeeding in a low-income area of the Western Cape province of South Africa. Africa Journal of Nursing and Midwifery 2014;16(1):13-31.

38. Matthysen M, Lombard MJ, Daniels LC. Factors that influence attitude, beliefs and barriers of caregivers regarding complementary feeding practices of infants aged 6-12 months in the Breede Valley District of the Western Cape. Stellenbosch University Library and Information Services. SUNscholar Research Repository. Dissertation.

2014. http://scholar.sun.ac.za/handle/10019.1/86713 (accessed 28 February 2015).

39. Bhutta ZA, Das JK, Rizvi A, et al. Evidence-based interventions for improvement of maternal and child nutrition: What can be done and at what cost? Lancet 2013;382(9890):452-477. [http://dx.doi.org/10.1016/S0140-6736(13)60996-4]

42

SAJCH

40. Department of Health. Regulations relating to the fortification of certain foodstuffs. 2003. http://www.doh.gov.za/docs/regulations/2003/Regulations_ Fortification.pdf (accessed 28 February 2015).

41. Mamabolo RL, Steyn NP, Alberts M. Can the high prevalence of micronutrient deficiencies, stunting and overweight in children at ages 1 and 3 years in the Central Region of Limpopo province be explained by diet? S Afr J Clin Nutr 2006;19(3):102-113.

42. Kimani-Murage EW, Kahn K, Pettifor JM, et al. The prevalence of stunting, overweight and obesity, and metabolic disease risk in rural South African children. BMC Public Health 2010;10:158. [http://dx.doi.org/10.1186/14712458-10-158] 43. Steyn NP, Labadarios D, Maunder EMW, et al. Secondary anthropometric data analysis of the national food consumption survey in South Africa: The double burden. Nutrition 2005;21:4-13. [http://dx.doi.org/10.1016/j. nut.2004.09.003] 44. Smuts CM, Faber M, Schoeman SE, et al. Socio-demographic profiles and anthropometric status of 0- to 71-month-old children and their caregivers in rural districts of the Eastern Cape and KwaZulu-Natal provinces of South Africa. S Afr J Clin Nutr 2008;21:117-124.

45. Ardington C, Case A. Health: Analysis of the NIDS Wave 1 dataset. NIDS Discussion paper 2. 2009. http://www.nids.uct.ac.za/documents/discussionpapers/93-nids-discussion-paper-no02 (accessed 28 February 2015).

46. Department of Health, Department of Social Development, Department of Performance Monitoring and Evaluation. Evaluation of Nutrition Interventions for Children from Conception to Age 5. Pretoria. 2014. http://www.google. co.za/url?sa=t&rct=j&q=&esrc=s&source=web&cd=1&ved=0CBwQFjAA& url=http%3A%2F%2Fwww.thepresidency-dpme.gov.za%2Fnews%2FDocu ments%2FAnnual%2520Report%2520on%2520Evaluations%2520compress ed%25202013-14%252014%252010%252022.pdf&ei=Bu3qVJviIoLtaszbgAg& usg=AFQjCNHLfV7Y_hc3wRwDs-jwWuucbVwX7A (accessed 28 February 2015).

47. Department of Health. Infant and Young Child Feeding Policy. Department of Health. 2013. http://www.health-e.org.za/2013/09/20/infant-young-child-feedpolicy-2013/#lightbox/0/ (accessed 28 February 2015).

48. Bryce J, Coitinho D, Darnton-Hill I, et al. for the Maternal and Child Undernutrition Study Group. Maternal and child undernutrition: effective action at national level. Lancet 2008;371(9608):510-526. [http://dx.doi. org/10.1016/S0140-6736(07)61694-8]

MARCH 2016 Vol. 10 No. 1

RESEARCH

A 3-year survey of acute poisoning exposures in infants reported in telephone calls made to the Tygerberg Poison Information Centre, South Africa C J Marks,1 BSc Pharm, Hons Pharmacology, MSc MedSci; D J van Hoving,2 MB ChB, DipPEC, MMed (Em Med), MSc MedSci (Clin Epi) Division of Clinical Pharmacology, Department of Medicine, Faculty of Medicine and Health Sciences, Stellenbosch University, Cape Town, South Africa 2 Division of Emergency Medicine, Stellenbosch University, Cape Town, South Africa 1

Corresponding author: C Marks (carinem@sun.ac.za) Background. Infants undergo rapid development changes and are particularly vulnerable to toxic chemicals. Identifying and evaluating the toxic risks that exist in this age group could be very valuable when making recommendations on how to prevent specific types of poisoning.

Objectives. This study analysed the toxic substances responsible for acute poisoning exposures in infants (<1 year of age) as well as the severity of the exposures.

Methods. A retrospective analysis of the Tygerberg Poison Information Centre (TPIC) database was conducted over a 3-year period (1 January 2011 to 31 December 2013). Descriptive statistics are provided for the entire study population as well as for the neonatal subgroup (<30 days old).

Results. The TPIC handled 17 434 consultations during the 3-year study period. Infants were involved in 1 101 cases (6.3%), of which 46 cases (4.2%) were neonates. Most enquiries about infants were associated with non-drug chemicals (n=824, 74.8%). Pharmaceuticals were involved in 185 cases (16.8%) followed by biological exposures (e.g. snake and spider bites, scorpion stings, plant and mushroom poisonings) (n=109, 9.9%). Most infants (n=987, 89.6%) presented with no or only minor symptoms. In neonates, 17 (37.0%) presented with moderate to severe toxicity. Six of these (35.3%) were poisoned by complementary and alternative medicines.

Conclusion. Most poisoning exposures in infants are not serious and can be safely managed at home after contacting a poison centre.

Identification and documentation of poisoning in this special population is of great importance.

S Afr J Child Health 2016;10(1):43-46. DOI:10.7196/SAJCH.2016.v10i1.1045

The diagnosis and management of poisoning in infants are challeng­ ing. Although fatalities among these young children are rare, poisoning still represents a frequent cause for hospital admission. [1] One would expect that most poisonings in infants would be ad­ minis­tered by a carer, but children are inquisitive by nature, and as infants become mobile (8 - 12 months), they start to explore their surroundings, inevitably putting things into their mouths. [2] This causes accidental self-poisoning; however, the true extent of acute poisoning exposures in infants (<1 year of age) is not well documented. Although poisoning trends change with increasing age, the agents responsible for these poisonings are usually found in and around the home.[2,3] Young children remain particularly vulnerable to poisoning, since exposure to a poison at a critical stage in their development may have severe long-term consequences for health.

Extrapolating from a population-based study in a developed country,[4] we can estimate that about 500 000 people might be poisoned each year in South Africa (SA). In 2004 the World Health Organization (WHO) estimated that unintentional poisoning caused 346 000 deaths worldwide, of which 91% occurred in low- and middle-income countries.[5] In 2012, unintentional injury, which included poisoning, was the fifth leading cause of death in infants in the USA.[6] Unintentional injuries in young children are largely preventable. Infants cannot remove themselves from danger and cannot read warning signs and labels. This important information should lead to the development of preventive strategies.

Parents often seek medical attention when a child has been exposed to a poison, more so if the poisoning involves an infant or neonate. This special population is more susceptible to poisoning because of their pharmacokinetic differences, their small blood 43

SAJCH

volume and small size.[7] Identifying and evaluating the toxic risks that exist in this age group could be very valuable when making recommendations on how to prevent specific types of poisoning. We aimed to describe which toxic substances are responsible for acute poisoning exposures and the severity of the poisoning as reported to the Tygerberg Poison Information Centre (TPIC).

Methods

Study design

A retrospective analysis of the TPIC database was conducted over a 3-year period (1 January 2011 to 31 December 2013). This study was approved by the Health Research Ethics Committee of the University of Stellenbosch (Ref: S13/04/088).

Study setting and data collection

The TPIC is located in the Western Cape Province of SA and forms an integral part of the Tygerberg Academic Hospital. It is located in the Division of Clinical Pharmacology, Faculty of Medicine and Health Sciences, on the Tygerberg campus of Stellenbosch University. The TPIC provides a free toxicology service to healthcare workers and members of the public throughout SA. The service covers toxicity assessments, as well as poisoning treatment recommendations, and is available 24 hours a day. Poisoning consultations are handled by telephone on a dedicated emergency number, which is advertised nationally. Poisoning enquiries are answered by pharmacists, medical scientists or medical doctors. Calls received are manually recorded on a standard TPIC consultation form and include: date and time; geographical area; medical background of the caller; patient’s age and gender; time, intent and route of exposure; substances involved;

MARCH 2016 Vol. 10 No. 1

RESEARCH advice given; and the poisoning severity score. Severity grading is done according to the European Association of Poisons Centres and Clinical Toxicologists poisoning severity score (severity 0 - 1 = no or minor symptoms; severity 2 - 3 = moderate to severe toxicity) and is allocated to the patient at the time of the call, when the observed clinical symptoms and signs are taken into account.[8] Data from consultation forms are entered on an electronic database by a permanent TPIC staff member; this is not cross-checked because of financial and time restraints.

Study population

All data in the TPIC database relating to infants (<1 year old) were included for the period 1 January 2011 to 31 December 2013. In the event of incomplete consultation forms, the cases were not excluded; instead, only the specific missing variable was indicated as unknown.

Data collection and analysis

Variables collected from the TPIC database included: time of call; geographical area; caller’s medical background; patient’s age and gender; time, intent and route of exposure; substances involved; and the poisoning severity score. Collected data were committed to a standard Microsoft Excel (USA) spreadsheet. Descriptive statistics are provided for the entire study population as well for the neonatal subgroup (<30 days old).

Almost all exposures were accidental (n=1 092, 99.2%) with only 8 cases (0.7%) reported as deliberate. The gender was recorded in 1 081 cases (males n=591, 54.7%).

Most enquiries about infants were associated with non-drug chemicals (n=824, 74.8%). One hundred and sixty-eight (15.4%) infants were exposed to pharmaceuticals, but often combinations of drugs were involved (e.g. antihistamine plus an analgesic).

Pharmaceuticals were involved in 185 cases (16.8%) followed by biological exposures (e.g. snake and spider bites, scorpion stings, plant and mushroom poisonings) (n=109, 9.9%).

Irritants and corrosives were responsible for most non-drug chemical exposures (n=255, 30.9%) (Fig. 4). Toilet cleaner discs were liable for 45 exposures (17.6%). One hundred and eightyfour exposures (22.3%) were related to pesticides, of which almost a third related to rodenticides (n=56, 30.4%) and a quarter to pyrethroids (n=45, 24.5%). Of the rodenticides, 40 (71.4%) involved anticoagulant rodenticides and 16 (28.6%) to aldicarb, a cholinesterase inhibitor that is often illegally sold as a rat poison.[10] Alcohols were responsible for 51 non-drug exposures (6.2%); 31 involved surgical spirits (60.8%). One hundred and two non-drug chemicals (12.4%) were categorised as mildly toxic substances; 52 of these cases (55.3%) involved silica gel. Of 109 biological exposures, plants and mushrooms were responsible for 80 cases (73.4%).

Results

Poisoning exposures in infants (n=1 101)

The TPIC handled 17 434 consultations during the 3-year study period. Infants were involved in 1 101 cases (6.3%), of which 46 cases (4.2%) were neonates (Fig. 1). The mean age of the study population was 9.8 months.

Most poisoning enquiries were received from the Western Cape (n=384, 34.9%) (Fig. 2) and the majority were made by healthcare professionals (n=659, 59.9%) (Fig. 3).

Number of infants, n

1 000

Healthcare professionals 59.9% (n=659)

Public 39.1% (n=430)

Unknown 1% (n=12)

Public medical facilities 32.4% (n=357)

n=857

800

Private hospitals and clinics 15.7% (n=173)

600 400 200 0

General practitioners 9.2% (n=357)

n=129

n=46

n=69

0 - 30 days

>1 - 6 months

>6 - 9 months

>9 - 12 months

Pharmacies 2.5% (n=28)

Fig. 1. Age distribution of infants exposed to poisons as reported to the TPIC (N=1 101).

Unknown

Fig. 3. Flow chart depicting the origin of calls received by the TPIC.

0.4%

Free State

0.6% 0.5% 2.2% 1.6%

North West

2%

Other countries Northern Cape

Mpumalanga Limpopo

Other Unknown

5.2%

3.5% 1.5%

Western Cape Eastern Cape KwaZulu-Natal Gauteng

3.3%

Essential oils

6.8%

Complementary and alternative medicine Alcohols

7.8% 10.4%

34.9%

11.4% 19.7%

24%

64

Irritants and corrosives

Fig. 2. Geographical distribution of calls received by the TPIC (light blue) v. SA’s population per province (dark blue).[9]

44

SAJCH

51 62

Pesticides

27.4%

30

Cosmetics Biological agents

24.3%

25

Volatile hydrocarbons Minimally or non-toxic substances

12.5%

42 9

102 113 184 255

Fig. 4. Number of enquiries about chemical and biological non-drug exposures (n=933) in infants.

MARCH 2016 Vol. 10 No. 1

RESEARCH 91% Infants (1 - 12 months) Neonates (0 - 30 days)

Table 1. Pharmaceuticals responsible for poisoning exposures in infants (N=185) n (%)

Analgesics

49 (26.5)

Antihistamines

18 (9.7)

Antimicrobials

17 (9.2)

Neuroleptics (antipsychotic drugs)

11 (5.9)

Antidepressants

9 (4.9)

Hypnotics

8 (4.3)

Antihypertensives

7 (3.8)

Fig. 5. Comparison of poisoning severity scores between neonates and older infants.

Anti-epileptics

7 (3.8)

Vitamins and minerals

6 (3.2)

Table 1 shows the pharmaceuticals responsible for poisoning in infants, with analgesics being the most common group of drugs (n=49, 26.5%). Paracetamol was involved in 23 cases (47.0%).

Most infants (n=987, 89.6%) presented with no or only minor symptoms. No initial deaths were recorded; however, fatality cannot be excluded since poisoned patients were not followed up. In neonates, 17 (37.0%) presented with moderate to severe toxicity. Six of these (35.3%) were poisoned by complementary and alternative medicines (CAM) (Table 2); four received SA-sold ‘Dutch remedies’ that contained herbal products and two received SA traditional medicines that contained unknown herbal preparations and plant extracts. Fig. 5 depicts the differences in the poisoning severity score between neonates and older infants.

Miscellaneous

53 (28.7)

Percentage (%)

Drug categories

63%

37%

9% Severity 0 & 1

Severity 2 & 3

Poisoning severity score

Discussion

Between 1 January 2011 and 31 December 2013 the TPIC received 17 434 poisoning-related calls. Of these, 1 101 (6.3%) involved infants. Our findings are similar to USA data where exposures involving infants ranged between 5.3% and 10.5%.[11,12] These USA data are derived from the American Association of Poison Control Centres which received and managed poisoning enquiries similar to those handled by the TPIC. SA is divided into nine provinces, but most calls came from only three (n=954; 86.6%) (Fig. 2). The skewed presentation of calls from the Western Cape, KwaZulu-Natal and Gauteng could be due to these provinces having larger populations, but it does not explain why we received only a few enquiries from the Eastern Cape. One reason may be that people can only call a poison information centre if they know that one exists![13] The Western Cape has two poison centres.

Both KwaZulu-Natal and Gauteng once had poison centres which have now closed. In addition, most of the calls were received from healthcare professionals (60%). Educational standards in SA are low and may make it likely that parents would be unaware of a poisons information centre, so would take children directly to a healthcare facility. The opposite happens in the USA, where poison information services are well known and integrated into emergency services.[11] Childhood poisoning exposures are almost always accidental.[14] Deliberate poisoning in children is unlikely to be reported to poison centres. A more accurate estimate of the prevalence of intentional poisoning in infants would be obtained from hospital admissions, as well as those from primary healthcare facilities. Eight children were poisoned intentionally during the study period. Two cases included intrauterine exposure to a chemical. It is not known when exactly in the gestation period these prenatal exposures took place, but both babies were born soon after the mother was exposed to the toxin. In six cases, poisoning was deliberate. The detection of intentional poisoning is worrying as this is a rare phenomenon.[15] Socioeconomic factors might play a role and require further investigation.

45

SAJCH

Table 2. Infants exposed to complementary and alternative medicine (N=30) Severity 0 - 1,* n (%) Severity 2 - 3,* n (%) Neonates, <1 month of age (n=10)

4 (40)

6 (60)

Infants, 1 - 12 months of age (n=20)

14 (70)

6 (30) [8]

*Severity 0 - 1 = no or minor symptoms; severity 2 - 3 = moderate to severe toxicity.

Our data revealed that non-drug exposures were responsible for almost 85% of reported cases. The three most commonly implicated non-drug toxins were irritant and corrosive substances, followed by pesticides and biological exposures. The high prevalence of toilet cleaner discs exposure (n=45, 17.6% of exposures to irritants and corrosives) was an interesting finding. Although most toilet detergents, household soaps and bleaches have irritant properties, it is impossible to predict whether these agents will cause injury to the oesophagus and stomach.

There is also a frustrating lack of information in respect of potentially toxic ingredients contained in commercially available household preparations. Labels on such products seldom provide adequate information on ingredients, and often do not contain warnings about their potential toxicity. Owing to the significant public ignorance regarding the toxicity of household non-drug chemicals, these preparations should not be displayed close to foodstuffs or within reach of children (as is often the case in households and supermarkets).[16] Pesticide exposures are a major public health problem in SA,[10] predominantly in settings of low education and poor regulatory frameworks. A previous study investigated accidental pesticide poisoning in children.[10] Poor and crowded areas are a breeding ground for most pests, and people seek cheap effective ways to deal with the problem. They often buy rodenticides intended for agricultural use instead of home use from street vendors (agricultural pesticides have a higher toxicity). In our study 16 infants were exposed to aldicarb, an illegally sold rodenticide.

Compared with exposures to biological toxic substances produced by animals (n=29), enquiries regarding poisoning exposures to plants and mushrooms were relatively high in our study (n=80).

The inherent toxicity of most plants is low, so the ingestion of small to moderate quantities is unlikely to produce toxic effects. Serious poisonings caused by plants are exceptional as a young child eats only a few leaves or seeds at a time.

In recent years there has been an increase in the popularity of CAM. This type of medicine includes medical products (e.g. herbal medicine,

MARCH 2016 Vol. 10 No. 1

RESEARCH homeopathy, traditional medicine) and medical practices (e.g. body manipulation, acupuncture) that are not part of standard medical care. Unlike conventional medicine, some CAM has not been through clinical testing and is often not safe.[17] It is regularly given to newborns to treat a diversity of symptoms, including colic. Owing to their small size and differences in pharmacokinetics, neonates are at an increased risk for toxicity.[7] In our study six neonates were seriously poisoned by CAM (Table 2). Two involved SA traditional medicine (a combination of animal, mineral and plant products to induce physiological or psychological healing effects) and four involved herbal products (a combination of plant seeds, berries, roots, leaves, bark, or flowers for medicinal purposes). Three of the four herbal products contained valerian which has been associated with toxicity.[18] In this study serious CAM-related poisonings occurred more in neonates (60%) compared with older infants (30%) (Table 2). The low weight and physiological differences in this vulnerable population could be a contributing factor.[7]

Study strengths

This study is valuable because data on poisoning are hard to obtain anywhere. They are particularly sparse in developing countries, such as SA. On the African continent there are very few poison information centres from which data on telephone calls can be obtained.

Study limitations

Incomplete data might have influenced the results in either direction.

However, the possible effect was limited by only indicating the specific missing variable as unknown and not excluding the entire case.

Our study only includes data from one poison centre in SA, and cannot be extrapolated to estimate the prevalence of poisoning in SA. Data on admissions to hospitals for acute poisoning are very different from data recorded from telephone enquiries by poison information centres. Telephone enquiries reflect a need for information for professionals and for the public. Future epidemiological studies of poisoning should use data from telephone enquiries and from admissions to hospital.

No cases reported to the TPIC were officially followed up because of time and financial constraints. Data regarding measurable clinical outcomes (e.g. morbidity and mortality) are therefore lacking and the clinical significance of the data remains unclear.

Conclusion

Most poisoning exposures in infants are not serious and can be safely managed at home after contacting a poison centre. Poisoning is more challenging in neonates because of physiological, behavioural and biological differences. Identification and documentation of poisoning in this special population is of great importance. The use of CAM in infants can be harmful and more information is needed to ensure its safe use.

46

SAJCH

Acknowledgements. We would like to express our sincere thanks and appreciation to the Tygerberg Poison Information Centre staff.

References 1. Azkunaga B, Mintegi S, Bizkarra I, Ferna´ndez J. Toxicology surveillance system of the Spanish Society of Paediatric Emergencies: First-year analysis. Eur J Emerg Med 2011;18(5):285-287. [http://dx.doi.org/10.1097/ MEJ.0b013e3283462504] 2. Chatsantiprapa K, Chokkanapitak JPN. Host and environment factors for exposure to poisons: A case-control study for preschool children. Thailand Inj Prev 2001;7:214-217.

3. Agran PF, Anderson C, Winn D, Trent R, Walton-Haynes L, Thayer S. Rates of pediatric injuries by 3-month intervals for children 0 to 3 years of age.

Pediatrics. 2003;111(6):e683–e692. http://pediatrics.aappublications.org/ content/111/6/e683.long (accessed 15 January 2015).

4. Kristinsson J, Palsson R, Gudjonsdottir GA, Blondal M, Gudmundsson S, Snook CP. Acute poisonings in Iceland: A prospective nationwide study. Clin Toxicol (Phila) 2006;46:126-132.

5. World Health Organization. Global Health Observatory Data Repository. Geneva: World Health Organization, 2009. http://apps.who.int/gho/data/?theme=main (accessed 16 August 2015).

6. Centers for Disease Control and Prevention, National Center for Injury Prevention and Control. Injury Prevention and Control: Data and Statistics.

Ten Leading Causes of Death and Injury in United States, 2015. http://www. cdc.gov/injury/wisqars/leadingcauses.html (accessed 23 February 2015).

7. Ginsberg G, Hattis D, Sonawane B. Incorporating pharmacokinetic differences between children and adults in assessing children’s risks to environmental toxicants. Toxicol Appl Pharmacol 2004;198(2):164-183. http://www.ncbi.nlm. nih.gov/pubmed/15236952 (accessed 18 January 2015).

8. Persson HE, Sjöberg GK, Haines JA, Pronczuk de Garbino J. Poisoning Severity Score grading of acute poisoning. J Toxicol Clin Toxicol 1998;36(3):205-213.

9. South Africa: Mid-year population estimates 2014. http://beta2.statssa.gov.za/ publications/P0302/P03022014.pdf (accessed 23 February 2015).

10. Balme KH, Roberts JC, Glasstone M, et al. Pesticide poisonings at a tertiary children’s hospital in South Africa: An increasing problem. Clin Toxicol (Phila) 2010;48:928-934. [http//:dx.doi.org/10.3109/15563650.2010.534482] 11. American Association of Poison Control Centers. National Poison Data System. Annual Report 2013. http://www.aapcc.org/data-system (accessed 22 January 2015).

12. Franklin RL, Rodgers GB. Unintentional child poisonings treated in United States hospital emergency departments: National estimates of incident cases, population-based poisoning rates, and product involvement. Pediatrics 2008;122:1244-1251. [http//:dx.doi.org/10.1542/peds.2007-355] 13. Litovitz T, Benson BE, Youniss J, Metz E. Determinants of U.S. poison center utilization. Clin Toxicol (Phila) 2010;48(5):449-457. [http//:dx.doi. org/10.3109/15563651003757947] 14. Even KM, Armsby CC, Bateman ST. Poisonings requiring admission to the pediatric intensive care unit: A 5-year review. Clin Toxicol (Phila) 2014;52(5):519-524. [http//:dx.doi.org/10.3109/15563650.2014.909601] 15. Dine MS, McGovern ME. Intentional poisoning of children – an overlooked category of child abuse: Report of seven cases and review of the literature.

Pediatrics 1982;70(1):32-35.

16. Savage EP, Tessari JD, Laurier P. Pesticides sold in grocery stores. Health Services Reports 1972;87(8):734-736.

17. Niggemann B, Grüber C. Side-effects of complementary and alternative medicine. Allergy 2003;58(8):707-716. [http//:dx.doi.org/10.1034/j.13989995.2003.00219.x] 18. Tomassoni AJ, Simone K. Herbal medicines for children: An illusion of safety?

Curr Opin Pediatr 2001;13(2):162-169. [http//:dx.doi.org/10.1097/00008480200104000-00014]

MARCH 2016 Vol. 10 No. 1

RESEARCH

Exploring sibling attitudes towards participation when the younger sibling has a severe speech and language disability M Hansen,1 M(AAC); M Harty,1,2 PhD; J Bornman,1 PhD 1 2

Centre for Augmentative and Alternative Communication, University of Pretoria, South Africa Division of Communication Sciences and Disorders, University of Cape Town, South Africa

Corresponding author: J Bornman (juan.bornman@up.ac.za) Background. Typically developing children who have a younger sibling with a disability often feel inadequately supported and excluded from family interactions. However, early intervention programmes often use family activities and routines as intervention settings. Siblings’ negative attitudes towards participation in such activities may therefore hamper intercession efforts.

Objective. To determine the attitudes of typically developing children toward their younger siblings with severe speech and language disabilities in four everyday life situations identified by the World Health Organization’s International Classification of Functioning, Disability and Health: Children and Youth Version, namely communication, domestic life, interpersonal interaction and relationships, and major life areas.

Methods. An adapted structured interview format was used to determine the attitude of the 27 participants, 6- to 10-year-old typically developing South Africans.

Results. Typically developing peers were most positive towards participation in play activities with their sibling with a disability. They were also positive towards participation in household tasks. They were less positive towards communication participation and least positive about participation in interpersonal relationships. A significant difference between certain components of attitudes was reported for three of the four domains.

Conclusion. The attitudes of typically developing children are generally positive towards participating with their younger siblings with severe speech and language disabilities. These results can be used to select activities for activity-based interventions and to guide interventions aimed at supporting the siblings of children with disabilities.

S Afr J Child Health 2016;10(1):47-51. DOI:10.7196/SAJCH.2016.v106i1.1046

Sibling relationships are one of the closest, strongest and most enduring relationships within the core family system, as they allow for sharing ideas and expressing feelings through experiences of loyalty, support and rivalry.[1,2] Interactions among siblings provide (and remain) opportunities to learn many crucial social, emotional, cognitive and communication skills that are necessary for participation in everyday life. Furthermore, the attitudes of either of the siblings towards sibling interactions will influence their participation in everyday life situations.

In families where one sibling has a disability, the typically developing sibling(s) can act as enablers if they have a positive attitude towards their sibling with a disability and towards their participation in everyday family activities.[3] Given the potential impact of a typically developing sibling’s attitude towards participation on the functional outcomes of a child with a disability, it is important that these attitudes are investigated.

Literature suggests that typically developing siblings often feel inadequately supported and excluded from participating with the sibling with the disability, resulting in atypical family interaction and participation.[2,4] Furthermore, they may feel excluded from their sibling’s intervention[5] and in these instances, sibling relationships may be viewed as a barrier to family-focused, activity-based intervention.

The World Health Organization[6] provides a conceptual frame­ work in the form of the International Classification of Functioning, Disability and Health: Children and Youth (ICF-CY) Version in terms of which children’s participation can be explored. The ICF-CY operationalises participation in nine domains and views participation as a non-linear, dynamic process that results from interaction between the child and its environment. Participation can 47

SAJCH

be assessed as frequency, i.e. ‘being there’, or as intensity, i.e. ‘being engaged while being there’,[7] in other words, how the child interacts with the social and physical environment. There is an increasing body of evidence that critically evaluates the use of the ICF-CY framework in clinical settings for certain populations, e.g. children with speech and language disabilities.[8-10] Current research is using the ICF-CY framework to identify some of the everyday activities and life situations that children frequently participate in, with a view to informing family-focused, activity-based intervention.[11] Based on an examination of the everyday activities and life situations captured in the ICF-CY, empirical research indicates that professionals rate many of these situations as being important for children.[12] Data from a study involving parents of preschool children who were receiving speech and language intervention, indicate that parents are more likely than professionals to describe intervention outcomes in terms of changes in their child’s participation,[10] which underscores the importance of using ‘participation’ as an outcomes measure in family-focused intervention. More importantly, however, parents emphasise that this participation is described as changes in engagement in everyday life situations such as play, socialisation with others, and behaviour at home and school.[10] Therefore, current data show that the participation domain in the ICF-CY is helpful when describing changes in participation in the everyday life situations of young children with speech and language disabilities.

Evidence suggests that everyday life situations in families with young children can be identified from the following ICF-CY domains: communication, self-care, domestic life, interpersonal relationships and interactions, major life areas (play and education) and, to a lesser extent, community life.[12] In spite of seminal works that acknowledge the role of siblings in the development of children with

MARCH 2016 Vol. 10 No. 1

RESEARCH disabilities[2,4,13] and that highlight the fact that siblings’ perceptions may differ from those of adults,[1] a literature search revealed a paucity of empirical data that describe the attitudes of siblings towards participating in everyday life situations with children with speech and language disabilities.

In this study we consequently used an adapted structured interview schedule format to determine the attitudes of typically developing children towards their younger siblings with severe speech and language disabilities in four everyday life situations, according to the ICF-CY, namely communication (chapter d3), domestic life (chapter d6), interpersonal interactions and relationships (chapter d7) and major life areas (chapter d8).

Methods

Research design

An exploratory descriptive design based on a structured interview schedule with the Talking Mats (Talking Mats Ltd, Scotland) visual framework was used (Fig. 1).

Participants

Participant selection criteria All participants had to be English speaking, typically developing (not having had to repeat a year at school) children between the ages of 6 and 10 years. They had to have a younger sibling who attended a preschool for children with severe speech and language disability.

Although South Africa (SA) has an inclusive education policy, the majority of children with disabilities who attend school still go to separate, ‘special’ schools for children with disabilities. The special needs preschool that was selected provides an integrated assessment, intervention and preschool environment for ambulatory children with severe speech and language delay and average intelligence quotient (IQ). All participants had to reside permanently with their sibling.

The parents of 35 potential participants who met the selection criteria were requested to consent to their child’s participation, of whom eight parents refused consent, which resulted in 27 parents consenting on behalf of their children. Assent was then sought from the children themselves, and all assented.

Description of the participants and their siblings The mean age of the 27 participants (9 male, 18 female) was 8 years (range 6 - 10 years), while the mean age of their siblings (21 male, 6 female) was 5 years (range 3 - 8 years).

Age and gender differences were calculated for each of the sibling dyads, of whom nine were same gender (6 male-male, 3 femalefemale). The remaining 18 were mixed gender (15 dyads with a female participant and male sibling, and 3 with a male participant and

Response category

Symbol options

Fig. 1. Example of a completed Talking Mat.

female sibling). Despite their severe speech and language disability, all the siblings were ambulatory and displayed age-appropriate cognitive abilities, as stipulated by the admission criteria of the preschool they attended.

Materials

Two core theoretical constructs – attitude and participation – were used in the development of the interview items. Attitudes included three components, namely beliefs (the cognitive component) that often attract strong feelings (the affective component) and may then lead to specific actions (the behavioural component).[14] Participation was discussed in terms of the participation domains of the ICFCY, namely major life areas, interpersonal relationships and domestic life.[10,12] Data from these studies informed the selection of three typical everyday life situations with which siblings were bound to interact and that represented three different ICF-CY domains.

Communication was added as an additional domain, as it was likely that the participants would have participation restrictions in this domain due to the nature of their sibling’s disability. A preliminary set of 40 items that tapped into the affective, behavioural and cognitive aspects of attitude was developed (Table 1).[14] An expert panel comprising three speechlanguage therapists, five remedial teachers and two occupational therapists with relevant experience subsequently evaluated the items and discarded nine of them. Picture communication symbols (PCSs) that visually depicted the specific item were developed for the remaining 31 items. All symbols were printed on cardboard, laminated and cut into 5 cm × 5 cm squares. Velcro was attached to the back of each symbol for easy placement on a regular short-pile mat. A structured interview schedule was developed to ensure consistent presentation of the stimuli to each participant, thereby heightening procedural integrity.

48

SAJCH

MARCH 2016 Vol. 10 No. 1

Procedures

Data were collected by using a structured interview schedule based on the Talking Mats visual framework[15] with a three-point response category: ‘Definitely yes’ (showing the PCS symbol of a thumbs up and a smiling face); ‘Not sure’ (showing the PCS symbol of a man shrugging his shoulders) and ‘Definitely no’ (showing the PCS symbol of thumbs down and a sad face). Talking Mats was developed to assist individuals to express their views or indicate their feelings. [15] Each item on the structured interview schedule was read aloud and the matched PCS symbol was shown to each participant individually.

When asked how they felt about each item, they indicated their choice by placing the matching PCS symbol under the relevant response category. After completion of all the items, the resulting Talking Mats image was captured with a digital camera. Fig. 1 shows an example of a completed Talking Mat. A series of three trial items was developed to ensure that participants understood the instructions and were able to complete the required task.

Prior to the main study, a stringent pilot study was conducted with 30 Grade 1 learners with typically developing sibling(s) to ensure that the language level and suggested procedures were appropriate for the target population.

The result was that minor modifications were made to the interview schedule. Next, a second pilot study was conducted with four children who matched the inclusion criteria for the study, which resulted in two items being reworded.[14] No changes were necessary to the 31 items and the scripted interview remained unchanged.

Ethical considerations

The research adhered to ethical principles as stipulated by the Declaration of Helsinki,[16] with ethics approval from the relevant university. Written informed consent was obtained from the principal, the preschool’s board of trustees and the parents, followed by assent from the children themselves.

RESEARCH Table 1. Everyday sibling life situations used in the development of the structured interview schedule Examples of items from the interview schedule

Everyday life situations

Definition

Use

Communication (ICF-CY Chapter d3: Communication)

Communication participation is interactive and reciprocal. Participants exchange ideas, needs, desires and information. It forms the basis of human interaction.

Items related to communication participation (receptive and expressive language) in order to determine the typically developing child’s attitude towards communication participation with his/her sibling with disability.

Affective item: ‘I enjoy talking with my brother/sister’ (#A3)

Household tasks include basic grooming activities, as well as chores in and around the house in which children are expected to participate.

Items related to participation in household tasks to determine the typically developing child’s attitude towards participation with his/her sibling with disability in household tasks.

Affective item: ‘I feel proud to be my brother/sister’s helper’ (#A8)

Household tasks (ICF-CY Chapter d6: Domestic life)

Behavioural item: ‘I ignore my brother/sister when I don’t understand him/her’ (#B2) Cognitive item: ‘My brother/sister understands everything I say’ (#C4)

Behavioural item: ‘I help my brother/sister do jobs around the house’ (#B6) Cognitive item: ‘My brother/sister pretends to struggle with easy tasks’ (#C8)

Time and responsibility situations (ICF-CY Chapter d7: Interpersonal interactions and relationships)

Interpersonal interactions focus on establishing and maintaining relationships.

Children with disabilities often have limited peer interaction opportunities, thus their interpersonal sibling interactions take on increased meaning.

Items pertaining to joint participation time, time alone and time spent with parents to determine the typically developing child’s attitude towards time and responsibility issues that relate to participation with his/her sibling with disability.

Affective item: ‘I like spending time by myself sometimes’ (#A10)

Play (ICF-CY Chapter d8: Major life areas)

Play participation is regarded as the vehicle for learning. It provides an environment free from pressure where children learn about language, social rules and interaction roles.

Items pertaining to play and recreational activities to determine the typically developing child’s attitude towards play participation with his/her sibling with disability.

Affective item: ‘I like playing with my brother/sister’ (#A1)

Data analysis

The means procedure was employed to determine the mean score for participation in the four everyday life situations. Higher values indicated a more positive attitude. The Friedman two-way analysis of variance (ANOVA) ranked values were used to evaluate differences among the four situations: play (mean (SD) 2.72 (0.24)); participating in household tasks (mean 2.54 (0.22)); communication (mean 2.41 (0.21)); and interpersonal relationships (mean 1.96 (0.29)). SPSS (IBM, USA) compensated for the number of groups when the pairwise comparisons for the Friedman procedure were conducted and allowed a 5% level of significance for each pair. This procedure was repeated to evaluate differences in means among the affective, behavioural and cognitive components of attitudes for each of the everyday life situations.

Results

Procedural integrity

All sessions were video-recorded. Procedural integrity was determined for 20% of randomly selected completed interviews. A rating agree­ ment of 97% was obtained, which indicated a high level of integrity.

Comparisons across the four everyday life situations

The 27 participants’ attitudes towards participation in play were positive (mean 2.72 (0.24)). They also showed fairly positive attitudes towards participation in household tasks (mean 2.54 (0.22)) and in communication (mean 2.41 (0.21)). The participants 49

SAJCH

Behavioural item: ‘I want to spend more time alone with Mom and Dad’ (#B5) Cognitive item: ‘Mom and Dad let me spend time by myself ’ (#C11)

Behavioural item: ‘I always include my brother/sister in games I play’ (#B1) Cognitive item: ‘My brother/sister enjoys playing with me’ (#C12)

were least positive about participation in interpersonal relationships (time and responsibility situations such as sharing parental attention), with a mean of 1.96 (0.29). The results of the Friedman procedure, which evaluated differences in means among the four everyday life situations, were significant (χ2(3) (N=27) = 52.38, p≤0.0001) (Table 2).

Follow-up pairwise comparisons were conducted. The null hypothesis was rejected if the z-value was larger than the critical z-value, where 1 – PHI(ZC)= ALPHA/(K(K – 1)). The critical z-value for an alpha of 0.05 was 2.64 with four groups.

The mean for play was significantly higher than the mean for communication (z=3.27) and for interpersonal relationships (time and responsibility situations) (z=6.96), although no statistically significant difference was observed between the mean values for play and participating in household tasks (z=1.79). The mean for interpersonal relationships was significantly lower than for communication (z=3.69), participation in household tasks (z=5.17) and play (z=6.96).

Comparisons across the components of the sibling attitude scale

The Friedman two-way ANOVA was conducted to explore how the three attitude components (affective, behavioural and cognitive) differed across the everyday life situations targeted in this study. The results of the Friedman test were significant for three of the four everyday life situations (Table 3).

MARCH 2016 Vol. 10 No. 1

RESEARCH Table 2. Comparison of the attitude means obtained for the four everyday life situations (N=27)

Communication (ICF-CY Chapter d3: Communication)

Time and responsibility situations (ICFCY Chapter d7: Interpersonal interactions and relationships)

Household tasks (ICF-CY Chapter d6: Domestic life)

Play (ICF-CY Chapter d8: Major life areas)

Mean

SD

Mean

SD

Mean

SD

Mean

SD

p-value

2.41*b

0.21

2.54ab

0.22

1.96c

0.29

2.72a

0.24

<0.0001†

* Pairwise comparisons are significant at the p≤ 0.05 level in the instances where the means of life situations being compared do not contain the same postscript (e.g. communication and play). † Significant at the p≤0.05 level.

Table 3. Comparison between attitude components in the four everyday life situations (N=27) Affective

Behavioural

Cognitive

Everyday life situations

Mean

SD

Mean

SD

Mean

SD

p-value

Communication (ICF-CY Chapter d3: Communication)

2.44*ab

0.36

2.67a

0.62

2.35b

0.31

0.0316†

Household tasks (ICF-CY Chapter d6: Domestic life)

2.89a

0.29

2.79a

0.32

1.81b

0.57

0.0001†

Time and responsibility situations (ICF-CY Chapter d7: Interpersonal interactions and relationships)

1.99a

0.33

1.74a

0.94

2.04a

0.55

0.2091

Play (ICF-CY Chapter d8: Major life areas)

2.84ab

0.25

2.41a

0.59

2.85b

0.33

0.0176†

* Pairwise comparisons are significant at the p≤ 0.05 level in the instances where the means of life situations being compared do not contain the same postscript (e.g. communication and play). † Significant at the p≤0.05 level.

Follow-up pairwise comparisons were con­ ducted for communication, play and house­ hold tasks. The null hypothesis was rejected if the z-stat was larger than the critical z-value, where 1 – PHI(ZC)= ALPHA/(K(K – 1)).

The critical z-value for an alpha of 0.05 was 2.39 with three groups. With regard to communication, the behavioural component was significantly higher than the cognitive component (z=2.59). However, the affective com­ ponent did not differ significantly from the behavioural component and neither did the affective and cognitive components differ. The analysis of the items related to participation in household tasks indicated a significant difference between the behavioural and cognitive component (z=4.35) and also between the affective and cognitive components (z=4.83), but not between the affective and behavioural components. Both the affective and behavioural components were higher than

the cognitive component. In play, the results indicated that the cognitive component was significantly higher than the behavioural component (z=2.59), but that there was no difference between the affective and behaviour components and neither between the affective and the cognitive components.

Discussion

The rating of the affective component was high for all the everyday life situations and no statistically significant difference was observed. Earlier research reported that siblings of children with disabilities displayed significantly more nurturing behaviours than siblings of children without disabilities,[17] and that they were kinder and more positive toward their siblings with disability than toward their typically developing siblings.[4] Similarly, Caro and Derevensky[18] reported frequent displays of spontaneous affectionate behaviour during 50

SAJCH

MARCH 2016 Vol. 10 No. 1

sibling interactions. Results from the present study concur with research evidence, which suggests that the presence of a disability within a sibling relationship increases the likelihood of a more positive sibling affect.

Rating the cognitive and behavioural components across the everyday life situations indicated a statistically significant difference between the behavioural and the cognitive components across three of the everyday life situations (play, communication and household tasks). The mean value for the cognitive component was higher than the behavioural component in both play and interpersonal relationships (time and responsibility situations such as sharing parental attention). The mean for the behavioural component was higher than the cognitive component for communication participation and participation in household tasks.

In exploring these results, it becomes necessary to incorporate the constructs of engagement and flow (as a measure of cognitive involvement). Participation in play as an everyday life situation requires high levels of engagement in activities that are intrinsically motivating and within the child’s initial competence, experience and interests. [19] Furthermore, play requires a high level of voluntary control, once again suggesting high levels of cognitive involvement in a functional context. The same holds true for interpersonal relationships (time and responsibility situations), which focus on establishing and maintaining relationships in the context of an individual’s social roles. When determining the level of cognitive involvement in the everyday life situations of participating in communication interactions and household tasks, it becomes clear that both of these situations require lower levels of cognitive involvement, because the activities that establish these everyday life situations require more ‘doing’ (behavioural component) than ‘thinking’ (cognitive component). It was evident from the present study that a measure (such as an attitude measure) that expresses the degree of cognitive involvement in everyday life situations in an ICF-CY domain could be used to make recommendations in respect of family-focused, activity-based interventions.

Sibling relationships are one of the key relationships in which children acquire and develop many social and cognitive skills.

If clinicians are able to identify specific everyday life situations in which cognitive involvement of both siblings is high, e.g. during play interactions, these situations could be used as contexts to coach older siblings to model specific communication strategies, e.g. key word signing or using augmentative and alternative communication systems and strategies. This would support

RESEARCH acquisition and generalisation of skills facilitated by the clinician, as well as contributing to long-term, positive family outcomes, as the responsibility of supporting intervention will not rest solely on the parents, but on the family structure as a whole.

Furthermore, by identifying the attitudes of siblings in specific everyday life situations, barriers to and opportunities for positive outcomes can be more specifically identified and defined. This will enable the intervention team to make more appropriate and timely referrals and to structure intervention goals and supports accordingly.

Study limitations

A methodological limitation was the relatively small sample size and the fact that it was demographically limited (e.g. in terms of socio-economic status). Moreover, participants were selected by means of a convenience sample (i.e. selecting siblings of children in a preschool for children with severe speech and language disability).

Therefore, further studies are needed in more demographically representative samples in SA to come to more general conclusions.

The comprehensive process of developing the adapted interview schedule, which included an in-depth pilot study,[14] allowed the researchers to refine the language level and length of the schedule to suit the needs of the target population in this study. It is, however, important to mention that certain allowances had to be made to accommodate the age of the participants, which had an impact on the statistical analysis. For instance, a three-point scale had to be used rather than a four-point scale and a limited number of response items had to be used under each sub-section, all of which subsequently contributed to the unequal distribution of response items from the different participation domains.

Conclusions

The impact of certain sibling dyad characteristics (e.g. age, gender, severity of disability) on the attitude of typically developing siblings in everyday life situations has yet to be determined. However, data from the current study indicate that the attitudes of typically developing siblings towards participation with their younger siblings with severe speech and language disabilities are generally positive. The data also reveal that a measure (such as an attitude measure) which indicates the degree of cognitive involvement in everyday life situations in an ICF-CY domain may be used to describe participation. The identification of additional everyday life situations for young children with disabilities, in conjunction with their siblings’ views of participation in these situations, can and should inform appropriate activity-based interventions for children with disabilities, as well as for sibling support programmes.[6,12]

51

SAJCH

References 1. Latta A, Rampton T, Rosemann J, et al. Snapshots reflecting the lives of siblings of children with autism spectrum disorders. Child Care Health Dev 2013;40(4):515–524. [http://dx.doi.org/10.1111/cch.12100].

2. McHugh M. Special Siblings: Growing up with Someone with a Disability.

Baltimore: Paul H Brookes, 2003.

3. Vig S. Young children’s object play: A window on development. J Devel Phys Disabil 2007;19(3):201-215. [http://dx.doi.org/10.1007/s10882-0079048-6] 4. Gamble WC, McHale SM. Coping with stress in sibling relationships: A comparison of children with disabled and nondisabled siblings. J Appl Devel Psychol 1989;10:353-373.

5. Opperman S, Alant E. The coping responses of the adolescent siblings of children with severe disabilities. Disabil Rehabil 2003;25(9):441-454. [http:// dx.doi.org/10.1080/0963828031000069735] 6. World Health Organization (WHO). International Classification of Functioning, Disability and Health: Children and Youth Version. Geneva; WHO, 2007.

7. Granlund M, Arvidsson P, Niia A, et al. Differentiating activity and participation of children and youth with disability in Sweden: A third qualifier for the ICFCY? Am J Phys Med Rehabil 2012;91(13):84-96. [http://dx.doi.org/10.1097/ PHM.0b013e31823d5376] 8. McCormack J, McLeod S, McAllister L, Harrison LJ. A systematic review of the association between childhood speech impairment and participation across the lifespan. Int J Speech-Lang Path 2009;11(2):155-170. [http://dx.doi. org/10.1080/17549500802676859] 9. McLeod S, Threats TT. The ICF-CY and children with communication disabilities. Int J Speech-Lang Path 2008;10(1-2):92-109. [http://dx.doi. org/0.1080/17549500701834690] 10. Thomas-Stonell N, Washington K, Oddson B, Robertson B, Rosenbaum P. Measuring communicative participation using the FOCUS©: Focus on the Outcomes of Communication Under Six. Child Care Health Dev 2013;39(4):474-480. [http://dx.doi.org/10.1111/cch.12049] 11. Raghavendra P. Participation of children with disabilities: Measuring subjective and objective outcomes. Child Care Health Dev 2013;39(4):461-465. [http:// dx.doi.org/10.1111/cch.12084] 12. Adolfsson M. Applying the ICF-CY to identify children’s everyday life situations: A step towards participation-focused code sets. Int J Soc Welf 2013;22(2):195206. [http://dx.doi.org/10.1111/j.1468-2397.2012.00876.x] 13. Powell TH, Gallagher PA. Brothers & Sisters: A Special Part of Exceptional Families. 2nd ed. Baltimore: Paul H Brookes, 1993.

14. Hansen M. The attitudes of typically developing children towards participation with their siblings with severe speech and language disabilities. Unpublished Master’s thesis in Augmentative and Alternative Communication. Pretoria: University of Pretoria, 2011.

15. Bornman J, Murphy J. Using the ICF in goal setting: Clinical application using Talking Mats. Disabil Rehabil Assist Tech 2006;1(3):145-154. [http://dx.doi.org /10.1080/17483100612331392745] 16. World Medical Association. Declaration of Helsinki – Ethical principles for medical research involving human participants. Seoul; World Medical Association, 2008. [http://dx.doi.org/10.4103/0022-3859.52846] 17. Abramovitch R, Corter C, Pepler DJ, Stanhope L. Sibling and peer interaction: A final follow-up and comparison. Child Dev 1986;57(1):217-229.

18. Caro P, Derevensky JL. An exploratory study using the Sibling Interaction Scale: Observing interactions between siblings with and without disabilities. Educ Treat Children 1997;20(4):383-403.

19. Almqvist L, Uys CJE, Sandberg A. The concepts of participation, engagement and flow: A matter of creating optimal play experiences. S Afr J Occup Ther 2007;37(3):8-13.

MARCH 2016 Vol. 10 No. 1

RESEARCH

Childhood and adolescent fatalities at the Pretoria Medico-Legal Laboratory: 2005 - 2009 G van den Ordel, BSc, BSc (Hons) (Medical Criminalistics); L du Toit-Prinsloo, MB ChB, DipForMed (SA) Path, FCForPath (SA), MMed (Path) (Forens); G Saayman, MB ChB, MMed (MedForens), FCForPath (SA) Department of Forensic Medicine, Faculty of Health Sciences, University of Pretoria, South Africa Corresponding author: L du Toit-Prinsloo (lorraine.dutoit@up.ac.za)

Background. Children and young adults are particularly vulnerable to intentional and accidental fatal injuries. The majority of deaths in Africa in children <5 years of age are due to infectious diseases. Road traffic fatalities constitute a large proportion of deaths in children and young adults worldwide.

Objective. To evaluate the demographic details, external cause/circumstance of death and manner of death in children and young adults admitted to the Pretoria Medico-Legal Laboratory (PMLL).

Methods. A retrospective case audit was conducted on all persons aged 1 - 20 years admitted to the PMLL from January 2005 through December 2009.

Results. A total of 965 cases were identified. Childhood deaths constituted 7 - 9% of all cases admitted. The majority were aged between 16 and 20 years (42%), followed by children aged between 1 and 5 years (33%). The most common cause of death was injury due to road traffic accidents (this was present in all age groups).

Conclusion. Children and young adults contribute a relatively large number of admissions to the medico-legal mortuary. The majority of these deaths are accidental. Many of the accidents could have been prevented with stricter legislation with regard to transportation of children and safety surrounding swimming pools.

S Afr J Child Health 2016;10(1):52-56. DOI:10.7196/SAJCH.2016.v10i1.1047

Children and young adults (adolescents) are particularly vulnerable to inflicted and accidental fatal injuries.[1,2] The United Nations Children’s Fund (UNICEF) convention on the right of the child defines a child as ‘a person below the age of 18, unless the laws of a particular country set the legal age for adulthood younger’.[3] According to Statistics South Africa (SA) (2008),[4] the leading underlying natural cause of death in children <14 years of age was intestinal infectious diseases (21.8% of deaths), followed by influenza and pneumonia – collectively accounting for a third of deaths in this age group. HIV/AIDS is a leading cause of mortality, with a prevalence rate of 10.6% in SA in 2009, and an estimated 5.21 million people living with the disease.[5,6] Individuals aged between 15 and 24 years are recorded as having the highest incidence of the disease.

In SA, the age group 15 - 49 years is most at risk for non-natural deaths, with 13.9% of deaths in this group being non-natural and the most common circumstance/external cause of death interpersonal violence (assault).[4] The high incidence of child fatalities due to road traffic accidents has been well documented worldwide. In San Diego, Fraga et al.[1] reported (in their study conducted at the medico-legal mortuary) that road traffic fatalites were the leading cause of accidental death (40.2% of cases). In Nebraska, Okoye and Okoye[7] reviewed medico-legal childhood deaths over a 7-year period and reported motor vehicle accidents as the cause of death in 23.6% of cases. In Brazil, Modelli et al.,[8] reporting on deaths in children <12 years, found the leading external cause of death to be road traffic fatalities (22% of cases). In Nigeria, Nwafor et al.[9] reviewed deaths in children ≤14 years; in a 20-year period, road traffic accidents accounted for 74.6% of accidental deaths and 15.4% of total deaths. In SA, Bass et al.[10] reviewed children <14 years who presented to Red Cross War Memorial Children’s Hospital in Cape Town. They indicated that 430 children were involved as pedestrians in accidents over a 12-month period, with 106 fatalities.

Deaths due to asphyxia in children was the second most common cause of death in the study by Fraga et al.,[1] accounting for 22.7% of cases.

52

SAJCH

Okoye and Okoye[7] indicated asphyxia mainly in children <5 years, with a total of 22 (15.7%) cases, and 7 (5%) cases of drowning. Modelli et al.[8] indicated asphyxia in 17% of cases.

Homicides in children <18 years accounted for 24.2% of deaths in San Diego,[1] 5% in Nebraska,[7] 1.5% in Nigeria,[9] and 13% in Istanbul.[11] Suicide in children accounted for 9.4% of deaths in San Diego,[1] 13.6% in Nebraska,[7] and 7.4% in the Istanbul study.[11] Cases due to natural causes on completion of the medico-legal investigation of death and that were admitted to the medico-legal mortuaries mentioned in the abovementioned studies accounted for 30% of deaths in the Okoye and Okoye[7] study, 50% in the Modelli et al.[8] study, 77.5% in the Nwafor et al.[9] study, and 12% in the Beyaztas et al.[11] study.

Numerous countries, including Austalia, the UK and the USA, have child fatality review programmes (also referred to as child death review programmes). These consist of a multidisciplinary approach to reviewing child deaths, ultimately providing prevention strategies to reduce childhood fatalities.[12] In 2009, SA had a population of 49.32 million, Gauteng being the most densely populated province with a population of 10.53 million.[5] According to Statistics SA,[4] the total number of registered deaths (including late submissions) were 598 131 in 2005, 612 778 in 2006, 603 094 in 2007 and 592 073 in 2008 (late submissions of deaths have not been completely analysed). In Gauteng, the total number of registered deaths for 2008 was 115 909.[4] Pretoria had an estimated population of 1.6 million in 2012.[13] The Inquests Act 58 of 1959 governs the medico-legal investiga­ tions of death in SA. This legislation stipulates that the presiding officer would ultimately determine the cause of death and whether or not someone else could be held accountable for the death (therefore determining the manner of death). In SA, forensic pathologists do not determine the manner of death. In 2001, according to the National Non-Natural Mortality Surveillance System (NMSS), pathologists could give their opinion with regard to the manner of death at the time of the postmortem examination – for statistical purposes.[14]

MARCH 2016 Vol. 10 No. 1

RESEARCH The objective of this study was to identify the demographic details, external cause/ circumstance and manner of death in child cases admitted to the Pretoria Medico-Legal Laboratory (PMLL) over a 5-year period. All children aged 1 - 20 years were included in the study, while infants (children <1 year of age) were excluded. The reasons for exclusion of such infants were twofold: (i) the report by Statistics SA indicated that the majority of deaths in infants are due to natural disease processes and previous studies conducted at the PMLL have indica­ ted that even in those admitted to medico-legal mortuaries, most deaths are due to natural causes, with respiratory tract infections being the most common;[4,15] and (ii) in international studies, such as the study by Lathrop[16] in New Mexico, 41.4% of all children ≤19 years admitted were infants, and they contributed 77% to all natural causes of death. The rationale for inclusion of children ≤20 years was that in the final year of school in SA, although the majority of children are aged 18, a proportion have repeated at least 1 year and in many instances 2 years at school.[17] The World Health Organization (WHO) defines an adolescent as a person ≤19 years.

[18] Adolescence is regarded as the transitional period between childhood and adulthood. In Canada, the Adolescent Health Committee indicated that adolescence ends when ‘an

adult identity and behavior are accepted’ and that the period roughly corresponds to the ages stated by the WHO, but that flexibility in the age span should be present.[19] Our secondary objective was to identify certain trends and prevent­able causes of death.

Methods

A retrospective descriptive case audit was conducted on all children aged 1 - 20 years admitted to the PMLL from January 2005 through December 2009. The admission register at the mortuary, case files and NMSS data collection sheets were reviewed. The total caseload admitted each year was also recorded.

The data collected included demographic details (age, gender and race), scene of accident, place of death, external cause/circumstance of death and manner of death.

Owing to the retrospective nature of the study, the completeness of the documentation at the PMLL was a limitation. Full ethical approval was obtained from the Faculty of Health Sciences, University of Pretoria Research and Ethics Committee before commencement of the study.

Results

A total study population of 965 cases was identified and included. The total caseload at the mortuary was 11 761 cases (over the 5-year period). Childhood deaths constitu­ ted 7 - 9% of all cases admitted annually.

Demographic details

The mean age of the entire population was 11 years. In 7 (1%) cases no specific age was recorded. Most deaths occurred in the

Table 1. Gender in relation to age groups Age group (years)

Male, n (%)

Female, n (%)

Total, n (%)

1-5

191 (59)

132 (41)

323 (33)

6 - 10

55 (50)

56 (50)

111 (12)

11 - 15

75 (65)

41 (35)

116 (12)

16 - 20

289 (71)

119 (29)

408 (42)

Unknown

3 (43)

4 (47)

7 (1)

Total

613 (64)

352 (36)

965 (100)

Table 2. External cause/circumstance of death in relation to age group External cause/ circumstance of death

1 - 5 years, n (%)

6 - 10 years, n (%)

11 - 15 years, n (%)

16 - 20 years, n (%)

Age unknown, n (%)

Total, n (%)

Road traffic fatalities

91 (27)

50 (15)

39 (11)

160 (47)

1

341 (35)

Railway fatalities

0

0

2 (20)

7 (70)

1 (10)

10 (1)

Burns

48 (64)

8 (11)

6 (8)

12 (16)

1 (1)

75 (8)

Electrocution

8 (73)

0

1 (9)

1 (9)

1 (9)

11 (1)

Hanging

0

0

6 (16)

31 (84)

0

37 (4)

Smothering/ strangulation

0

1 (9)

0

10 (91)

0

11 (1)

Drowning

44 (66)

10 (15)

9 (13)

4 (6)

0

67 (7)

Drug overdose/ poisoning

11 (30)

4 (11)

6 (16)

16 (43)

0

37 (4)

Assault

2 (5)

0

4 (11)

31 (85)

0

37 (4)

Jump/fall/push from height

9 (31)

4 (14)

3 (10)

13 (45)

0

29 (3)

Gunshot wounds

5 (8)

2 (3)

10 (16)

44 (72)

1 (1)

62 (6)

Stab wounds

0

0

2 (7)

25 (89)

1 (4)

28 (3)

Natural causes

34 (50)

10 (14)

7 (10)

18 (26)

0

69 (7)

Unknown

39 (45)

13 (15)

9 (10)

25 (29)

0

86 (9)

Other

32 (49)

9 (14)

12 (18)

11 (17)

1 (2)

65 (8)

Total

323 (33)

111 (12)

116 (12)

408 (42)

7 (1)

965 (100%)

53

SAJCH

MARCH 2016 Vol. 10 No. 1

RESEARCH 16 - 20-year group (n=408; 42%), followed by 1 - 5-year-olds (n=323; 33%). The 6 - 10-year group had 111 deaths and the 11 - 15-year group 116 deaths (12% each).

The majority were male (n=613; 64%), with 352 (36%) females. Table 1 indicates the gender in relation to the age groups.

The racial distribution showed that 693 (72%) children were black, 221 (23%) white, 31 (3%) coloured and 20 (2%) Asian.

This is in keeping with the population of the mortuary.

male:female ratio of 2.3:1 and 2.0:1, respectively.

In other studies,[8,9] the ratio was less, with proportionately more female admissions, i.e.

1.4:1 and 1.2:1, respectively.

The majority of the deaths in our study were due to road traffic accidents (35%). This was also the finding in other studies (Istanbul – 38.8%[11] and Nebraska – 50.7%[7]). The most common external cause of accidents in children in all the abovementioned studies[1,7-9,11] was road traffic fatalities, which was also found in the study conducted at the PMLL. In our study, there were nearly equal numbers of passengers (n=126) and children as pedestrians (n=121) involved in accidents. In San Diego,[1] there were more passengers than pedestrians involved.

Worldwide, emphasis has been placed on reducing childhood mortality. The use of preventive measures, such as child restraint devices, has been well documented to reduce

Suicidal deaths The most common method of suicide was hanging (n=37; 52%), followed by poisoning/ overdose on prescription drugs (n=13; 18%) (Table 4).

Discussion

In our study children aged 1 - 20 years contri­ buted 7 - 9% to the annual caseload in the mortuary. In Nebraska,[7] children aged 0 18 years consituted 10.9% of the caseload to the medico-legal mortuary, and in Nigeria,[9] children ≤14 years contributed 12.3% to the total caseload. It should be noted that both of the latter studies included infants.

The gender differentiation in our study showed a male:female ratio of 1.7:1. This was the same as in the study by Fraga et al.[1] Some studies[7,11] indicated a relatively larger proportion of male descendants, with a

External cause/circumstance of death

The most common external cause/circum­­ stance of death was road traffic accidents (among all age groups) (n=341; 35%). Table 2 indicates the external cause/circumstance of death in relation to the age groups. Deaths cate­ gorised in the group ‘other’ included mostly anaes­t hetic/procedure-associated deaths (n=36), mainly in the 1 - 5-year group. The remaining 21 cases included aviation accidents, sport-related inju­ ries, explosion deaths, lightning death (n=1) and other accidents such as electric gates falling on children.

The types of road traffic fatalities are summa­ rised in Fig. 1. There were an equal number of deaths from pedestrians involved in accidents and from children as passengers. The highest number of pedestrian fatalities was in the 1 5-year group.

60 50

6 - 10 years 11 - 15 years

39

40 n

1 - 5 years

54

50

16 - 20 years

34 28

30 20

26 18 15

12 1

0

Pedestrian

Passenger

0

The manner of death (Fig. 2) was mostly acci­ dental deaths in all age groups (n=525; 54%).

Homicidal deaths accounted for 136 (14%) cases and suicidal deaths for 72 (7%). The majority of homicides (76%) and suicides (82%) were in the 16 - 20-year group. Natural disease processes were present in 69 (7%) (50% of these occurred in the 1 - 5-year group), and in 162 (17%) cases the manner was undetermined.

Age unknown

15

10

Manner of death

29

0 00

2

0

Driver

45

0

0

Unspecified

232

0

Cyclist

01 1

0

Motorcycle

Fig. 1. Type of road traffic accident (N=341).

250

Accident

208

Homicide

181

200

Suicide Natural

150 n

100

Homicidal deaths The most common external circumstance/ cause of homicidal deaths (Table 3) was gun­ shot wounds (n=45; 33%), followed by blunt force trauma (inluding assault) (n=35; 26%) and sharp-force injuries (n=28; 21%).

Undetermined

103 73

76

59

56

34

50

8 1 10 16

8 0

15 12

26

46 19

7

0 1 - 5 years

6 - 10 years

11 - 15 years

4 2 0 0 1

16 - 20 years

Unknown

Fig. 2. Manner of death.

Table 3. Homicidal deaths: external cause/circumstance in relation to age group External cause/circumstance

1 - 5 years

6 - 10 years

11 - 15 years

16 - 20 years

Age unknown

Total, n (%)

Gunshot wounds

2

1

7

34

1

45 (33)

Assault/blunt force

3

1

4

30

0

38 (26)

Sharp force

0

0

2

25

1

28 (21)

Strangulation

0

1

0

10

0

11 (8)

Other

3

5

2

4

0

14 (10)

Total

8 (6%)

8 (6%)

15 (11%)

103 (76%)

2 (1%)

136 (100)

54

SAJCH

MARCH 2016 Vol. 10 No. 1

RESEARCH Table 4. Suicidal deaths: method in relation to age group Method

6 - 10 years, n 11 - 15 years, n 16 - 20 years, n Total, n (%)

Hanging

0

6

31

37 (52)

Poisoning/drug overdose

1

2

10

13 (18)

Gunshot wounds

0

3

9

12 (17)

Jumped from height

0

0

6

6 (8)

Other

0

1

3

4 (5)

Total, n (%)

1 (1%)

12 (17%)

59 (82%)

72 (100)

the mortality of children as passengers in road traffic accidents.[20,21] The current legis­ lation pertaining to the use of child restraint devises in SA does not make its use mandatory, but prescribes that the restraint seat should be used ‘if available’ (National Road Traffic Act 93 of 1996, Regulation 213).

The large number of pedestrians involved in accidents is worrisome. In our study, the majority of pedestrians were in the 1 - 5-year group. In a study conducted in Cape Town in 1990 - 1991,[10] it was reported that the majority of pedestrian accidents in children occurred in the later part of the afternoon and in residential areas, and that only 24.3% of children were supervised by an adult at the time of the accident. The annual Road Traffic Report (2010 - 2011) released

by Arrive Alive,[22] indicated that 17.5% of fatalities due to road traffic accidents related to children <19 years.

In our study, burns were the second most common cause of accidental deaths in children, accounting for 8% of the total caseload. The WHO reported that the majority of burns occur in lower- and middle-income countries.[23] The WHO also indicated that burns are the 11th leading cause of death in children aged between 1 and 9 years and emphasised that in a large proportion of burn cases in children, lack of adult supervision contributed to the child sustaining the burn injury.[23] In the Nigerian study,[9] accidental deaths due to burns occurred in 17.5% of cases compared with the study in Nebraska,[7] which reported only 3 cases (n=140) over a 7-year period.

Deaths due to drowning in the study at the PMLL accounted for 7% of the total caseload (n=67), with the majority (n=44; 66%) in children aged 1 - 5 years. The WHO reported that low- and middle-income countries account for 96% of unintentional deaths from drowning.[24] Generally (with the exception of Canada and New Zealand), children <5 years have a higher drowning mortality rate.[24] In Istanbul,[11] drowning accounted for 22% of accidental deaths and 8.7% of total deaths. In Nebraska,[7] 7 cases of drowning were reported in a 7-year period (5% of the total caseload). Several practices

are well advertised to attempt to reduce the risk of drowning in children. These include safety nets (some with attached alarm systems), fences around swimming pools and adult supervision at all times. Most of the drownings in the PMLL study occurred in and around the house, with some of the younger children drowning in buckets and bathtubs. Stricter legislation pertaining to the enclosement of swimming pools and use of safety nets could possibly prevent at least some of these deaths.

Natural disease processes were encounter­ ed in 7% of children in the PMLL study, with 50% occurring in the 1 - 5-year group. Our study excluded infants (children <1 year of age). Nwafor et al.[9] reported natural disease processes in 77.5% of their cases, including infants. According to Statistics SA,[4] the majority of infants in SA die from natural disease processes. Lathrop[16] indicated that natural causes were the most common cause of death among all age groups, accounting for 44.8% of all deaths, with 77% of these in infants. In 86 (9%) cases no cause of death could be ascertained at autopsy alone. The possibility exists that some of these deaths could have been due to underlying natural disease processes such as channelopathies, which could increase the number of natural disease processes. As this could not be established scientifically, these deaths were described as a separate group (unknown).

Homicidal deaths accounted for 14% of the caseload in children at the PMLL. This is less than the 24.2% reported by Fraga et al.,[1] nearly the same as in the Beyaztas et al.[11] study, but more than that in the Okoye and Okoye[7] study (5%). Nwafor et al.[9] reported homicidal deaths in children ≤14 years in only 1.5% of cases. The majority of homicidal deaths occurred in the PMLL and Fraga et al.[1] studies (in children >15 years).

This could explain the fairly low number of homicides in the studies by Nwafor et al.[9] and Okoye and Okoye.[7] In the PMLL study, the majority of homicidal deaths were the result of gunshot wounds (among all age groups). In the studies by Beyaztas et al.[11] and Fraga et al.,[1] the majority of homicides were due to stabbings.

55

SAJCH

MARCH 2016 Vol. 10 No. 1

Seven percent of cases in the PMLL study were as a result of suicide. In Okoye and Okoye’s[7] study suicides accounted for 14% of the caseload (a higher number, also taking into account that the majority of suicides in the PMLL study were among 16 - 20-year- old children; their study only included children ≤14 years[7]). Suicides accounted for 9.4% of cases in the Fraga et al.[1] study and 7.4% of those in the Beyaztas et al.[11] study. The most prevalent external cause/circumstance of suicidal deaths in our study was as a result of hangings (among all age groups), followed by poisonings/overdose of prescription medication. The youngest child who commited suicide was 9 years old (prescription medication overdose). Hanging was the method mostly used in suicides in some studies.[1,7,11] Firearm-related deaths accounted for 6% of cases in the PMLL, Beyaztas et al.[11] and Okoye and Okoye[7] studies, but for 16% of cases in the Fraga et al.[1] study. In the study

by Modelli et al.,[8] only two gunshot-related deaths were reported in a 1-year study in children <12 years. SA has strict legislation pertaining to firearm ownership and use and on storage in appropriate gun safes.

Lathrop[16] indicated that ‘complete medicolegal investigation of childhood fatalities is needed to provide public health agencies with adequate data to evaluate and prevent childhood deaths’. This author also suggests reporting the causes of death separately for each year of life rather than in groups.[16] Accurate statistics on childhood and adoles­cent mortality in SA are lacking. McKerrow et al.[25] indicated that using more than one data set could aid in better statistics, but that ‘it is accepted that these data are incomplete’. The analysing of data sets from medico-legal mortuaries can aid in improving the statistics on deaths in children and young adults in SA and identifying preventable deaths.

Conclusion

Our study indicated that cases of children and young adults comprise a relatively large number of admissions to the medico-legal mortuary. The majority of these deaths are accidental. Many of the accidents could have been prevented with the application of stricter legislation with regard to transportation of children and safety surrounding swimming pools. A child death review programme can aid in identifying causes of death in children, which can be reduced with appropriate changes in legislation.

Acknowledgements. We thank Ms B English, Faculty of Health Sciences, University of Pretoria, for the language editing of the article.

References

1. Fraga AM, Fraga GP, Stanley C, Costantini TW, Coimbra R. Children at danger: Injury fatalities among children in San Diego County.

Eur J Epidemiol 2010;25(3):211-217. [http://dx.doi. org/10.1007/s10654-009-9420-1]

RESEARCH 2. Kanchan T, Menezes RG, Monteiro FNP. Fatal unintentional injuries among young children – a hospital based retrospective analysis. J Forensic Leg Med 2009;16(6):307-311. [http://dx.doi.org/10.1016/j.jflm.2008.12.017] 3. The convention on the rights of the child. http://www.unicef.org/crc/files/ Guiding_Principles.pdf (accessed 28 August 2013).

4. Mortality and causes of death in South Africa: Findings from death notification, 2008. www.statssa.gov.za/publications/P03093/P030932008.pdf (accessed 27 February 2011).

5. Mid-year population estimates, 2009. www.statssa.gov.za/publications/P0302/ P03022009.pdf (accessed 27 February 2011).

6. Chakraborty R. Infections and other causes of death in HIV-infected children in Africa. Paediatr Respir Rev 2004;5:132-139. [http://dx.doi.org/10.1016/j. prrv.2004.01.005] 7. Okoye CN, Okoye MI. Forensic epidemiology of childhood deaths in Nebraska, USA. J Forensic Leg Med 2011;18(8):366-374. [http://dx.doi.org/10.1016/j. jflm.2011.07.013] 8. Modelli ME dos S, Vasques MAdA, Pratesi R. Medicolegal autopsies in children: Experience of a department of legal medicine in Brazil. Forensic Medicine and Anatomy Research 2013;1:40-46. [http://dx.doi.org/10.4236/fmar.2013.13008] 9. Nwafor CC, Ugiagbe EE, Akhiwu WO. A retrospective study of paediatric medicolegal autopsies at the University of Benin Teaching Hostpital, Benin City, Nigeria. Journal of Biomedical Sciences 2013;12(1):76-80.

10. Bass D, Albertyn R, Melis J. Child pedestrian injuries in the Cape metropolitan area – final results of a hospital based study. S Afr Med J 1995;85(2):96-99.

11. Beyaztas FY, Dokgoz H, Saka E, Ҫitici I, Butun C. Evaluation of childhood deaths in Istanbul, Turkey. Middle East Journal of Family Medicine 2007;5(2):38-41.

12. Christian CW, Sege RD, The Committee on Child Abuse and Neglect. Policy statement – child fatality review. Pediatrics 2010;126:592-596. [https://dx.doi. org/10.1542/peds.2010-2006] 13. Population of Pretoria, South Africa. http://population.mongabay.com/ population/south-africa/964137/pretoria (accessed 29 August 2012).

56

SAJCH

14. Butchart A, Peden M, Matzopoulos R, et al. The South African National NonNatural Mortality Surveillance System – rationale, pilot results and evaluation.

S Afr Med J 2001;91(5):408-417.

15. Du Toit-Prinsloo L, Dempers JJ, Verster J, et al. Towards a standardized investigation protocol in sudden and unexpected deaths in infancy in South Africa – a multicenter study of medico-legal investigation procedures and outcomes. Forensic Sci Med Pathol 2003;9:344-350. [http://dx.doi.org/10.1007/s12024-013-9427-5] 16. Lathrop SL. Childhood fatalities in New-Mexico: Medical examiner-invested cases, 2000 - 2010. J Forensic Sci 2013;58(3):700-705. [https://dx.doi.org/10.1111/15564029.12106] 17. Anderson K, Case A, Lam D. Causes and consequences of schooling outcomes in South Africa: Evidence from the survey data. Social Dynamics 2001;27:1-23.

18. WHO. Adolescent health. http://www.who.int/topics/adolescent_health/en/ (accessed 29 April 2015).

19. Canadian Paediatric Society Position Statement. Age limits and adolescents.

Paediatr Child Health 2003;8(9):577.

20. Lennon A, Siskind V, Haworth N. Rear seat safer: Seating position, restraint use and injuries in children in traffic crashes in Victoria, Australia. Accident Analysis and Prevention 2008;40:829-834. [https://dx.doi.org/10.1016/j.aap.2007.09.024] 21. Braver ER, Whitfield R, Ferguson SA. Seating positions and children’s risk of dying in motor vehicle crashes. Inj Prev 1998;4:171-187. [http://dx.doi. org/10.1136/ip.4.3.181] 22. Road Traffic Management Corporation. Arrive Alive – Road Traffic Report, 2011. http://www.arrivealive.co.za/documents/March%202011%20Road%20 Traffic%20Report.pdf (accessed 3 August 2012).

23. WHO. Burns. http://www.who.int/mediacentre/factsheets/fs365/en/ (accessed 10 October 2013).

24. WHO. Drowning. http://www.who.int/mediacentre/factsheets/fs347/en/ (accessed 10 October 2013).

25. McKerrow N, Mulaudzi M. Child mortality in South Africa: Using existing data. http://www.healthlink.org.za/uploads/files/sahr10_5.pdf (accessed 22 January 2016).

MARCH 2016 Vol. 10 No. 1

RESEARCH

Factors present on admission associated with increased mortality in children admitted to a paediatric intensive care unit (PICU) C L Hendricks,1 MB ChB, FC Paed (SA); N H McKerrow,2 MB ChB, BA, DCH (SA), FC Paed (SA), MMed (Paeds), PG Dip Int Res Ethics; R J Hendricks,3 BComm Hons (Actuarial Science), Fellow of the Institute of Actuaries (FIA) Department of Paediatrics and Child Health, Nelson R Mandela School of Medicine, Faculty of Health Sciences, University of KwaZuluNatal, Durban, South Africa 2 Department of Paediatrics and Child Health, University of KwaZulu-Natal, Durban, and Department of Health, KwaZulu-Natal, South Africa 3 Independent consultant, Durban, South Africa 1

Corresponding author: C L Hendricks (cl.hendricks@telkomsa.net) Background. The admission of children to an intensive care unit (ICU) necessitates the selection of children who will benefit most from scarce ICU resources. Decisions should be based on objective data available on outcomes related to particular conditions and resource availability.

Objective. To determine which sociodemographic factors and paediatric scoring systems can be used on admission to identify patients who would derive the most benefit.

Methods. A retrospective review was undertaken of the charts of children admitted to a paediatric ICU (PICU) over a 6-month period.

Charts were analysed according to health status, biographical and demographic data, as well as Pediatric Risk of Mortality (PRISM), Pediatric Logistic Organ Dysfunction (PELOD) and Paediatric Index of Mortality 3 (PIM3) scores to determine which factors were associated with an increased mortality risk.

Results. Two hundred and two children were admitted during the study period. Ninety-six children were included in the study, 79 files were not found and 27 children were ineligible. The median age was 14 months and the mortality rate was 15.6%. The significant factor associated with mortality was severe malnutrition. In total 88% of required data were available for the calculation of both the PRISM and PELOD scores and 95% for PIM3 score. The PRISM, PELOD and PIM3 standardised mortality ratios were 2.5, 4.8 and 2.9, respectively.

P-values for PRISM, PELOD and PIM3 were <0.05.

Conclusion. Severe malnutrition is a statistically significant factor in predicting mortality. This possibly reflects the social context in which the children live. PRISM, PELOD and PIM3 underpredict mortality in our setting. A larger sample is required to verify these outcomes and to determine whether other factors play a role.

S Afr J Child Health 2016;10(1):57-62. DOI:10.7196/SAJCH.2016.v10i1.1048

The availability of intensive care unit (ICU) resources varies widely among low-, middle- and high-income countries, with demand exceed­ ing capacity in many settings.[1] The decision to admit patients to ICU has to take into consideration available resources and the likelihood of a successful outcome.[2] Admission criteria should therefore assess whether ICU will add value to the patient’s overall management and eventual outcome, and consider that a patient’s health status, diagnosis and severity of illness all influence his/her risk of death.[3] Social factors determining access to health services include income, location and transport. These all contribute to a delay in presentation, possibly leading to more advanced disease and poorer outcomes. Access to social grants is associated with decreased illness and improved growth monitoring and therefore plays an important role in mitigating these factors.[4] HIV has placed increased demands on health services in South Africa (SA),[5] with many infected children presenting in early infancy with life-threatening illnesses. The introduction of antiretroviral (ARV) therapy has improved the outcome of HIV-infected children, leading to a review of paediatric ICU (PICU) admission criteria with greater accommodation of these children.[5] Globally severe malnutrition affects 13 million children under the age of 5 years, with case fatality rates between 20% and 30%.[6] The second highest burden is seen in sub-Saharan Africa.[6] A low weight centile has been shown to be an independent risk factor for mortality in intensive care.[7] Clinical scoring systems are objective measures that may be used to assess the performance of ICUs.[8] They can thus aid the 57

SAJCH

optimal use of PICU resources by evaluating the quality of medical care received. [9] The Pediatric Risk of Mortality (PRISM) score uses clinical, physiological and laboratory variables in the first 24 hours of PICU admission, to attain a score assessing severity of illness. [10] Although validated for use in the USA,[11] its validity in an SA population has been questioned.[12] The Paediatric Index of Mortality (PIM) score is a simpler tool using eight parameters to assess mortality risk on admission.[8] The latest version of PIM, PIM3, has not been evaluated in an SA setting.

Additionally, organ dysfunction scores can be used to assess the severity of illness at various points during the ICU admission. The Pediatric Logistic Organ Dysfunction (PELOD) score is one such score quantifying organ dysfunction on admission or throughout the clinical course where parameters are collected daily.[13] An eight-bed PICU was commissioned in Grey’s Hospital in 2003 with facilities to ventilate up to four patients simultaneously. This unit serves the 1.2 million children in the western half of KwaZuluNatal, admitting patients from both inside and outside the hospital.

Externally referred patients include those from hospitals within a 300-km radius of Pietermaritzburg. The decision to admit is made by the intensivist in charge of the unit, who balances multiple factors, including the availability of beds, the possibility of discharging patients and the stability for transfer of the patient.

In this study, demographic factors (age, gender, and source of referral), underlying health status (immunisation, nutritional and HIV status), diagnosis on admission and scoring systems (PRISM, PELOD and PIM3) were reviewed to assess their association with

MARCH 2016 Vol. 10 No. 1

RESEARCH survival in this setting. The purpose of the review was to establish more objective measures to be used in considering PICU admission.

Table 1. Demographic characteristics Died, n (% of admissions in category)

Variable

Methods

Survived, n (% of admissions in category)

Total, n (% of total admissions)

Gender

The PICU admission register at Grey’s Hospital was screened to identify those patients meeting the inclusion criteria during the 6-month period, 1 January - 30 June 2011. A retrospective chart review was undertaken for all these admissions. Patients were identified from their clinical records retrieved from the medical records department. Inclusion criteria were age <13 years, first admission to PICU, and baseline blood investigations done within 3 hours of admission. Exclusion criteria were readmissions for the same condition during a single hospital stay, elective surgical procedures, and children in the home-based tracheotomy programme. Data were collected by the primary author (CLH) and one additional doctor. Data checks were performed in Microsoft Access (USA) by author RJH to identify missing or incorrect values, and these were corrected where possible.

The corrected data were entered into an Excel spreadsheet (Microsoft, USA) and statistical software EpiInfo 7 (Centers for Disease Control, USA) was used for analysis.

The nutritional classification of the study population is presented according to the nutritional classification mentioned in the clinical notes, the Wellcome classification for children older than 10 years and the World Health Organization (WHO) weight-for-age z-score (WAZ score) for those below 10 years of age. Surrogate WAZ scores were allocated to children older than 10 years as follows: weight above the 25th centile equated to a WAZ score >−2, weight between the 5th and 25th centile were classified as a WAZ score between −2 and −3 and those below the 5th centile were given a WAZ score <−3.

Weights were documented for 95 patients but the heights were documented for only 18, making it impossible to apply the weightfor-height z-score (WHZ score).

The PRISM, PELOD and PIM3 scores were calculated according to the prescribed calculations.

The PRISM and PELOD data were ~88% complete (i.e. had ~12% missing data) while information for the PIM3 was 95% complete. In order to quantify the possible effect of missing data elements in PRISM and PELOD, normal values were allocated.

PIM3 gives instructions on what to do if a data element is missing and this was applied accordingly. The deficiencies in the data were accounted for as accurately as possible.

The analysis assessed the association of biographical factors, health status and disease pattern, as well as PRISM, PELOD and PIM3 scores with outcome. A multivariate analysis analysed the relationship between age, grant

0.986

Male

8 (15.7)

43 (84.3)

51 (53.1)

Female

7 (15.6)

38 (84.4)

45 (46.9)

0 - 1 mo

1 (11.1)

8 (88.8)

9 (9.4)

>1 - 6 mo

8 (25.8)

23 (74.1)

31 (32.3)

>6 - 12 mo

1 (20.0)

4 (80.0)

5 (5.2)

>1 - 5 yr

3 (14.3)

18 (85.7)

21 (21.9)

>5 - 10 yr

1 (5.3)

18 (94.7)

19 (20.0)

>10 - 13 yr

1 (9.1)

10 (90.9)

11 (11.0)

Internal

4 (8.7)

42 (91.3)

46 (47.9)

External

11 (22.0)

39 (78.0)

50 (52.1)

Age

0.342

Referral status

0.0729

Grant status

0.22

Yes

4 (8.9)

41 (91.0)

45 (46.9)

No

8 (17.8)

37 (82.2)

45 (46.9)

Unknown

3 (50.0)

3 (50.0)

6 (6.2)

Immunisations up to date

0.833

Yes

11 (15.3)

61 (84.7)

72 (75.0)

No

2 (13.3)

13 (86.6)

15 (15.6)

Unknown

2 (22.2)

7 (77.7)

9 (9.4)

HIV status

0.347

Positive

2 (18.2)

9 (81.8)

11 (11.5)

Exposed

5 (29.4)

12 (70.5)

17 (17.7)

Negative

6 (12.2)

43 (87.7)

49 (51.0)

Unknown

2 (10.5)

17 (89.4)

19 (19.8)

Distance from hospital (km)*

0.7154

0 - 50

2 (13.3)

13 (86.7)

15 (15.6)

>50 - 100

3 (25.0)

9 (75.0)

12 (12.5)

>100 - 200

6 (27.3)

16 (72.7)

22 (22.9)

>200

0 (0)

1 (100)

1 (1.0)

Nutritional status: clinical notes

0.003

SAM

7 (46.7)

8 (53.3)

15 (15.6)

MAM

2 (40.0)

3 (60.0)

5 (5.2)

Normal

4 (5.8)

64 (94.2)

68 (70.8)

Not growing well

2 (25.0)

6 (75.0)

8 (8.3)

WAZ-score

0.033

>−2

4 (7.3)

51 (92.7)

55 (57.3)

−2 - −3

5 (26.3)

14 (73.7)

19 (19.8)

<−3

6 (27.3)

16 (72.7)

22 (22.9)

SAM = severe acute malnutrition; MAM = moderate acute malnutrition. *Externally referred patients.

58

SAJCH

p-value

MARCH 2016 Vol. 10 No. 1

RESEARCH status and nutritional status. R2 is a number indicating how well data fit the statistical model.

The Wald χ2 test was used to determine significance using a 5% significance level.

The outcome measures were death or survival. Ethical approval was obtained from the Biomedical Research Ethics Committee, University of KwaZulu-Natal.

Table 3. Diagnostic categories

During the study period there were 202 ad­ missions to Grey’s Hospital PICU. Seventynine files were not found and 27 patients were excluded, including 13 readmissions, 4 children older than 13 years and 10 patients admitted following elective surgery. Con­ sequently, 96 patients were included in the analysis. Eight male and 7 female patients died, with a total mortality rate of 15.6%.

Table 1 illustrates the demographic data of the study population. Male patients constituted 53%. Ages ranged from 0 to 142 months with a median of 14 months.

Forty-three percent were under 1 year of age, including nine neonates who were admitted when the neonatal ICU was full. Of the 15 deaths, 9 were of children under the age of 6 months (60.0%) even though this age group constituted only 22% of the total admissions. Two deaths (13.3%) occurred in children older than 5 years although this age group constituted 30% of admissions.

Forty-eight percent of admissions were internal referrals from within Grey’s Hospital and 52% were referred from outside. The mortality rate among internal referrals was 8.7% compared with 22.0% among those referred from outside the hospital. This difference was not statistically significant however, with a p-value of 0.0729.

The social grant status was known for 90 children, 50% of whom were receiving a grant. Of these, 42 were receiving a child support grant, 2 a care dependency grant and in 1 patient the type of grant was unknown. Those children who were not receiving a grant had a two-fold higher mortality than those who were (17.8% v.

8.9%). This difference was not statistically significant, with a p-value of 0.22.

In 75% of children the vaccination status was up to date, in 16% it was delayed and in 8% it was not documented. In all three of the above groups the mortality rate ranged from 12 to 15% (p=0.833).

Age

0.13

Age and grant status

0.14

Age and nutritional status

0.32

Total, n (% of total admissions)

Medical

13 (20.0)

52 (80.0)

65 (67.7)

0 (0)

9 (100)

9 (9.4)

Meningitis

0 (0)

1 (100)

1 (1.0)

Seizures/status epilepticus

0 (0)

2 (100)

2 (2.1)

Encephalopathy

0 (0)

5 (100)

5 (5.2)

Guillian Barré

0 (0)

1 (100)

1 (1.0)

0 (0)

4 (100)

4 (5.9)

Cardiac failure

0 (0)

2 (100)

2 (2.2)

Pericardial effusion

0 (0)

1 (100)

1 (1.0)

Takayasu arteritis

0 (0)

1 (100)

1 (1.0)

0 (0)

5 (100)

5 (5.4)

0 (0)

5 (100)

5 (5.4)

2 (66.7)

1 (33.3)

3 (3.2)

Liver failure

2 (100.0)

0 (0)

2 (2.1)

CVS

Endocrine Diabetic ketoacidosis GIT Biliary atresia

0 (0)

1 (100)

1 (1.0)

Other: poisoning

0 (0)

1 (100)

1 (1.0)

Renal

0 (0)

5 (100)

5 (5.2)

Renal failure

0 (0)

2 (100)

2 (2.1)

Nephrotic syndrome

0 (0)

1 (100)

1 (1.0)

APSGN

0 (0)

1 (100)

1 (1.0)

HIVAN

0 (0)

1 (100)

1 (1.0)

8 (25.8)

23 (74.2)

31 (32.3)

Pneumonia

8 (30.7)

18 (69.2)

26 (26.0)

Upper airway obstruction

0 (0)

4 (100)

4 (4.2)

Pneumothorax

0 (0)

1 (100)

1 (1.0)

Respiratory

Shock

3 (42.9)

4 (57.1)

7 (7.3)

Septic shock

2 (40.0)

3 (60)

5 (5.2)

Hypovolaemic shock

1 (50.0)

1 (50)

2 (2.1)

2 (6.5)

29 (93.5)

31 (32.3)

Surgical non-trauma

2 (7.7)

24 (92.3)

26 (27.1)

Ruptured appendix

0 (0.0)

7 (100)

7 (7.3)

Bowel obstruction

1 (16.7)

5 (83.3)

6 (6.2)

Bilateral testicular torsion

0 (0)

1 (100)

1 (1.0)

Post adenotonsillectomy

0 (0)

1 (100)

1 (1.0)

Foreign body aspiration

0 (0)

3 (100)

3 (3.1)

Oesophageal stricture

0 (0)

2 (100)

2 (2.1)

Septic arthritis/myositis

0 (0)

2 (100)

2 (2.1)

Biliary atresia

0 (0)

1 (100)

1 (1.0)

Relaparotomy

Surgical

Table 2. Multivariate analysis of age, grant status and nutritional status R2

Survived, n (% of admissions in category)

CNS

Results

Model

Diagnostic category

Died, n (% of admissions in category)

1 (33.3)

2 (66.7)

3 (3.1)

Surgical trauma

0 (0)

5 (100)

5 (5.2)

Polytrauma

0 (0)

2 (100)

2 (2.1)

Burns

0 (0)

1 (100)

1 (1.0)

Snake bites

0 (0)

2 (100)

2 (2.1)

CNS = central nervous system; CVS = cardiovascular system; GIT = gastrointestinal tract; APSGN = acute post-streptococcal glomerulonephritis; HIVAN = HIV-associated nephropathy.

59

SAJCH

MARCH 2016 Vol. 10 No. 1

RESEARCH Eleven patients were HIV-infected (11%), 17 were HIV-exposed (18%), 49 were HIV-negative (51%) and in 19 (20%) the HIV status was unknown. Seventy-two percent of HIV-infected patients were already on ARVs (8/11). HIV status was not found to have a statistically significant effect on survival (p=0.347).

The distance individual children were transferred and their eventual outcome is also demonstrated. Seventy-three percent of the total deaths occurred in children referred from outside Grey’s Hospital, while these children made up 52% of admissions.

However, the apparent increased mortality rate with increasing distance from referral facility was not statistically significant (p=0.7154).

There was a significantly increased mortality rate in children who were classified as severely malnourished. Of the 15 children with severe acute malnutrition (SAM), 7 died (46.7%), compared with 4 deaths (5.9%) in those children with a normal nutritional status.

This was statistically significant (p=0.003). For the WAZ score, those patients who plotted below −2 z-score made up 73% of the deaths (p=0.033).

In Table 2, a multivariate analysis shows the relationship between grant status, age category and nutritional status. All three variables had a statistically significant correlation with each other. The majority of children who were not receiving a grant (66.7%) were under the age of 12 months, indicating that age is the covariate factor. These children also had a higher rate of malnutrition (64.3%).

Seventy-one percent of children who were receiving a grant were older than 12 months.

Table 3 portrays the diagnostic categories of the patient population. Only the primary admission diagnosis was recorded.

The majority of children (67.7%) had a medical diagnosis, with respiratory illness comprising 32.3% of all admissions. Over half of all deaths (53.3%) were associated with respiratory illnesses (n=8/15) and both patients with gastrointestinal tract (GIT) problems had liver failure and died. The diagnoses of the two patients who died in the surgical category were bowel obstruction and post relaparotomy for abdominal sepsis.

The average length of stay of patients was 4 days, with a range of 0 - 20 days. A fifth of deaths occurred within 24 hours and almost half (46.7%) within the first 48 hours.

Table 4 indicates the number of parameters available for the calculation of the PRISM, PELOD and PIM3 scores. Twenty-five patients had all the information required to calculate the PRISM score and 23 the PELOD score. In total 88% of data were complete for the calculation of both the PRISM and PELOD scores. The data completion rate for PIM3 was much higher (95%).

The dominant missing parameters included INR (61%) and partial pressure of carbon dioxide (PaCO2) (22%) for the PRISM calculation and serum glutamic oxaloacetic transaminase (SGOT) (41%) for the PELOD calculation.

As expected, higher scores were associated with a higher mortality.

Noteworthy, however, is that all three scores significantly underpredicted the number of deaths.

Fig. 1 illustrates the relationship between the actual and expected deaths for the PRISM, PELOD and PIM3 scores. The standardised mortality ratios (SMRs) for PRISM, PELOD and PIM3 were 2.5, 4.8 and 3.3, respectively.

Table 5 displays the diagnostic categories used to calculate PIM3.

Only 20% of the study population were able to be categorised according to this list. Of note is that HIV, nutritional status and other communicable diseases do not form part of this categorisation.

Table 4. Parameters retrieved (n) to complete the PRISM, PELOD and PIM3 scores Data quality percentage

PRISM, n (% of total admissions)

PELOD, n (% of total admissions)

PIM3, n (% of total admissions)

100

25 (26)

23 (24)

81 (84.3)

>75 - 99

62 (64)

62 (64)

9 (9.4)

>50 - 75

5 (5)

6 (6)

2 (2.1)

>25 - 50

2 (2)

3 (3)

3 (3.2)

≤25

2 (2)

2 (3)

1 (1.0)

Total

96

96

96

Table 5. PIM3 diagnostic categories[3] Diagnostic category

Zero is chosen if not sure

Low-risk diagnosis

0. None 1. Asthma is the main reason for ICU admission 2. Bronchiolitis is the main reason for ICU admission 3. Croup is the main reason for ICU admission 4. Obstructive sleep apnoea is the main reason for ICU admission 5. Diabetic ketoacidosis is the main reason for ICU admission 6. Seizure disorder is the main reason for ICU admission

High-risk diagnosis

1. Spontaneous cerebral haemorrhage 2. Cardiomyopathy or myocarditis 3. Hypoplastic left heart syndrome

16

4. Neurodegenerative disorder

14

5. Necrotising enterocolitis is the main reason for ICU admission

12 Deaths, n

0. None

Very-high-risk diagnosis

10 8

0. None 1. Cardiac arrest preceding ICU admission

6

2. Severe combined immune deficiency

4

3. Leukaemia or lymphoma after first induction

2

4. Bone-marrow transplant recipient

0 Actual

Expected PRISM

Expected PELOD

Expected PIM3

Fig. 1. Actual v. expected deaths as determined by PRISM, PELOD and PIM3.

60

SAJCH

MARCH 2016 Vol. 10 No. 1

5. Liver failure is the main reason for ICU admission

RESEARCH Discussion

Although 8 - 10% of SA’s gross domestic product (GDP) is spent on health,[1] critical care remains a scarce resource.[14] In light of these resource difficulties one needs to ask ‘Who are the most likely to survive on the best evidence available?’[15] This question highlights the difficult ethical questions raised in providing these services.

Ultimately, the provision of any health service has to take into account basic and preventive medical care[1] to mitigate the need for critical care. The limitations to this study are the retrospective nature, small sample size and missing data required to calculate PRISM and PELOD scores. It does, however, highlight some important issues.

The overall mortality rate for children in this study was 15.6%, which correlates with the 14.9% reported from Tygerberg Children’s Hospital in Cape Town.[16] This is lower than other middle-income countries such as Egypt (33.1%)[17] Saudi Arabia (37.4%)[11] and India (24.3%).[9] In China, however, this figure drops to 2.64%,[18] which is more in keeping with high-income countries where rates are generally <10%. The Netherlands, USA and Australia have total mortality rates of 6.6%, 4.86% and 4.25%, respectively.[18] Children <1 year of age comprised only half (46.9%) of all ad­ missions yet accounted for two-thirds of the deaths, with a mortality rate of 22.2% in the age group. This correlates with experience in Cairo[17] and Saudi Arabia,[11] where the highest mortality rates (43.95% and 65.6%, respectively) were found in patients <1 year of age. These facts point to increased vulnerability in this age group and probably a need to be more aggressive in their management to improve outcomes.

Malnutrition remains a major public health burden in the developing world. Numa et al.[7] demonstrated that extremes of weight have a significant impact on survival in ICU. Patients with weights below the 3rd centile had more than double the mortality of those at the 75th centile.[7] Nutritional status was found to be statistically significant in this study at 5% (p<0.05). Sixty-four percent of infants were malnourished and 68% of malnourished children were not in receipt of a social grant. On multivariate analysis with age and nutritional status, malnutrition remained significant (R2 significant change with malnutrition in the model). As children with primary malnutrition are not usually accepted into the ICU, the study population consisted of patients with malnutrition secondary to other underlying medical conditions. This probably reflects the severity or chronicity of the underlying condition and should be considered a poor prognostic marker when present and possibly a reason to preclude admission.

Forty-six percent of deaths occurred within 48 hours of admission.

This high early mortality may reflect the quality of care in the periphery and during transfer, as well as resource and logistical challenges in the referral system. It was beyond the scope of this study to determine the impact of these factors. In KwaZulu-Natal however, 69% of public hospitals have no high-care or ICU facilities,[14] making stabilisation and presumably monitoring of patients prior to transfer a challenge.

There was a 50% greater chance of dying among children who were not receiving a grant. On multivariate analysis it was found that a significant proportion of these children (66.7%) were under the age of 1 year and that age, not grant status, was the significant factor (R2 insignificant change with grant status in the model). This illustrates that age, and the vulnerabilities associated with infancy, is the covariate factor in interpreting the difference in mortality between those receiving and not receiving a social grant.

HIV status was not found to affect mortality in this study, although a large proportion of the study population had an unknown HIV status on admission. With the advent of ARV therapy, mortality in HIV-infected children has decreased significantly in both high- and middle-income countries.[5] Ultimately the aim should be to reduce 61

SAJCH

the need for intensive care in these children and instead focus on improved systems to prevent mother-to-child transmission and the early initiation of ARVs where necessary.[16] Increasing PRISM, PELOD and PIM3 scores were all associated with increasing mortality rates, as expected (p<0.05). More remarkable, however, is that all three scores significantly underpredicted deaths in this population. The SMRs for PRISM, PELOD and PIM3 were 2.5, 4.8 and 3.3, respectively. These rates are higher than those found in the Netherlands[8] (0.95 with PRISM and 0.88 with PIM) while in India Taori et al.[9] reported an SMR of 1 with PRISM. Wells et al.[12] suggested that the different demographic characteristics and disease patterns of SA ICU patients may influence PRISM scoring.[12] PIM and PIM2 showed SMRs of 1.10 and 0.90, respectively, in an SA study.[19] The increased SMR in the study population may point towards failure of these scores to include underlying factors unique to our patient population. In PIM3, failure to include diagnoses common to our population (HIV infection, malnutrition, tuberculosis and other communicable diseases) leads to a large proportion of patients being allocated a score of zero. Many of these children would probably qualify for the high- and very-high-risk categories in an SA model.

The study population, however, was too small to make accurate analyses in this regard.

The value of using the PRISM, PELOD and PIM3 in our setting may be to use the SMR to monitor the performance of the unit itself over time or to compare different units with the same resource limitations.[9] Solomon et al.[19] commented that PIM and PIM2 should not be used as screening tools as they are not accurate enough. The scores may need to be recalibrated or recalculated for the SA population[11] to more accurately determine disease severity.

Conclusion

This study highlights severe malnutrition as a statistically significant factor associated with mortality regardless of the cause of the malnutrition. This suggests there may be no value in differentiating between primary and secondary malnutrition in PICU admissions as outcomes remain poor.

In addition, children under the age of 1 year had the highest proportion of malnutrition and deaths. This factor points strongly towards the need to practise early aggressive intervention in infants admitted to an ICU to improve outcomes in this age group.

PRISM, PELOD and PIM3 were all found to significantly underpredict death in our setting. It may not be relevant to use these scores to predict death in SA but the SMR can be used by units to compare their individual performance. The authors found the PIM3 to be the score with the most potential for reproducibility in that it has fewer variables and clear instructions on how to use it.

References 1. Fowler RA, Adhikari NK, Bhagwanjee S. Clinical review: Critical care in the global context – disparities in burden of illness, access, and economics. Crit Care 2008;12(5):225. [http://dx.doi.org/10.1186/cc6984] 2. Cowburn C, Hatherill M, Eley B, et al. Short-term mortality and implementation of antiretroviral treatment for critically ill HIV-infected children in a developing country. Arch Dis Child 2007;92(3):234-241. [http://dx.doi.org/10.1136/ adc.2005.074856] 3. Stranley L, Clements A, Parslow RC, et al. Paediatric Index of Mortality 3: An updated model for predicting mortality in pediatric intensive care. Pediatric Critical Care Medicine 2013;14(7):673-681. [http://dx.doi.org/10.1097/ PCC.0b013e31829760cf] 4. DSD, SASSA and UNICEF. Heinrich C, Hoddinott J, Samson M, Mac Queen K, van Niekerk I, Renaud B, eds. The South African Child Support Grant Impact Assessment: Evidence from a Survey of Children, Adolescents and their Households. Pretoria: UNICEF South Africa, 2012. http://www.unicef.org/ southafrica (accessed February 2015).

5. Argent AC. Managing HIV in the PICU: The experience at the Red Cross War Memorial Children’s Hospital in Cape Town. Indian J Pediatr 2008;75(6):615. http://link.springer.com/article/10.1007%2Fs12098-008-0118-2 (accessed May 2015).

MARCH 2016 Vol. 10 No. 1

RESEARCH 6. Collins S. Treating severe acute malnutrition seriously. Arch Dis Child 2007;92(5):453-461. [http://dx.doi.org/10.1136/adc.2006.098327 (accessed 19 July 2015).

7. Numa A, McAweeney J, Williams G, Awad J, Ravindranathan H. Extremes of weight centile are associated with increased mortality in paediatric intensive care. Crit Care 2011;15(2):R106). [http://dx.doi.org/10.1186/cc10127] 8. Gemke RJ, van Vught J. Scoring systems in pediatric intensive care: PRISM III versus PIM. Intensive Care Med 2002;28(2):204-207. [http://dx.doi. org/10.1007/s00134-001-1185-2] 9. Taori RN, Lahiri KR, Tullu MS. Performance of PRISM (Pediatric Risk of Mortality) score and PIM (Pediatric Index of Mortality) score in a tertiary care pediatric ICU. Indian J Pediatr 2010;77(3):267-271. [http://dx.doi.org/10.1007/ s12098-010-0031-3] 10. Rady H, Mahomed S, Mohssen N, ElBaz M. Application of different scoring systems and their value in pediatric intensive care unit. Gaz Egypt Paediatr Assoc 2014;62:5964. http://www.sciencedirect.com/science/article/pii/S1110663814000494 (accessed October 2014).

11. Alsuheel A, Shati A. Factors predicting mortality in pediatric intensive care unit in a tertiary care center Southwest Region, Saudi Arabia. J Med Med Sci 2014;5(5):113-120. [http:/dx.doi.org/10.14303/jmms.2014.085] 12. Wells M, Riera-Fanego JF, Luyt DK, Dance M, Lipman J. Poor discriminatory performance of the Pediatric Risk of Mortality (PRISM) score in a South African intensive care unit. Crit Care Med 1996;24(9):1507-1513. [http:// dx.doi.org/10.1097/00003246-199609000-00013]

62

SAJCH

13. Leteurtre S, Duhamel A, Grandbastein B, et al. Daily estimation of the severity of multiple organ dysfunction syndrome in critically ill children.

Can Med Assoc J 2010;182(11):1181-1187. [http://dx.doi.org/10.1503/ cmaj.081715] 14. Bhagwanjee S, Scribante J. National audit of critical care resources in South Africa â&#x20AC;&#x201C; unit and bed distribution. S Afr Med J 2007;97(12 Pt 3):1311-1314.

15. Jeena PM, McNally LM, Stobie M, Coovadia HM, Adhikari MA, Petros AJ. Challenges in the provision of ICU services to HIV infected children in resource poor settings: A South African case study. J Med Ethics 2005;31(4):226-230. [http://dx.doi.org/10.1136/jme.2003.004010] 16. Rabie H, de Boer A, van den Bos S, Cotton MF, Kling S, Goussard P. Children with human immunodeficiency virus infection admitted to a pediatric intensive care unit in South Africa. J Trop Pediatr 2007;53(4):270-227. [http://dx.doi.org/10.1093/ tropej/fmm036] 17. Rady H. Profile of patients admitted to pediatric intensive care unit, Cairo University Hospital: 1-year study. Ain-Shams J Anesthesiol 2014;7(4):500-503. [http://dx.doi.org/10.4103/1687-7934.145680] 18. Bilan N, Galehgolab BA, Emadaddin A, Shiva SH. Risk of mortality in paediatric intensive care unit, assessed by PRISM III. Pak J Biol Sci 2009;12(6):480-485. [http://dx.doi.org/10.3923/pjbs.2009.480.485] 19. Solomon LJ, Morrow BM, Argent AC. Paediatric Index of Mortality scores: An evaluation of function in the paediatric intensive care unit of the Red Cross War Memorial Childrenâ&#x20AC;&#x2122;s Hospital. S Afr J Crit Care 2014;30(1):8-13. [http://dx.doi. org/10.7196/SAJCC.166]

MARCH 2016 Vol. 10 No. 1

RESEARCH

The association between chronic undernutrition and malaria among Ethiopian children aged 6 - 59 months: A facility-based case-control study H Y Hassen,1 BSc, MPH; J H Ali,2 MD, MSc, CRM, CME 1 2

Department of Public Health, College of Health Sciences, Mizan Tepi University, Mizan Teferi, Ethiopia School of Public Health, College of Health Sciences, Addis Ababa University, Ethiopia

Corresponding author: H Y Hassen (abdulhamidy71@gmail.com) Background. Malaria and undernutrition remain two major causes of childhood mortality in sub-Saharan Africa, including Ethiopia. The synergetic relationship between undernutrition and infection is widely documented but the relationship with malaria remains controversial. More studies are needed to address this and provide information to develop effective strategies for malaria control and prevention of undernutrition.

Objective. To assess the relationship between malaria and chronic undernutrition in children aged 6 - 59 months at Bahir-Dar special zone, Ethiopia.

Methods. A facility-based case-control study was employed to assess the relationship between malaria and chronic undernutrition at Bahir-Dar. A total of 621 children aged between 6 and 59 months, with a 1:3 ratio of confirmed malaria cases to controls, were enrolled.

Clinical data and anthropometric measurements were taken, and blood film taken and examined. Anthropometric data were converted into nutritional indices using World Health Organization Anthro software version 3.2.2 and exported to SPSS for cleaning and analysis.

Results. Prevalence of stunting and underweight was 50.3% and 34.2% among cases, respectively. Stunting and underweight were significantly associated with confirmed cases of malaria after sociodemographic and other variables were controlled. Other important predictors were rural residence, sleeping under long-lasting impregnated nets, and using indoor residual spray.

Conclusion. Chronic undernutrition was closely associated with malaria infection. Major predictors for contracting malaria were stunting, underweight, rural residence, not using long-lasting impregnated nets and indoor residual spray, male sex and low educational status of mother/caregiver. Integrated actions targeting these factors are necessary to reduce the prevailing problem.

S Afr J Child Health 2016;10(1):63-67. DOI:10.7196/SAJCH.2016.v10i1.1052

Malaria is the most significant human parasitic disease and remains a major cause of morbidity and mortality worldwide, particularly in developing countries. It affects children below the age of 5 years and pregnant women more severely. Globally, an estimated 3.3 billion people were at risk of malaria, with ~80% of cases and 90% of deaths in 2011 occurring in the World Health Organization (WHO) African region.[1] Hunger and malnutrition also remain among the most devastating problems facing the majority of the world’s poor and needy, and continue to dominate the health of the world’s poorest nations.

The United Nations Food and Agriculture Organization estimates that nearly 870 million people, or one in eight people globally, suffered from chronic undernourishment between 2010 and 2012. [2] Nearly 30% of the population of the developing world, including infants, children, adolescents, adults and older persons, are currently suffering from one or more of the multiple forms of malnutrition.[3] Based on the 2013 United Nations Children’s Fund report, 38% of children below the age of 5 years suffer from chronic malnutrition or stunting in sub-Saharan Africa, with malaria and undernutrition being the two major causes of childhood mortality.[3] In Ethiopia, both undernutrition and malaria are major public health problems, with an estimate of 55.7 million people (68% of the population) at risk for malaria, and nearly 80% of the existing 736 districts malaria endemic.[4] In 2009/2010, malaria was the leading cause of outpatient visits and health facility admissions, accounting for 14% of outpatient visits and 9% of admissions.[4] Chronic malnutrition prevalence reported for the same period, manifesting in the form of stunting and underweight, was 44% and 29%, respectively, indicating the burden of the problem in the country.[5] The synergy of malnutrition and infections is widely documented elsewhere,[6,7] though the relationship between malnutrition and 63

SAJCH

malaria is conflicting.[8-13] This presents a challenge in attempting to control the issue. This study was conducted to investigate the association between chronic malnutrition and malaria, and provide necessary information to develop effective strategies for the control of malaria and prevention of malnutrition.

Methods

Study design and setting

An institution-based case-control study was conducted to assess the association of undernutrition and malaria among children in BahirDar special zone, which has 10 health centres (3 rural and 7 urban).

The zone is situated on the southern shore of Lake Tana, 564 km north-west of Addis Ababa, with an elevation of 1 840 m above sea level. The zone is heterogeneous with various ethnic populations and malaria is endemic with seasonal transmission, which makes the site ideal for investigating the association of malaria and undernutrition. From the available health facilities at the time of the study, two rural and four urban facilities were purposively selected for our assessment.

Ethical considerations

Ethical approval was obtained from the School of Public Health Research Ethics Review Committee of Addis Ababa University, College of Health Sciences. Permission was also secured from the Amhara Regional Health Bureau. Officials of each facility were also contacted, and permission was granted to conduct the study.

Informed written consent was obtained from each subject for their participation and the right to withdraw from the study at any time was also communicated to all of them. Antimalarial treatment was also provided for children with confirmed malaria and all undernourished children were referred to a nutritional rehabilitation centre, according to the national guideline.

MARCH 2016 Vol. 10 No. 1

RESEARCH Sample size determination and subjects

The required sample size was determined using the double population proportion formula, by considering the magnitude of the two different types of malnutrition of the region: stunting (52.0%) and underweight (33.4%),[4] with level of significance α=0.05 and margin of error d=5%. Power was 80% with non-response rate of 10% and expected odds ratio of 1.75. Two different sample sizes were calculated based on underweight prevalence (n=621) and stunting prevalence (n=448). The sample size that yielded the largest number (n=621) was taken as our working sample since it accommodated all the above scenarios.

A sample size of 621 was estimated and allocated proportionally across the six health centres: Tis-Abay (n=108), Meshenti (n=108), Shinbit (n=104), Abaymado (n=100), Bahir-Dar (n=104), and Han (n=97). All children who had confirmed malaria were enrolled as cases until the required sample sizes for the respective health centres were reached. For every confirmed malaria case, three controls that visited the facility for any reason other than malaria (upper respiratory tract infection, ear infection, eye infection, minor injury and minor dermal cases) were interviewed based on their order of visiting. Children who had acute lower respiratory tract infection, diarrhoea, measles, severe illness or chronic illness (such as tuberculosis, HIV and AIDS) were excluded from the study.

Data collection and processing

The important variables included in the questionnaire were socio­ demographic, clinical and laboratory data. Twelve data collectors fluent in the local languages of the study zone (two clinical nurses from each health centre), laboratory technicians and two supervisors with relevant experience were recruited and trained for 2 days on the method of data collection. The training addressed issues such as the content of the questionnaire, basic interviewing skills, filling out the questionnaire, and weight and height measurements.

Anthropometric measurements

Weight was measured in kilograms without shoes, using a digital bath scale with good precision, and recorded to the nearest 100 g. During weight measurement, the scale was repeatedly checked for accuracy against a known weight. Height or length was taken barefoot using a stadiometre. A vertical tape fixed perpendicular to the ground on the wall or recumbent length measurement tape fixed for children under 2 years of age was used and height or length recorded to the nearest centimetre.

Clinical assessments

Clinical information that included illnesses in the preceding 2 weeks and type of treatments received was assessed by trained outpatient staff of the respective health centre. Oedema was checked by grasping both feet with thumbs on the top of the feet and pressing thumbs gently for 3 seconds (or count 101, 102, 103 if there was no stopwatch); if pitting remained in both feet after lifting the hand, the child was considered as having nutritional oedema. Axillary temperature was also measured using a digital thermometer two times and the average was recorded.

USA) for cleaning and analysis. For computing the nutritional indices, the data were exported to the WHO’s Anthro statistical package version 3.2.2 (Switzerland). Children were classified as stunted (chronic undernutrition) and underweight when the height-for-age z-scores (HAZ) and weight-for-age z-scores (WAZ) were below minus two z-scores, respectively. Severe stunting and underweight were defined when both HAZ and WAZ were below minus three z-scores.

The results are presented in percentages and graphs accordingly.

Binary logistic regression was employed to examine the associations between sociodemographic variables and undernutrition with malaria and to identify variables for consideration in multivariate analysis.

To ascertain the association between the dependent variables and the explanatory variables, simultaneously controlling for the aforementioned explanatory variables (all sociodemographic characteristics and other covariates associated in bivariate with p<0.2 were used and entered), stepwise logistic regression was applied and adjusted odds ratios (AORs) and 95% confidence intervals (CIs) were calculated. Multicollinearity was also checked for WAZ and HAZ using variance inflation factor. In all analyses, p<0.05 was considered to be statistically significant.

Results

Of the total enrolled subjects, 607 (149 cases and 458 controls) participated in the study, making the response rate 96.2% for cases and 98.3% for controls. Over three-quarters (78%) of caregivers were between 20 and 39 years old and most of them were female and married, in both groups. The majority in both groups were housewives and less than half had no formal education. The median household size was five people for both cases and controls, respectively. Less than a quarter were earning above ETB2 000 (USD100) and urban residents constituted more than half in both groups (Table 1).

Measures of children’s sociodemographic characteristics and nutritional status

Various child characteristics are displayed by type of study group (Table 2). As shown, more than half (58.4%) of cases and (51.5%) of controls were males. Their mean age was 27.5 and 32.2 months for cases and controls, respectively, and this difference was significant (p=0.04). The mean (standard deviation) temperature of cases and controls was 37.8 (0.8)°C and 36.4 (0.5)°C, respectively. Overall, the proportion of stunting was higher among cases (50.3%) than controls (40.2%) and the difference was significant (p=0.01). When stunting was further disaggregated by severity, both severe and moderate stunting was higher among cases than controls, though the difference was only significant among those who had severe stunting (p=0.02). Likewise, children who had malaria significantly suffered underweight (34.2%), more than those without malaria or the control group (32.3%). Furthermore, malaria was significantly higher among those suffering moderate underweight (p=0.01).

Identified Plasmodium spp.

Of the total 149 enrolled malaria cases, 79 (53.0%) were Plasmodium falciparum, 62 (41.6%) were P. vivax and the remaining 8 (5.4%) were mixed cases (P. falciparum and P. vivax; data not shown).

Biological assessments

Association of malaria with stunting and underweight

Statistical analysis

Association of malaria with selected sociodemographic variables

A finger prick following aseptic techniques was done on all subjects to prepare thick and thin films for confirmation of malaria parasites, and stained with Giemsa in the respective health centres. Each slide was read by senior laboratory technicians who had participated in the training. Absence of malaria parasite in 200 high-power ocular fields of the thick film was considered as negative.

Data collected in the health centres were checked for completeness, coded, and entered using EpiInfo version 3.5.4 (Centers for Disease Control and Prevention, USA), then exported to SPSS version 21 (IBM, 64

SAJCH

Overall, stunting was significantly associated with confirmed malaria. The odds of developing malaria were 1.6 times more among stunted than non-stunted children (AOR 1.61, 95% CI 1.19 - 2.51).

Likewise, the odds of developing malaria were 1.7 times more among underweight children than the referent group (AOR 1.69, 95% CI 1.11 - 2.90; Table 3).

Place of residence, sex of the child, long-lasting insecticide-treated net (LLIN) utilisation, and use of indoor residual spray (IRS) were

MARCH 2016 Vol. 10 No. 1

RESEARCH Table 1. Respondents' characteristics by study groups, BahirDar special zone, Ethiopia, 2014 Characteristics

Cases (N=149), Controls (N=458), n (%) n (%) p-value

Characteristics

Cases (N=149), Controls (N=458), n (%) n (%) p-value

Residence

Age (years) 10 - 19

3 (2.0)

39 (8.5)

20 - 29

43 (28.9)

200 (43.7)

0.05

30 - 39

75 (50.3)

163 (35.6)

0.90

40 - 49

25 (16.8)

52 (11.4)

0.01*

≥50

3 (2.0)

4 (0.9)

0.02*

Sex

Urban

87 (58.4)

392 (85.6)

Rural

62 (41.6)

66 (14.4)

Alive

130 (87.2)

429 (93.7)

Died

19 (12.8)

29 (6.3)

0.01*

Parent of the child 0.01*

*p<0.05.

Male

20 (13.4)

74 (16.2)

Female

129 (86.6)

384 (83.8)

Married

129 (86.6)

414 (90.4)

Single

1 (0.7)

4 (0.9)

0.42

Divorced

11 (7.4)

28 (6.1)

0.26

Widowed

1 (0.7)

12 (2.6)

0.08

Separated

7 (4.7)

0 (0.0)

-

Housewife

72 (48.3)

223 (48.7)

Farmer

43 (28.9)

161 (35.2)

0.19

Daily labourer

18 (12.1)

23 (5.0)

0.01*

Employer

8 (5.4)

45 (9.8)

0.07

Merchant

7 (4.7)

6 (1.3)

0.01*

Other

1 (0.7)

0 (0)

-

0.21

Marital status

Cannot read and 73 (49.0) write

170 (37.1)

Read and write

19 (12.8)

53 (11.6)

0.27

Primary (Grades 25 (16.8) 1 - 8)

56 (12.2)

0.44

Secondary (Grades 9 - 12)

20 (13.4)

107 (23.4)

0.01*

Grade 12 and above

12 (8.0)

72 (15.7)

0.01*

Household members ≤2

5 (3.4)

19 (4.1)

3-4

75 (50.3)

158 (34.5)

0.13

≥5

69 (46.3)

281 (61.4)

0.45

Mean (SD)

4.7 (1.80)

5.4 (2.15)

39 (26.2)

79 (17.2)

500 - 999

36 (24.2)

181 (39.5)

0.01*

1 000 - 1999

53 (35.6)

133 (29.0)

0.20 0.09

21 (14.1)

65 (14.2)

1 106.3 (820.1)

1 026.9 (778.5) Continued ...

65

Controls (N=458), n (%)

p-value

Male

87 (58.4)

236 (51.5)

0.14

Female

62 (41.6)

222 (48.5)

6 - 23

70 (47.0)

173 (37.8)

24 - 59

79 (53.0)

285 (62.2)

Age (months), mean (SD)

27.5 (12.9)

32.2 (16.3)

0.01*

Temperature (°C), mean (SD)

37.8 (0.8)

36.4 (0.7)

0.01*

Stunted (overall)

75 (50.3)

184 (40.2)

0.01*

Severe stunting

24 (16.1)

63 (13.8)

0.02*

Moderate stunting

51 (34.2)

121 (26.4)

0.10

Normal

74 (49.7)

274 (59.8)

-

Underweight (overall)

51 (34.2)

102(32.3)

0.01*

Severe underweight

14 (9.4)

31 (6.8)

0.07

Moderate underweight

37 (24.8)

71 (15.5)

0.01*

98 (65.8)

356 (77.7)

-

149 (100)

458 (100)

Sex

0.04*

Underweight

Normal Total *p<0.05

Monthly family income (EB)

≥2 000

Cases (N=149), n (%)

Characteristic

Stunting

Educational status

Mean (SD)

Table 2. Child characteristics by study groups, Bahir-Dar special zone, June 2014

Child age (months)

Occupation

<500

Table 1. (continued) Respondents' characteristics by study groups, Bahir-Dar special zone, Ethiopia, 2014

SAJCH

important variables significantly associated with occurrence of malaria, after controlling for confounding effects in the multivariate model.

Almost one-third (28.9%) of children in cases and two-thirds (66.2%) in controls slept the night preceding the interview under a LLIN. As a result, those children who slept under a LLIN were 68.6% less likely to have confirmed malaria than controls (AOR 0.314, 95% CI 0.15 - 0.36). Similarly, 42.3% of children in cases and 73.8% of controls were from a house sprayed with IRS in the last 12 months.

As a result, those children from houses sprayed with IRS were 66.3%

MARCH 2016 Vol. 10 No. 1

RESEARCH Table 3. Correlates of stunting and confirmed malaria in preschool children at Bahir-Dar special zone, 2014, after adjustment for other variables Variable

Cases (N=149), n (%) Controls (N=458), n (%) Crude OR (95% CI) AOR (95% CI)

Stunting Yes

75 (50.3)

184 (40.2)

1.509 (1.041 - 2.188)

No

74 (49.7)

274 (59.8)

1

Yes

51 (34.2)

102 (22.3)

1.816 (1.213 - 2.720)

No

98 (65.8)

356 (77.7)

1

1.614 (1.192 - 2.514*)

Underweight 1.690 (1.112 - 2.903*)

Residence Rural

62 (41.6)

66 (14.4)

4.233 (2.789 - 6.424)

1.546 (1.007 - 3.179*)

Urban

87 (58.4)

392 (85.6)

1

1

Sex Male

87 (58.4)

236 (51.5)

1.320 (0.908 - 1.918)

1.416 (0.931 - 1.980)

Female

62 (41.6)

222 (48.5)

1

1

6 - 23

70 (47.0)

173 (37.8)

0.685 (0.472 - 0.995)

1.128 (0.649 - 1.676)

24 - 59

79 (53.0)

285 (62.2)

1

1

Child age (months)

Slept under LLIN last night Yes

43 (28.9)

303 (66.2)

0.208 (0.139 - 0.311)

0.314 (0.148 - 0.360*)

No

106 (71.1)

155 (33.8)

1

1

Yes

63 (42.3)

338 (73.8)

0.260 (0.177 - 0.383)

0.337 (0.204 - 0.530*)

No

86 (57.7)

120 (26.2)

1

1

73 (49.0)

170 (37.1)

1

1

Indoor residual spray within last 12 months

Education status Cannot read and write Read and write

19 (12.8)

53 (11.6)

0.835 (0.462 -1.508)

0.811 (0.401 - 1.716)

Grades 1 - 12

45 (30.2)

163 (35.6)

0.643 (0.419 - 0.987)

0.778 (0.510 - 1.311)

Above Grade 12

12 (8.1)

72 (15.7)

0.388 (0.199 - 0.758)

0.618 (0.414 -1.286)

Occupation Housewife

72 (48.3)

223 (48.7)

1

1

Farmer

43 (28.9)

161 (35.2)

0.827 (0.539 - 1.270)

0.880 (0.487 - 1.884)

Merchant or other

26 (17.4)

29 (6.3)

2.777 (1.536 - 5.021)

1.621 (0.988 - 7.164)

Employer

8 (5.4)

45 (9.8)

0.551 (0.248 - 1.222)

0.994 (0.553 - 2.046)

*p<0.05

less likely (AOR 0.33, 95% CI 0.20 - 0.53) to have confirmed malaria than those who did not, after controlling for nutritional and other sociodemographic variables.

Males were 1.4 times more likely to have confirmed malaria than females, though the difference was not significant (AOR 1.41, 95% CI 0.93 - 1.98).

Children from rural areas were 1.5 times more likely to have confirmed malaria than those from urban areas after stunting, underweight and other important factors were controlled (AOR 1.5, 95% CI 1.01 - 3.18).

Discussion

This facility-based case-control study attempted to establish the association between chronic undernutrition and malaria, as well as other important sociodemographic variables of major public health significance in this and other developing countries, to enable the incorporation of programme initiatives in malaria prevention and control guidelines.

In this study, chronic undernutrition, in the form of stunting and underweight, was common, and undernourished children were at higher 66

SAJCH

risk of contracting malaria than well-nourished children. The odds of having confirmed malaria were 1.6 times higher among stunted children than non-stunted children and this finding was consistent with a longitudinal follow-up study conducted in Gambia and a cross-sectional study in Kenya, both of which reported that stunted children were at higher risk of developing malaria.[8,9] Furthermore, Fillol et al.,[10] through a cellular-level nested case-control study, substantiated the aforementioned findings by showing that antimalarial immune response is significantly lower among stunted children, suggesting that undernutrition could predispose to clinically overt malaria. Our study enrolled symptomatic malaria cases that could have been a result of lower antimalarial immune response, even though the study did not differentiate whether cases were new malaria infections or subsequent malarial attacks. In contrast to these findings, a Senegalese cohort study of young children from rural areas, a Ghanaian study and western Ethiopian studies have all reported no association between stunting and malaria infection.[11-13] The discrepancy could be explained by the difference in case definition used in both studies â&#x20AC;&#x201C; the studies in Ghana and Ethiopia compared asymptomatic malaria

MARCH 2016 Vol. 10 No. 1

RESEARCH and the Senegal study focused on patients with chronic malarial attack, differing entirely from the present study which used confirmed malaria cases. In addition, the Ghanaian study had sample size limitations, unlike the present study which had a large sample size and controlled for the confounding effects of important variables. The study in Ethiopia also found low malaria prevalence, making the analysis non-significant.[13] The comparative nature of our study alleviated the problem of low malaria prevalence, enabling comparison of the relationships with adequate sample size for cases and controls.

Similarly, the odds of having a confirmed malaria case were 1.7 times higher among underweight children than their counterparts and this finding is concordant with a WHO Comparative Risk Assessment project, which found that children who were moderately to severely underweight had an increased risk of a clinical malaria attack, compared with those who were better nourished, although this difference was not significant.[14] Nonsignificance could have been due to low prevalence of clinical malaria, which was described as a study limitation. A similar explanation was put forward by a northern Guinea study, which used a large sample size of 4â&#x20AC;&#x201A; 000 children.[15] The higher odds of confirmed malaria among underweight children are because of increased susceptibility to malaria for different reasons, such as through a reduction in the function of the immune system. A study conducted by Scrimshaw and SanGiovanni[16] indicated that undernourished children are unable to mount an appropriate immune response to the malaria parasite due to a reduction in T-lymphocytes, impairment of antibody formation, decreased complement formation, and atrophy of thymus and other lymphoid tissues. In contrast to the present findings, a study in Port Harcourt, Nigeria, showed that underweight children were not at a higher risk of asymptomatic malaria attack than wellnourished children (RR 1.02, 95% CI 0.34 - 2.37).[17] The discrepancy could be due to the difference in cases, as asymptomatic malaria cases were used, unlike our study, which enrolled confirmed symptomatic cases. Similarly, studies in Kenya and Uganda had reported no relationship between baseline nutritional status and subsequent incidence of malaria.[18,19] The discrepancy noted with the present study could be explained by malaria transmission in Uganda being holoendemic and perennial, while in the present study transmission is seasonal. The Kenyan study included children up to 8 years old,[18] another difference from our study.

Other important factors strongly associated with malaria attack were use of LLIN for children during sleeping, use of IRS in the last 12 months and place of residence.

Use of LLIN for children during sleeping was strongly associated with clinical malaria after controlling for undernutrition and other demographic variables. Those children who slept under LLIN were 68.6% less likely to have confirmed malaria than children who did not. This is concordant with a community-based comparative interventional study conducted in Nigeria, which showed a 93% decrease in malaria prevalence for those who slept under LLIN compared with those who did not.[20] Children from houses sprayed with IRS in the last 12 months were 66.3% less likely to have confirmed malaria than their counterparts, after controlling for undernutrition and other confounders.

In this study, children from rural areas were 2.3 times more likely to have confirmed malaria than children from urban areas. This could be explained by higher irrigation and drainage in rural than urban areas, which is favourable for mosquito breeding and malaria transmission.

The strengths of the study are that it compared cases and controls using an appropriate sample size, included anthropometric measurements, and used the laboratory for confirmation of malaria diagnosis, to decrease the subjectivity of diagnosis/bias. Micronutrient status, anaemia and the level of parasitaemia, which might have an impact, were not assessed and are considered limitations of the study.

Conclusion

Chronic undernutrition was closely associated with malaria infection.

The major predictors for contracting malaria were being stunted, 67

SAJCH

underweight, rural residence, not using LLIN or IRS, male sex, and low educational status of the mother/caregiver. It is recommended that health offices and non-government organisations operating in malaria endemic areas distribute LLINs for households and create awareness on the importance of maintaining good nutrition, as well as implementing other preventive measures addressing the previously mentioned predictors.

Integration of malaria intervention programmes with essential nutrition actions would be an effective strategy to reduce morbidity and mortality from malaria and undernutrition. Further studies with more biochemical tests might be helpful in addressing some of the study limitations.

Acknowledgement. The Addis Ababa School of Public Health Research Ethics Review Committee, and the staff from Bahir-Dar health offices, the respective health centres and the municipality, are duly acknowledged.

This work would not have reached its present state without the unreserved participation of the respondents and the support rendered by our families.

References 1. World Health Organization (WHO). Global Malaria Programme: World Malaria Report 2012. www.who.int/malaria (accessed 10 December 2013).

2. Food and Agricultural Organization of the United Nations (FAO). The State of Food Insecurity in the World, 2012. http://www.fao.org/docrep/016/i3027e/ i3027e00.htm (accessed 13 February 2014).

3. United Nations International Childrenâ&#x20AC;&#x2122;s Fund (UNICEF). Improving child nutrition: The achievable imperatives for global progress. New York: UNICEF, 2013.

4. Ethiopian Health and Nutrition Research Institute (EHNRI). Ethiopian malaria indicator survey report 2011. Addis Ababa, Ethiopia: EHNRI, 2011.

5. Central Statistical Agency (Ethiopia) and ICF International. 2012. Ethiopia Demographic and Health Survey 2011. Addis Ababa, Ethiopia and Calverton, USA: Central Statistical Agency and ICF International, 2011:156-157.

6. Cunningham RS, McNeeley DF, Moon A. Mechanisms of nutrient modulation of the immune response. J Allergy Clin Immunol 2005;115(1129):1119-1128. [http://dx.doi.org/10.1016/j.jaci.2005.04.036] 7. Roth DE, Caulfield LE, Ezzati M, Black RE. Acute lower respiratory infections in childhood: Opportunities for reducing the global burden through nutritional interventions. Bull World Health Organ 2008;86(4):356-364. [http://dx.doi. org/10.2471/blt.07.049114] 8. Deen JL, Walraven GE, von Seidlein L. Increased risk for malaria in chronically malnourished children under 5 years of age in rural Gambia. J Trop Pediatr 2002;48(2):78-83. [http://dx.doi.org/10.1093/tropej/48.2.78] 9. Friedman JF, Kwena AM, Mirel LB, et al. Malaria and nutritional status among pre-school children: Results from cross-sectional surveys in Western Kenya.

Am J Trop Med Hyg 2005;73(4):698-704.

10. Fillol F, Sarr JB, Boulanger D, et al. Impact of child malnutrition on the specific anti-Plasmodium falciparum antibody response. Malar J 2009;8(2):116. [http:// dx.doi.org/10.1186/1475-2875-8-116] 11. Fillol F, Cournil A, Boulanger D, et al. Influence of wasting and stunting at the onset of the rainy season on subsequent malaria morbidity among rural preschool children in Senegal. Am J Trop Med Hyg 2009;80(2):202-208.

12. Crookston BT, Alder SC, Boakye I, et al. Exploring the relationship between chronic undernutrition and asymptomatic malaria in Ghanaian children. Malar J 2010;9(2):39. [http://dx.doi.org/10.1186/1475-2875-9-39] 13. Deribew A, Alemseged F, Tessema F, et al. Malaria and undernutrition: A community based study among under-five children at risk of malaria, south-west Ethiopia. PLoS One 2010;5(5):e10775. [http://dx.doi.org/10.1371/journal.pone.0010775] 14. Caulfield LE, Richard SA, Black RE. Under-nutrition as an underlying cause of malaria morbidity and mortality in children less than five years old. Am J Trop Med Hyg 2004;71(2 Suppl):55-63.

15. Snow RW, Byass P, Shenton FC, Greenwood BM. The relationship between anthropometric measurements and measurements of iron status and susceptibility to malaria in Gambian children. Trans R Soc Trop Med Hyg 1991;85(5):584-589.

16. Scrimshaw NS, SanGiovanni JP. Synergism of nutrition, infection, and immunity: An overview. Am J Clin Nutr 1997;66(2):464S-477S. 17. Jeremiah ZA, Uko EK. Childhood asymptomatic malaria and nutritional status among Port Harcourt children. East Afr J Public Health 2007;4(2):55-58.

18. Nyakeriga AM, Troye-Blomberg M, Chemtai AK, Marsh K, Williams TM. Malaria and nutritional status in children living on the coast of Kenya.

Am J Clin Nutr 2005;80(23):1604-1610. [http://dx.doi.org/10.1111/j.03009475.2004.01423o.x] 19. Arinaitwe E, Gasasira A, Verret W, et al. The association between malnutrition and the incidence of malaria among young HIV-infected and -uninfected Ugandan children: A prospective study. Malar J 2012;27(11):90. [http://dx.doi. org/10.1186/1475-2875-11-90] 20. Ashikeni MA, Envuladu EA, Zoakah AI. Malaria and the use of the insecticidetreated net (ITN) among under-five children in Kuje Area Council of the Federal Capital Territory Abuja, Nigeria. J Mosq Res 2013;3(6):45-53. [http:// dx.doi.org/10.5376/jmr.2013.03.0006].

MARCH 2016 Vol. 10 No. 1

RESEARCH

Epidemiology of paediatric poisoning reporting to a tertiary hospital in Ghana D Ansong,1,2,4 MD, MSc; C Nkyi,3 MD; C O Appiah,2 BSc; E X Amuzu,2 BSc; C A Frimpong,2 BSc; I Nyanor,2 MPH; S B Nguah,4 MD; J Sylverken,4 MD Department of Child Health, School of Medical Sciences, Kwame Nkrumah University of Science and Technology, Kumasi, Ghana Research and Development Unit, Komfo Anokye Teaching Hospital, Kumasi, Ghana 3 Department of Internal Medicine, Komfo Anokye Teaching Hospital, Kumasi, Ghana 4 Department of Child Health, Komfo Anokye Teaching Hospital, Kumasi, Ghana 1 2

Corresponding author: D Ansong (ansongd@yahoo.com) Background. Childhood poisoning is an important cause of morbidity in both developed and developing countries. Epidemiological studies on accidental poisoning in children show a consistent pattern regarding age and gender. Childhood poisoning is predominant in children <6 years of age and has a male preponderance, as boys are more active with a drive to explore the environment.

Objective. To document the epidemiology of home poisonings in Kumasi and its environs.

Methods. We conducted a retrospective study from January 2007 to January 2012 at the Komfo Anokye Teaching Hospital, a tertiary hospital in Ghana.

Results. Poisoning is a significant health problem in the study area. A total of 253 children reported to the hospital with poisoning over the 61-month period, with an average of four cases per month. The male to female ratio was 1.58:1. The median age of the children was 24 months (interquartile range 24 - 48 months). Kerosene was the leading cause of poisoning (39.5%).

Conclusion. Paediatric poisoning is a major health hazard in children living in Kumasi and its environs. This can possibly be attributed to a lack of adequate supervision of children and poor storage of harmful substances in homes. Multidisciplinary interventions are needed to reduce the occurrence of the condition in the population at risk.

S Afr J Child Health 2016;10(1):68-70. DOI:10.7196/SAJCH.2016.v10i1.1055

In 2004, an estimated 346 000 people died worldwide from unintentional poisoning, according to the World Health Organiza­ tion; 91% of the deaths occurred in low- and middle-income countries.[1] Poisoning is defined as exposure of an individual to a substance that can cause symptoms and signs of organ dysfunction leading to injury or death.[2] Ingestion of poisons and other household products has the potential to increase morbidity and mortality in children worldwide.[3,4] The ingestion can be accidental or non-accidental in younger children, but in older children – especially in high-income countries – it is mostly intentional.[5] As children explore their environment as part of their natural development, they are exposed to poisons without knowing that they may be harmful. Medicinal and non-medicinal substances have been identified as common agents causing poisoning in children as well as adolescents.[6] Common medicinal substances ingested by children are analgesics, anti-inflammatory agents, psychotropic drugs such as antidepressants and benzodiazepines, and related agents.[7] Non-medicinal chemicals such as organophosphates, pesticides, insecticides, organic solvents and household agents such as bleach and caustic soda are also common causes of poisoning among children.[8] Several studies have reported on childhood poisoning in both developed and developing countries.[4,6,7] A pre­ viously reported comparative analysis carried out from January to June 2009 and January to June 2010 in the current study hospital revealed an increase in chemical poisoning. A six-fold increase in the incidence of caustic soda poisoning occurred during the period. [9] Our objective was to study the records of paediatric patients presenting with poisoning between January 2007 and January 2012, in order to describe the demographic characteristics and identify the types of childhood poisonings in this paediatric population.

68

SAJCH

Methods Setting

A retrospective study was conducted at Komfo Anokye Teaching Hospital in Kumasi, Ghana. The hospital is a tertiary hospital situated in the second largest city, with a population base of about two million.

The hospital has 1 000 beds with an average occupancy rate of about 150%. The Paediatric Emergency Unit runs a 24-hour service, and it serves as the major referral centre for all emergencies in children aged 3 months to 16 years.

Design

A retrospective study design was used to collect all reported cases of poisoning in the Paediatric Emergency Unit between January 2007 and January 2012 (61 months). This period was chosen based on the reliability and availability of medical records for the study.

The study defined poison as any medicinal or non-medicinal substance, including food and food products, which was ingested inappropriately. A systematic review of all medical records of home poisoning reported to the emergency unit was conducted. Data on age, gender and type of poisons were extracted from records.

Data collected were entered on the Epi-Info 3.2.1 database system (Centers for Disease Control, USA) and exported into STATA 8.1 software (StataCorp LP, USA) for analysis. Basic summary statistics of age and gender were conducted, and bivariate analysis of type of poisoning and age was also conducted.

Permission to conduct the study was obtained from the Committee on Human Research Publications and Ethics (CHRPE) and the Research and Development Unit of Komfo Anokye Teaching Hospital.

Results

A total of 253 cases of home poisoning had reported in person to the emergency unit of the hospital. (The data gathered did not include telephone enquiries but only cases that had reported to the emergency

MARCH 2016 Vol. 10 No. 1

RESEARCH

15 Frequency, n

10

5

Oct 2011

Jan 2012

Jul 2011

Apr 2011

Jan 2011

Oct 2010

Jul 2010

Apr 2010

Jan 2010

Time (month, year)

Fig. 1. Trend in accidental poisoning occurring in children reporting to a tertiary hospital in Ghana.

Table 1. Distribution of patients by sex and age groupings

Discussion

The study found that home poisonings are a cause of morbidity in children living in Kumasi and its environs. The most common causes of poisoning found in the study were kerosene, sodium hypochlorite and caustic soda. In communities in Ghana these chem­ icals are readily available for domestic activities such as cooking, cleaning and local soap preparation. The chemicals are stored in plastic water bottles that resemble those used for drinking water. Storage of these substances serves as a potential source of danger if they are within easy reach of the children in their homes. Kerosene poisoning was noted to be the leading cause of poisoning in the study (39.5%). This result is similar to that in other studies that showed kerosene to be the leading source of poisoning in children. [10,11] The study also showed that children <3 years of age were more commonly involved in childhood poison­ing (74.3%). Studies by Abbas et al.[12] showed simi­lar findings, with children <3 years hav­ ing the highest incidence of poisoning (46.5%) in their study sample. This may be attributable to lack of adequate supervision for children in this age group, who like to explore their environment. From January to July 2007 there was a spike in the number of cases of poisoning reported at the Paediatric Emergency Unit of Komfo Anokye Teaching Hospital. This spike may be attributable to the prolonged period of low electrical power generation in the country, with a resultant increase in the use of petroleum products as alternatives for sources of light.[13] There was no marked difference in terms of age for caustic soda ingestion. Caustic soda poisoning is associated with life-threatening complications such as corrosive oesophagitis and oesophageal stricture, and its proportion in this study (7.5%) is worrying. The use of

Oct 2009

Jul 2009

Apr 2009

Jan 2009

Oct 2008

Jul 2008

Apr 2008

Jan 2008

Oct 2007

Jul 2007

Jan 2007

0 Apr 2007

department for treatment.) On average, about four cases of home poisoning were reported to the Paediatric Emergency Unit of the Komfo Anokye Teaching Hospital every month over the study period of 61 months (Fig. 1).

The median age of children affected by poisoning was 24 months (interquartile range 24 - 48 months). There were more males (n=155, 61.3%) than females (n=98, 38.7%) reported with poisoning, giving a male to female ratio of 1.58: 1 (Table 1).

Over 60% of the affected children were <3 years old. Out of 253 children identified, 25 (9.9%) suffered from sodium hypochlorite ingestion, 19 (7.5%) patients had ingested caustic soda, 16 (6.3%) had food poisoning, 11 (4.4%) had iron poisoning, and 4 (1.6%) had ingested rat poison. The most frequently recorded poison was kerosene (100 subjects, 39.5%), followed by other chemical ingestion in 70 patients (27.7%) (Table 2).

Sex, n (%) Age in months

Female

Male

Total, n (%)

<12

21 (21.43)

29 (18.71)

50 (19.76)

12 - 24

26 (26.53)

62 (40.00)

88 (34.78)

25 - 36

24 (24.49)

26 (16.77)

50 (19.76)

>36

27 (27.55)

38 (24.52)

65 (25.69)

Total

98 (100)

155 (100)

253 (100)

Median

24

Table 2. Source of poisoning by age groupings Age category (months), n (%) Diagnosis

<12

12 - 24

25 - 36

>36

Total, n (%)

Caustic soda ingestion

4 (21.5)

5 (26.3)

5 (26.3)

5 (26.3)

19 (7.5)

Drug ingestion

1 (12.5)

1 (12.5)

4 (50.0)

2 (25.0)

8 (3.2)

Food poisoning

2 (12.5)

2 (12.5)

1 (6.3)

11 (68.8)

16 (6.3)

Iron poisoning

5 (45.5)

3 (27.3)

1 (9.1)

2 (18.2)

11 (4.4)

Kerosene ingestion

15 (15.0)

40 (40.0)

20 (20.0)

25 (25.0)

100 (39.5)

Sodium hypochlorite (parazone) ingestion

7 (28.0)

11 (44.0)

2 (8.0)

5 (20.0)

25 (9.9)

Rat poisoning

1 (25.0)

1 (25.0)

0 (0.0)

2 (50.0)

4 (1.6)

Other chemical ingestions*

15 (21.4)

25 (35.7)

17 (24.3)

13 (18.6)

70 (27.7)

Total

50 (19.8)

88 (34.8)

50 (19.8)

65 (25.7)

253 (100)

*Other chemical ingestions refer to any other poisons including weedicides, pesticides, etc. that appeared only once in the data.

this agent in households for making local soap and the poor methods of storage are contributing to the availability of this poison to children. Detailed studies need to be done on poisoning in children of Ghanaian homes to influence public health policies and interventions.

Food poisoning was noted to have a high incidence in children >36 months (68.8%).

69

SAJCH

MARCH 2016 Vol. 10 No. 1

This is an expected finding as children in this age group are starting to eat independently and have more freedom to explore the variety of food options with less supervision. Iron poisoning had a significant representation (4.4%). This may be attributable to the attractive packaging and poor storage of iron-containing tablets, making them readily available and appealing to children.

RESEARCH Other drugs and rat poisoning had 3.2% and 1.6% representation, respectively. This is a large proportion and may be due to poor storage and poor compliance with safety measures in homes. The recent liberalisation of policies governing the sale of agrochemicals and animal poisons in Ghana may have contributed to this problem.

The liberalisation has also raised safety concerns regarding the storage and use of these poisons within homes.

This retrospective study provided limited data but more details will emerge from our current longitudinal study aimed at monitoring all children reporting to the study hospital with poisoning. The additional information will support the evidence from the retrospective study and highlight the need to provide a comprehensive public health approach to this health problem.

Conclusion

Childhood poisoning is a major public health problem in developing country such as Ghana. Data gathered from research aimed at understanding the underlying causes may be used to formulate public health policies including comprehensive health education of the public. Children will continue to be exposed to poisonings until the lack of education of the caregivers is addressed.

References 1. World Health Organization. Poisoning Prevention and Management. http:// www.who.int/ipcs/poisons/en/ (accessed 5 January 2013).

70

SAJCH

2. Osterhoudt K, Shannon M, Henretig F. Toxicological emergences. In: Fleisher G, Ludwig S, eds. Textbook of Pediatric Emergency Medicine. 4th ed.

Philadelphia: Lippincott William & Wilkins, 2000:887-897.

3. Walton WW. An evaluation of the Poison Prevention Packaging Act. Pediatrics 1982;69(3):363-370.

4. Lawson GR, Craft AW, Jackson RH. Changing pattern of poisoning in children in Newcastle, 1974 - 1981. Br Med J Clin Res Ed 1983;287(6384):15-17.

5. Sibert J, Davies P. Poisoning, accidents and sudden infant death syndrome. In: Campbell A, McIntosh M, eds. Forfar and Arneil’s Textbook of Paediatrics. 4th ed. London: Churchill Livingstone, 1992:1777-1800.

6. Aslam M, Boluch GR, Wagar H, Akbar M, Aniga H. Accidental poisoning in children. Pak Paediatr J 2002;26(2):67-72.

7.

James LP, James L, Abel K, Wilkinson J, Simpson PM, Nichols MH. Phenothiazine, butyrophenone, and other psychotropic medication poisonings in children and adolescents. Clin Toxicol 2000;38(6):615-623. [http://dx.doi. org/10.1081/clt-100102010] 8. Lam L. Childhood and adolescence poisoning in NSW, Australia: An analysis of age, sex, geographic, and poison types. Inj Prev 2003;9(4):338-342. [http:// dx.doi.org/10.1136/ip.9.4.338] 9. Weldon E, Martey PM. Caustic soda poisoning in Ghana – an alarming increase. Paediatr Int Child Health 2012;32(3):158-160. [http://dx.doi.org/10.

1179/2046905512y.0000000007] 10. Balme K, Roberts JC, Glasstone M, Curling L, Mann MD. The changing trends of childhood poisoning at a tertiary children’s hospital in South Africa. S Afr Med J 2012;102(3):142-146.

11. Brata Ghosh V, Jhamb U, Singhal R, Krishnan R. Common childhood poisonings and their outcome in a tertiary care center in Delhi. Indian J Pediatr 2013;80(6):516-518. [http://dx.doi.org/10.1007/s12098-012-0879-5] 12. Abbas SK, Tikmani SS, Siddiqui NT. Accidental poisoning in children. J Pak Med Assoc 2012;62(4):331-334.

13. Boateng MOA. New load shedding programme out. http://edition.myjoyonline. com/pages/news/200703/2836.php (accessed 12 February 2016).

MARCH 2016 Vol. 10 No. 1

RESEARCH

An exploratory study of the implementation of early intervention workshops for primary caregivers in Johannesburg S Medhurst,1 BA (Speech & Hearing Therapy); S Abdoola,2 BComm Pathology (SLT & A), MECI; L Duncan,3 BSocSci (Psych), BEd Hon (School Guidance and Counselling), MECI Speech-Language Therapist and Audiologist, Johannesburg, South Africa Department of Speech Pathology and Audiology, Faculty of Humanities, University of the Witwatersrand, Johannesburg, South Africa; and Department of Speech Language Pathology and Audiology, Faculty of Humanities, University of Pretoria, South Africa 3 Remedial Therapist, Reddford House, Pretoria, South Africa 1 2

Corresponding author: S Abdoola (786.shabnam@gmail.com) Background. If primary caregivers are able to stimulate their children’s development effectively, then the prevalence of children at risk of cognitive and language developmental delays could decrease and the shortage of available services for the identified children could be addressed, as hopefully fewer children would require extensive early intervention (EI) services later on in life.

Objective. To develop and implement an EI workshop with primary caregivers on how to provide language and cognitive stimulation through daily living activities (DLAs).

Methods. Two workshops were conducted at two daycare centres, focusing on daily language stimulation, with the primary caregivers of children aged between 0 and 3 years. A pre-workshop, semi-structured group interview was conducted to gain insight into the participants’ knowledge and expectations. This was followed with a post-workshop, semi-structured group interview to gain insight into and feedback on how the participants were able to carry over the techniques in order to stimulate their children in DLAs. The data were analysed using thematic data analysis.

Results. It was found that although participants demonstrated the basic understanding of the concept of EI, their knowledge improved with the workshop. The participants reported that they were able to implement the techniques gained from the workshop and noticed a change in their children’s behaviour and communication within the space of 1 week. The participants also reported on the ease of stimulating their children through DLAs and that no additional time had to be scheduled for stimulation.

Conclusion. The workshops have the potential to target populations regardless of their socioeconomic status, cultural beliefs, linguistic differences, and access to medical institutions.

S Afr J Child Health 2016;10(1):71-74. DOI:10.7106/SAJCH.2016.v10i1.1057

The literature indicates a direct correlation between poor academic performance (failure to either complete primary or secondary school, or failure to obtain the requirements to complete each grade of schooling) in primary caregivers (main parent or family member who takes care of and spends most time with the child) and poor cognitive and language development in the children of these primary caregivers. Children from low-income households generally experience less consistent caregiving, support and cognitively stimulating home environments, alongside external environment influences.[1] It is therefore essential to enrich these children’s home environments in order to support cognitive development[2] and, consequently, language development.

Owing to the fact that there is a lack of speech-language pathologists in the lower socioeconomic status areas in South Africa (SA), it becomes necessary to provide informative workshops to these families.

The objective of the workshops provided in the current study was to provide techniques to primary caregivers on how to stimulate cognitive and language development of their children in daily living activities (DLAs). The use of these techniques by the primary caregivers might decrease the need for future intervention for these children at risk from cognitive and language developmental delays.

Early intervention (EI)

EI refers to comprehensive activities designed to improve and develop the cognitive, language, motor and sensory development of young children.[3,4] Three major objectives of EI are to identify at-risk children, to provide EI as a preventive measure, and to identify and intervene with 71

SAJCH

young children that have existing developmental delays.[3] EI services can be implemented regardless of the socioeconomic status of the participants.[5] The evidence supports and indicates that EI has positive effects on development if implemented correctly by the therapists in partnership with the parents or caregivers.[5] However, there appears to be insufficient literature discussing how primary caregivers can be empowered to stimulate their children optimally in order to prevent developmental delays and to address the lack of intervention services in the public health services in SA. Family interaction influences the cognitive and language development of young children.[6] Therefore, it is crucial to involve primary caregivers in EI services and consequently empower them.

Caregivers and EI services

Caregivers are the most critical providers of care, support, holistic development and stimulation, and therefore play a vital role in EI.[7] Furthermore, open and honest communication between all parties is essential for the implementation of EI. It is also vital that caregivers understand the different stages of development and how to stimulate their child’s development. An audit conducted by the SA Department of Education in 2000 revealed that 84% of young children between the ages of 0 and 5 years of age did not have access to formal early childhood development (ECD) services and relied on their primary caregivers for EI.[8] Children from low-income socioeconomic environ­ ments face external factors that affect their development,[1] which further stresses the importance for primary caregivers to have access to EI services and implement the services.

MARCH 2016 Vol. 10 No. 1

RESEARCH The purpose of this qualitative study was to develop and implement an EI workshop for primary caregivers on how to provide language and cognitive stimulation. The caregivers were required to implement the shown skills at home in DLAs and provide feedback as to whether or not it was challenging to stimulate their child’s language and cognition, and if any changes in their child’s behaviour, cognition or language were noted within a week. As previously mentioned, there is a lack of speech-language pathologists in SA and providing workshops in a group setting is more effective than locating individual caregivers and providing information.

Methods

Participants

A combination of convenience and snowball sampling was used in selecting the participants from two daycare centres in Johannesburg.

The daycare centres had to each be registered as a child development centre. The daycare centres were contacted and asked if any EI services were being offered to parents and if not, would they be interested in receiving EI services? A needs analysis was conducted at the two selected daycare centres to further specify the needs in terms of EI services. The participants had to be caregivers of children between the ages of 0 and 3 years; this is the age group stipulated as the target group for EI in SA.[4] The sample size comprised 15 participants, combined of primary caregivers and staff members from the daycare centres.

Data collection

Data were collected through semi-structured group interviews with the participants. Semi-structured group interviews are more flexible than standardised questionnaires but are not unstructured discussions, and they encourage participants to respond to openended questions; this in turn reveals the participants’ knowledge, opinions and concerns regarding the topics being discussed.[9] Semi-structured group interviews are seen to have high validity because the participants are able to discuss items in great detail and depth. [10] There were two group interviews: a pre-workshop interview to determine the primary caregivers’ general understanding and thoughts around EI and their expectations of the workshops, and a post-workshop interview to receive the primary caregivers’ feedback from implementing and carrying over the techniques shown into the DLAs at home. The pre-workshop interview was conducted on the same day as the workshop. The post-workshop interview was conducted 1 week later in order to allow the participants to implement the knowledge and skills acquired from the workshop.

The time period between the data collection from Daycare Centre 1 and Daycare Centre 2 was ~2 weeks. The information obtained from the interviews was recorded with an audio recorder.

Workshops

The workshops were conducted and facilitated by the same student speech-language therapist. Although there are few standardised EI programmes available, the workshop was based on the principles of Guralnick’s Developmental Systems Approach (DSA). The workshop provided information pertaining to EI, and importance of EI services, age-appropriate communication and cognitive developmental mile­ stones, and general motor developmental milestones. It is essen­tial for the primary caregivers to have a general understanding of the holistic development of their children, and when and where to seek help if their children are not developing at age-appropriate levels.

The primary objective of the workshop was to look at how primary caregivers can stimulate children through DLAs, including feeding, bathing and dressing, in order to foster the carryover of skills.[11]

Reliability and validity

Reliability was ensured by conducting the same methodology at the two different daycare centres, and utilising a field journal or 72

SAJCH

researcher’s ‘reflective commentary’ to limit researcher bias and influence of the researcher’s emotions when analysing the data.

Validity was ensured through triangulation, specifically of data sources, to verify the viewpoints and experiences of the participants against each other’s.[12] A variety of participants (primary caregivers of children between the ages of 0 and 3 years, and daycare staff members) were included in the sample population.

Ethical considerations

Ethical clearance for this study was obtained from the Human Research Ethics Committee (Non-Medical) of the University of the Witwatersrand (protocol number: H14/02/04).

Confidentiality and anonymity could not be guaranteed because of the workshops and semi-structured group interviews, which were discussed with the participants. However, personal details and names of participants have been omitted in the presentation of findings to ensure confidentiality and anonymity.

Results

Workshops

There were 12 participants from Daycare Centre 1. Of these, three were male and nine were female. Nine participants were primary caregivers and three were staff members from that daycare centre.

There were 10 participants from Daycare Centre 2. Of these, one was male and nine were female. Six participants were primary caregivers and four were staff members (one from that daycare centre and three from neighbouring centres). Both Daycare Centres 1 and 2 had one child receiving EI services such as physiotherapy or speech and language therapy. No other EI services were being provided at either Daycare Centre 1 or 2.

Thematic content analysis revealed five main themes.

Awareness of EI

The participants from both Daycare Centres 1 and 2 had limited awareness and understanding of the term EI. One participant stated: ‘I’ve heard of EI on programmes, children’s programmes, but I’m not clear of its meaning’. Another participant demonstrated a basic understanding of the term: ‘You intervene if you see there’s a problem or before the problem starts,’ but could not identify the age group at which EI is targeted. The primary caregivers demonstrated an understanding of how their child’s current development affects school readiness: ‘ECD is very much important. If you can go wrong there, then you mess up big time. I think most of the parents understand, but I think it’s also ignorance,’ but there were a few participants who stated that their child is not talking now but they will wait until they are around 4 - 5 years old before asking for help.

Access to information

The three main areas from which participants receive information are clinics, daycare centres and television/technology. The majority of the participants from Daycare Centres 1 and 2 stated that they received information pertaining to their child’s development from the nurses or doctors at the clinics, especially when the child goes for immunisation. The daycare centres are also responsible for reporting to caregivers about their child’s development. Information is readily available through the internet, and participants admitted to using the internet to gather information about their own children and about conditions or syndromes that other children present with at the daycare centre or that they have heard of.

Current beliefs

When the topic of cultural belief was addressed during the preworkshop interviews, all the participants agreed on seeking medical intervention instead of consulting traditional healers. When probed further, some of the participants stated that religion is a major

MARCH 2016 Vol. 10 No. 1

RESEARCH influence in the change in belief, as many community members are becoming Christian/‘born-again Christian’, which influences the way they view medical intervention versus traditional intervention. It was also mentioned that education and access to information through technology also influences the participants’ beliefs.

Obtaining information

All the participants stated that they attended the workshops to receive information about development and how to stimulate their child’s development. Most participants from Daycare Centre 1 wanted information with regard to tantrums, and participants from both Daycare Centres 1 and 2 expressed specific concerns about their children’s communication skills or behaviour.

Daycare centre feedback

Daycare Centre 1 The participants showed a slight increase in understanding of the term EI: ‘I think I can say it’s how the child develops and how to interact with the child’; ‘…you can also identify a phase where the child needs help and work on that.’ The participants provided examples of how they stimulated their children during that week and the outcomes of the stimulation: ‘You know I even tried to make her to sit down and eat by herself at least. It worked, and I really can see that she knows what she is doing and she is listening. She was well behaved this week, but I can see that before workshop she was just toying with me’; ‘Just to speak to my child, what do you want to wear, this colour or this colour? I know what colours he likes now’; ‘To deal with the tantrums – before he would cry for nothing and now I can control them. He is better and I can give him options to choose what he wants and there is a change now’; ‘I think I was able to involve him in preparing the food, he likes that and that changed his eating a lot, because he was very fussy.’ The participants also stated that they spoke to other parents at the daycare centre about the workshops, and shared information that they learnt from the workshop.

Daycare Centre 2 Although the post-workshop could not be conducted, the email received about feedback stated: ‘I really gained a lot from the workshop and everyone else that was in the workshop. I think if it wasn’t for the other commitments that they had, they would’ve attended again, because I received good feedback from them. They told me that they’ve been practising the things that you told us to practise, like talking to your child like you are talking to an adult and that every time that you get to spend time with your child make it a learning opportunity.’

Discussion

The awareness of EI is important, because if caregivers are not aware that these services are available, then preventive and curative management for at-risk children or children who require the services cannot be and are not being implemented.[3] Caregivers’ knowledge and thoughts regarding EI are affected by the means in which they gather information about child development. Cultural beliefs, especially towards medical intervention, influence whom caregivers seek information from pertaining to a child’s development, as well as whether or not they receive or continue with EI services.

Grandparents and parents might also have diverse beliefs regarding interventions, which can be obstacles when young children are in need of EI. Although only two workshops were conducted, the feedback from the workshops provided insight into the potential and the importance of holding such workshops. There was good feedback from all participants, as they were able to implement the strategies given to them into their DLAs, and they noted changes in their children’s behaviour and became more aware of their children’s 73

SAJCH

abilities over the course of 1 week. It would have been more beneficial to monitor this over a longer period of time, as more insight into the effectiveness of the workshops could have been gained. It is vital that primary caregivers are aware of how to stimulate a child’s cognitive and language skills in order to foster age-appropriate development.

If a child is not developing age appropriately, the primary caregivers need to be able to identify this, be aware that EI services are available and know how such services can assist with the improvement of the child’s development. Primary caregivers also need to be made aware of how a child’s current development will affect school readiness and academic performance. The workshop covered all these areas, provided information on how to stimulate the children’s language and cognition functionally through DLAs, and covered relevant areas that the participants could relate to. There was a slight increase in understanding of the concept and benefits of the term ‘EI’ after the workshop.

Participants stated that the places from which they access inform­ ation regarding children’s development include clinics, daycare centres and television or internet. Various allied professionals could provide information at clinics through pamphlets and posters, and possibly implement workshops or support groups at the clinics, but this would have to be discussed with therapists at the clinics and the Department of Health. This study indicated that direct preventive and curative intervention can occur at the daycare centres. Cultural beliefs did not seem to have an effect on medical intervention; change in beliefs seems to be due to religion and primary caregivers having more access to information. This finding was common among participants at both daycare centres.

This study focused on the primary caregivers using the techniques shown to them through the implementation of EI workshops to provide language and cognitive stimulation through DLAs. However, the involvement of the daycare centres became apparent, including the importance of their roles within the community and in terms of sharing the information gained during the workshops. The daycare centres are now responsible for facilitating the transfer of the workshop information to primary caregivers who were unable to participate in the study, and the participants can share knowledge with other caregivers who did not participate in the study. Although it was not an objective of the study, Daycare Centre 1 held a workshop for the other caregivers who were unable to participate in the original workshop. This workshop was based on the workshop presented during the study. This provides hope for the extension of workshops should these be implemented in more daycare centres in the future.

Study limitations

It was difficult to identify and compare themes as the post-workshop interview could not be conducted at Daycare Centre 2. However, the owner of Daycare Centre 2 emailed feedback. If this study were to be conducted in future, more daycare centres should be involved and the study should be conducted over a longer time period to better measure the effectiveness of the workshops.

Recommendations

In order to measure the effectiveness of the workshops, it would be essential to implement the workshops at more daycare centres, and conduct sufficient follow-up interviews. The workshops can also be implemented in areas with different socioeconomic statuses and with varying access to health facilities where EI services are available, to determine how the workshops can target different populations.

Conclusion

Although only two workshops and one post-workshop interview were conducted, this study provided valuable insight into the importance of providing EI workshops to primary caregivers. The study also indicated that primary caregivers are able to stimulate their children’s

MARCH 2016 Vol. 10 No. 1

RESEARCH communication and cognition more easily through DLAs than scheduling additional time in the evening or on weekends to provide the stimulation. There is potential for conducting workshops that can incorporate different health and allied professionals and which can target populations regardless of their socioeconomic status, cultural beliefs, linguistic differences, and access to medical facilities.

References 1. Votruba-Drzal E, Coley RL, Chase-Lindsdale PL. Child care and low-income children’s development: Direct and moderated effects. Child Dev 2004;75(1):296312. [http://dx.doi.org/10.1111/j.1467-8624.2004.00670.x] 2. Woolfolk A. Educational Psychology (10th ed.). Boston, USA: Pearson Education, Inc., 2007.

3. Shonkoff JP, Meisels SJ. Handbook of Early Childhood Intervention (2nd ed.).

Cambridge: Cambridge University Press, 2000.

4. Department of Social Development. Guidelines for Early Childhood Development Services. Pretoria, South Africa: United Nations Children’s Fund, 2006.

74

SAJCH

5. Guralnick MJ. Effectiveness of early intervention for vulnerable children: A developmental perspective. Am J Ment Retard 1998;102(4):319-345.

6. Guralnick MJ. Family influences in early development: Integrating the science of normative development, risk and disability and intervention. In: McCartney K, Phillips D (eds). Handbook of Early Childhood Development. Oxford; Blackwell Publishers, 2006:44-61.

7. UNICEF. Guidelines for Early Childhood Development Services. Pretoria, South Africa: Department of Social Development, 2006.

8. UNICEF. Young Lives: Statistical Data on the Status of Children Aged 0 - 4 in South Africa. Pretoria, South Africa: United Nations Children’s Fund, 2007.

9. Lodico MG, Spaulding DT, Voegtle KH. Methods in Educational Research: From Theory to Practice. (2nd ed.). San Francisco: Wiley & Sons, Inc., 2010.

10. Dixit JB, Kumar R. Structured System Analysis and Design. New Delhi: Laxmi Publications, 2007.

11. Sussman F. The Power of Using Everyday Routines to Promote Young Children’s Language and Social Skills. Toronto: Hanen Early Program Centre, 2014.

12. Shenton AK. Strategies for ensuring trustworthiness in qualitative research projects. Educ Inf 2004;22:63-75.

MARCH 2016 Vol. 10 No. 1

RESEARCH

Usefulness of ultrasonography and biochemical features in the diagnosis of cholestatic jaundice in infants M S Choopa,1 MB ChB, FC Paed (SA); C Kock,1 MB ChB, FC Paed (SA), Cert Gastroenterology Paed (SA); S O M Manda,2 PhD (Stats); A J Terblanche,1 MB ChB, FC Paed (SA), Dip Allerg (SA), Cert Gastroenterology Paed (SA); D F Wittenberg,1 MD, FCP (Paed) (SA) Division of Gastroenterology, Department of Paediatrics and Child Health, School of Medicine, Faculty of Health Sciences, University of Pretoria, South Africa 2 Biostatistics Research Unit, South African Medical Research Council, Pretoria, South Africa, and School of Mathematics, Statistics and Computer Science, University of KwaZulu-Natal, Pietermaritzburg, South Africa 1

Corresponding author: M S Choopa (mchoopa@yahoo.co.uk) This work was a Master of Medicine in Paediatrics research study.

Background. Biliary atresia is a common cause of cholestasis. In our experience, patients with biliary atresia are referred late, with the diagnosis based on an absent gall bladder at ultrasonography. Such late referrals may render patients inoperable and not acceptable for formal intraoperative diagnosis.

Objectives. To determine the usefulness of an absent gall bladder on ultrasonography, and of biochemical features, in differentiating biliary atresia from other causes of cholestasis, using liver needle biopsy as a gold standard.

Methods. A retrospective file review of 150 infants presenting with cholestasis to Steve Biko Academic Hospital Paediatric Gastroenterology and Hepatology Unit from January 2008 to August 2014 was undertaken. Clinical, serum biochemical, abdominal ultrasonography and liver histology findings were analysed. Three groups were compared, based on liver histology findings, consisting of patients with biliary atresia, neonatal hepatitis, and other diagnoses, respectively.

Results. A total of 66/150 patients had biliary atresia, based on liver histology findings of extrahepatic obstruction. Their mean age was 4.7 (2.9) months, higher than in the other groups. In those with biliary atresia, the age at diagnosis, splenomegaly, and gamma glutamyl transferase (GGT), aspartate transaminase (AST) and GGT/AST ratio values were significantly different from the other groups. Total and conjugated bilirubin levels were similar among the groups. Ultrasonography was 69.7% sensitive, 98.8% specific and had a positive predictive value of 97.9% for biliary atresia. Ultrasonography missed 30.3% of patients with biliary atresia.

Conclusion. Ultrasonography has poor sensitivity but good specificity in screening for biliary atresia; however, other investigations are necessary to confirm the diagnosis.

S Afr J Child Health 2016;10(1):75-78. DOI:10.7196/SAJCH.2016.v10i1.1075

Cholestasis results from reduced excretion of bile due to a number of disorders[1] and is defined as conjugated bilirubin more than 20% of the total bilirubin when the total bilirubin is 85 µmol/L or more, and 17 µmol/L or more when the total bilirubin is 85 µmol/L or less. [2] It is good clinical practice to determine the conjugated fraction of bilirubin in any infant who remains jaundiced beyond the age of 2 weeks[3,4] and ultrasonography is one of the first investigations that is performed.[2] Identification of a normal gall bladder on sonogram is highly predictive of the absence of biliary atresia.[5] Visualisation of the gall bladder may be operator dependent and is also influenced by the patient feeding or fasting, even though biliary atresia can be accurately diagnosed when performed by an experienced sonographer.[6] The presence or absence of the triangular cord sign (TCS) on ultrasonography of the gall bladder has been described to assist in the diagnosis of biliary atresia and is the visualisation of a triangular or tubular echogenic density cranial to the portal vein bifurcation on a transverse or longitudinal ultrasonograhic scan. [7] After finding a TCS, explorative laparotomy should be done to confirm biliary atresia, while patients in whom the TCS is absent should have hepatobiliary scintigraphy.[7] Cholestasis can also be assessed by using serum biochemical markers such as liver function tests (total and conjugated bilirubin) and liver enzymes (alanine transaminase (ALT), aspartate 75

SAJCH

transaminase (AST), alkaline phosphatase (ALP) and gamma glutamyl transferase (GGT)). The serum GGT is usually higher in biliary atresia than in other causes of neonatal cholestasis when correlated with age.[8] Liver biopsy shows extrahepatic biliary obstruction by varying degrees of portal tract fibrosis, oedema, ductular proliferation and cholestasis with appearance of bile plugs.[8] It has been described as the most accurate investigation (99 - 100%) in differentiating biliary atresia from other causes of cholestasis.[2,4] Idiopathic neonatal hepatitis and biliary atresia are common and make up 70 - 80% of causes of cholestasis in term infants.[1,9,10] The incidence of biliary atresia ranges from 1 in 5 000 in Taiwan to 1 in 17 000 - 19 000 live births in the UK and France.[8] Biliary atresia is characterised by progressive obliterative cholangiopathy.[11] Therefore, it is crucial to diagnose biliary atresia for early surgical intervention[3] using the Kasai procedure, which is the resection of all the extrahepatic bile ducts and anastomosis of a jejunal en Y loop at the porta hepatis.[11] This procedure improves outcome and long-term survival of the infant’s native liver when performed between the ages of 45 and 60 days.[3,4] Liver transplantation is the only alternative therapeutic procedure[12] in those for whom the Kasai procedure is not feasible or has failed.

In view of our clinical experience that many patients are only referred for evaluation of cholestasis after the finding of an absent

MARCH 2016 Vol. 10 No. 1

RESEARCH gall bladder on ultrasonography examination, we aimed to determine the accuracy of ultrasonography in the diagnosis of biliary atresia and to determine the differences in clinical, biochemical and histological findings in infants who presented with different forms of cholestasis.

Methods Data

A retrospective review of files of patients aged 0 - 12 months, who presented with cholestasis to the Paediatric Gastroentero­ logy and Hepatology Unit, Steve Biko Academic Hospital, Pretoria from January 2008 to August 2014, was undertaken.

Ethical approval (Protocol number: 33/2015) was obtained from the University of Pretoria Health Sciences Research Ethics Committee.

Permission to review the hospital files was obtained from the Chief Executive Officer of the hospital. Out of 199 patient files reviewed, 150 patients were included in the study. Forty-nine patients had neither ultrasound of liver, liver biopsy nor both, or were above 1 year of age.

The following details were extracted from the files: gender, age at diagnosis, biochemical results of liver enzymes (ALT, AST, GGT, ALP) and liver function (total and conjugated bilirubin) at first presentation, ultrasonography reports and histological report of liver biopsies.

Patients aged 0 - 12 months who presented with cholestasis, and had either liver ultrasonography or liver biopsy or both, were included. Patients aged >12 months, or who did not have cholestasis, or who did not have an ultrasound or biopsy of the liver, were excluded. Biliary atresia was defined as the presence of extrahepatic obstruction on liver biopsy, based on bile duct proliferation, portal fibrosis and canalicular bile stasis.

We used liver histology findings even though this is an imperfect gold standard, because the majority of our patients were referred too late to be deemed operable for an intraoperative diagnosis during laparotomy.

Only 12 of the 66 patients with histological evidence of extrahepatic obstruction had exploratory laparotomies, and none had cholangiograms.

Hepatomegaly was defined as a liver that was palpable ≥2 cm below the right costal margin in the midclavicular line and splenomegaly was defined as any-sized spleen palpable below the left costal margin.

The TCS was infrequently reported on, therefore we included patients with absence of a gall bladder on ultrasonography as possibly having biliary atresia.

The liver enzyme, total and conjugated bilirubin, and histology results of the liver biopsies and abdominal ultrasonography, were tabulated.

Statistical methods

Continuous data were expressed as means with standard deviations (SDs) and categori­ cal data were summarised using frequencies and percentages. The independent t-test was used for comparison of normally distributed data; otherwise non-parametric alternatives were used. Performance of the diagnostic methods was evaluated using the area under the receiver-operating curve (ROC), known as the C-index. We used the kappa (κ) statistic to assess the proportion of observed-to-expected agreement between ultrasonography findings and liver enzymes with liver histology. Stata 13 (StataCorp, Texas, USA) was used for data analysis and p<0.05 was considered to be statistically significant.

Results

The liver histology findings were divided into three groups: group 1 consisted of those with the features of biliary atresia, group 2 had the features of neonatal hepatitis due to sepsis or cytomegalovirus infection, and group 3 consisted of other diagnoses. The last

group included intrahepatic paucity of bile ducts, Caroli disease, veno-occlusive disease, alpha-1 antitrypsin deficiency and steatosis.

The distribution of clinical and serum biochemical findings across the groups is shown (Table 1).

The overall mean (SD) age at diagnosis of the 150 patients was 3.7 (2.85) months, and differed significantly across the groups, with the biliary atresia group having the highest age of 4.7 (2.9) months (p<0.05) at diagnosis.

There were 89 boys and 61 girls of different races, with no significant difference between the sexes in those with biliary atresia.

There were significant differences in the proportions of patients with splenomegaly in the three groups, with more patients with biliary atresia presenting with splenomegaly compared with the other two groups. The GGT, AST and their ratio had different distributions across the groups, with GGT and the ratio to AST being higher in patients with biliary atresia (p<0.05). AST was significantly higher in the neonatal hepatitis group than in the other two groups (p<0.05). Of the biochemical features in the

Table 1. Clinical, ultrasonography and biochemical parameters by cause of cholestasis Biliary atresia (N=66, 44%)

Neonatal hepatitis (N=37, 24.7%)

Other diagnoses (N=47, 31.3%)

p-value

4.7 (2.9)

2 (1.74)

3.6 (2.7)

0.000

Male

32 (48.5)

26 (70.3)

31 (66.0)

0.052

Female

Age (months), mean (SD) Gender, n (%)

34 (51.5)

11 (29.7)

16 (34.0)

0.052

Hepatomegaly, n (%)

48 (72.0)

29 (80.6)

32 (68.1)

0.444

Splenomegaly, n (%)

44 (66.7)

14 (37.8)

30 (63.8)

0.018

Mean (SD)

604.5 (717.1)

197.1 (232.9)

408.3 (491)

0.0026

Median (IQR)

382.5 (139 - 714)

114 (85 - 212)

159 (68 - 688)

Mean (SD)

277.8 (184.7)

428.6 (396.4)

269.9 (199.5)

Median (IQR)

219 (142 - 374)

323 (181 - 478)

213 (140 - 326)

Mean (SD)

2.9 (3.7)

1.1 (1.8)

2.4 (3.93)

Median (IQR)

1.37 (0.53 - 3.58)

0.35 (0.18 - 1.13)

0.97 (0.3 - 3.2)

Mean (SD)

219.6 (76.8)

231.3 (115.8)

241.9 (120.7)

Median (IQR)

211.5 (160 - 261)

218 (160 - 274)

210 (151 - 319)

Mean (SD)

133.3 (55.5)

131.9 (79.9)

141.9 (76.7)

Median (IQR)

125 (97 - 151)

109 (71 - 161.5)

115 (92 - 164)

20/66 (30.3)

37/37 (100)

46/47 (97.9)

GGT (IU/L)

AST (IU/L) 0.0082

GGT/AST ratio 0.039

Total bilirubin (IU/L) 0.5179

Conjugated bilirubin (IU/L)

Gall bladder visible on ultrasound, n (%)

76

SAJCH

MARCH 2016 Vol. 10 No. 1

0.7569 0.000

1.00

Sensitivity

69.7

Specificity

98.8

Positive predictive value

97.9

Negative predictive value

80.5

0.00

0.25

Sensitivity 0.50

Table 3. Ultrasonography performance in determining biliary atresia in infants with cholestasis (%)

0.75

RESEARCH

0.00

0.25

0.50 1 – Specificity

bili_t ROC area: 0.5947 ggt ROC area: 0.6935 ast ROC area: 0.5425

0.75

1.00

alp ROC area: 0.5926 alt ROC area: 0.546 Reference

Fig. 1. ROC with area under the curve (AUC) of liver enzymes. The C-index of GGT was 0.67 (95% CI 0.58 – 0.76) while the C-indices of AST, ALT and ALP with 95% CI were 0.47, 0.52 and 0.58, respectively.

Table 2. ROC curve and AUC data to determine diagnostic accuracy of biliary atresia using serum biochemical investigations Obs, n

Area

SE

95% CI

ALP

149

0.585

0.047

0.492 - 0.678

GGT

149

0.671

0.045

0.582 - 0.759

ALT

149

0.521

0.048

0.428 - 0.615

AST

149

0.468

0.048

0.375 - 0.562

GGT to AST ratio

150

0.652

0.045

0.564 - 0.739

0.50 0.00

0.25

Sensitivity

0.75

1.00

Obs = observations; SE = standard error.

0.00

0.25

0.50 1 – Specificity

0.75

1.00

Area under ROC curve = 0.6519

Fig. 2. GGT to AST ratios shown as ROC and AUC. Graph indicates the AUR and ROC of GGT/AST ratio performance in the diagnosis of biliary atresia. The C-index of the GGT/AST ratio was 0.65 (95% CI 0.56 – 0.73).

biliary atresia group, the C-indices of GGT (Fig. 1, Table 2), and of the GGT to AST ratio (Fig. 2, Table 2), were statistically significant because their 95% confidence intervals did not include 0.5.[13] The total and conjugated bilirubin were similarly raised across all groups.

The performance of the ultrasonography finding of an absent gall bladder in determining biliary atresia in infants with cholestasis is outlined (Table 3). The sensitivity of ultrasonography was 69.7%, 77

SAJCH

specificity was 98.8%, and positive predictive value was 97.9%. Thus, in our series, ultrasonography was only able to identify biliary atresia in 69.7% of the 66 patients who had features of biliary atresia on histology. Of the 84 patients who did not have features of biliary atresia on histology, 98% had gall bladders on ultrasonography examination. The accuracy of the ultrasonography was therefore determined to be between 70 and 99%, and the agreement, measured by κ, was 0.71 (p<0.05).

Discussion

Of the 150 infants reviewed, 66 (44%) had biliary atresia (extrahepatic biliary obstruction on liver histology). Clinical, serum biochemistry and ultrasonography parameters were compared against liver histology results to assess their usefulness in the diagnosis of biliary atresia in infants with cholestasis. The mean age at diagnosis of biliary atresia was 4.7 (2.9) months and was higher than that of the other groups. Many patients with cholestasis were referred for further management of suspected biliary atresia after initial ultrasonography had indicated the presence of a gall bladder. In most of these patients, a diagnosis of biliary atresia was made. In our study, 30.3% of patients with extrahepatic biliary obstruction on histology were diagnosed as having gall bladders on ultrasonography examination. Many patients presented in advanced liver disease with liver dysfunction and portal hypertension, and the Kasai procedure would not have improved their outcome, which is improved if surgery is performed before the age of 8 weeks.[3] Other factors that may have contributed to the late presentation of patients with biliary atresia include failure of clinicians to realise the importance of persisting cholestasis and failure to detect jaundice clinically in dark-skinned babies.

There was no significant difference in the size of the liver among the groups and more than 70% of all patients had hepatomegaly.

Splenomegaly was more frequent in those with biliary atresia due to the late presentation of patients, indicating portal hypertension. [8] One of the reasons fewer patients with neonatal hepatitis had splenomegaly may have been due to the shorter duration of the disease. One study showed hepatomegaly and splenomegaly in 50% of cases of cholestasis in infants;[14] however, in our study, 73% of infants with cholestasis had hepatomegaly while 59% had splenomegaly.

Park et al.[7] described the use of TCS in the diagnosis of biliary atresia, the finding of which was followed by mandatory explorative laparotomy. The features of TCS are a triangular or tubular shaped echogenic density cranial to the portal vein bifurcation on ultrasonography of the gall bladder.[7] In our study, this sign was reported on very infrequently, therefore we assessed the value of visualising the gall bladder in the diagnosis of biliary atresia. Ultrasonography had a sensitivity of 69.7%, a specificity of 98.8% and an accuracy of 86.0% in the diagnosis of biliary atresia. A total of 20 of the 103 patients who had gall bladders on ultrasonography had extrahepatic biliary obstruction on histology.

Therefore, ultrasonography did not identify 30.3% of the 66 cases with biliary atresia. Ultrasonography was only able to exclude biliary atresia correctly in 80.5% of the 103 patients with a visible gall bladder (negative predictive value). These findings were marginally comparable to those of Park et al.,[7] who reported that the use of

MARCH 2016 Vol. 10 No. 1

RESEARCH ultrasonography in the diagnosis of biliary atresia using the TCS had a sensitivity, specificity and accuracy of 85%, 100% and 95%, respectively. Dehghani et al.[9] reported a sensitivity of 52% and specificity of 76.1% using ultrasonography in the diagnosis of biliary atresia. However, experience in performing ultrasonography is important.[6] Serum GGT levels were significantly higher in infants with biliary atresia (median 382.5, IQR 139 - 714) than with neonatal hepatitis (median 114, IQR 85 - 212) or with other diagnoses (median 159, IQR 68 - 688), p<0.05. GGT is usually higher in biliary atresia than in other causes of neonatal cholestasis, especially when correlated with age.[8] The GGT levels in our series were similar to those reported by Tang et al.,[15] which were significantly higher in patients with biliary atresia than in neonatal hepatitis, and with a GGT/AST ratio >2 in 55/68 infants with biliary atresia and 15/54 with neonatal hepatitis.

In that study, a GGT level >300 IU/L had a sensitivity of 39.7% and a GGT/AST ratio >2 showed a high possibility of biliary atresia.[15] In our series, the GGT/AST ratio was significantly higher in those with biliary atresia (mean 2.9, SD 3.7) than in those with neonatal hepatitis (mean 1.1, SD 1.8) and other diagnoses (mean 2.4, SD 3.9; p<0.05). AST, ALT and ALP levels were high in all groups, with AST significantly higher in those with neonatal hepatitis (Table 1). Even though the AST and ALT may only be marginally elevated in early cases of biliary atresia, these enzymes increase as progressive liver damage occurs, as seen in our patients who presented late. The causes of neonatal hepatitis in our series included sepsis, cytomegalovirus infection and idiopathic neonatal hepatitis. The total and conjugated bilirubin levels were high in all the groups and were not significantly different (p>0.05) between the three groups.

Conclusion

Ultrasonography has poor sensitivity but good specificity in screening for biliary atresia; however, other investigations are necessary to confirm the diagnosis. Many patients with biliary atresia were referred too late for surgical intervention. Therefore, all babies with cholestatic jaundice must be referred early for specialist evaluation and exclusion of biliary atresia.

Acknowledgement. We thank Prof. G van Biljon of the Department of Paediatrics and Child Health, for advice on a scientific writing course by the Department of Research and Innovation Support, University of Pretoria.

78

SAJCH

References 1. Yang J, Ma D, Peng Y, Song L, Li C. Comparison of different diagnostic methods for differentiating biliary atresia from idiopathic neonatal hepatitis. Clin Imaging 2009;33(6):439-446. [http://dx.doi.org/10.1016/j. clinimag.2009.01.003] 2. Poddar U, Thapa B, Bhattacharya A, Rao K, Singh K. Neonatal cholestasis: Differentiation of biliary atresia from neonatal hepatitis in a developing country. Acta Paediatr 2009;98(8):1260-1264. [http://dx.doi.org/10.1111/ j.1651-2227.2009.01338.x] 3. Cartledge P, Kevlani N, Shapiro L, McClean P. An audit of a community protocol for identifying neonatal liver disease. Arch Dis Child 2012;97(2):166168. [http://dx.doi.org/10.1136/adc.2009.179812] 4. Moyer V, Freese DK, Whitington PF, et al. Guideline for the evaluation of cholestatic jaundice in infants: Recommendations of the North American Society for Pediatric Gastroenterology, Hepatology and Nutrition. J Pediatr Gastroenterol Nutr 2004;39(2):115-128. [http://dx.doi.org/10.1097/00005176200408000-00001] 5. Nwomeh B, Caniono D, Hogan M. Definitive exclusion of biliary atresia in infants with cholestatic jaundice: The role of percutaneous cholecystocholangiography. Pediatr Surg Int 2007;23(9):845-849. [http:dx.doi. org/10.1007/s00383-007-1938-2] 6. Takamizawa S, Zaima A, Muraji T, Akasaka Y, Satoh S, Nishijima E. Can biliary atresia be diagnosed by ultrasonography alone? J Pediatr Surg 2007;42(12):2093-2096. [http://dx.doi.org/10.1016/j.jpedsurg.2007.08.032] 7. Park W, Choi S, Lee H, Kim S, Zeon S, Lee S. A new diagnostic approach to biliary atresia with emphasis on ultrasonographic triangular cord sign: Comparison of ultrasonography, hepatobiliary scintigraphy and liver needle biopsy in the evaluation of infantile cholestasis. J Pediatr Surg 1997;32(11);1555-1559. [http://dx.doi.org/10.1016/S0022-3468(97)904516] 8. Hartley JL, Davenport M, Kelly DA. Biliary atresia. Lancet 2009;374(9702):17041713. [http://dx.doi.org/10.1016/S0140-6736(09)60946-6] 9. Dehghani S, Mahmood H, Imanieh M, Geramizadeh B. Comparison of different diagnostic methods in infants with cholestasis. World J Gastroenterol 2006;12(36):5893-5896.

10. Deghady MAA, Abdel-Fattah M, Abdel-Kader M, Naguib M, Madina E, Abd El Gawad M. Diagnostic evaluation of cholestasis in infants and young children in Alexandria. Internet J Pediatr Neonatol 2005;6(1). http://www.ispub.com/ IJPN/6/1/8863 (accessed 26 February 2015).

11. Chardot C, Serinet M. Prognosis of biliary atresia: What can be further improved? J Pediatr 2006;148(4):432-434. [http://dx.doi.org/10.1016/j. jpeds.2006.01.049] 12. Davenport M, Puricelli V, Farrant P, et al. The outcome of the older (â&#x2030;Ľ100 days) infant with biliary atresia. J Pediatr Surg 2004;39(4):575-581. [http://dx.doi. org/10.1016/j.jpedsug.2003.12.014] 13. Tripepi G, Jager KJ, Dekker FW, Zoccali C. Statistical methods for the assessment of prognostic biomarkers (Part 1): Discrimination. Nephrol Dial Transplant 2010;25:1399-1401.

14. Elfaramawy A. Cholestasis in neonates and infants. Egypt J Med Hum Genet 2008;9(2):135-147.

15. Tang K, Huang Y, Lai C, et al. Gamma glutamyl transferase in the diagnosis of biliary atresia. Acta Paediatr Taiwan 2007;48(4):196-200.

MARCH 2016 Vol. 10 No. 1

RESEARCH

An evaluation of the screening for retinopathy of prematurity in very-low-birth-weight babies at a tertiary hospital in Johannesburg, South Africa Z Dadoo, MB BCh, FCPaed (SA); D E Ballot, MB BCh, FCPaed (SA), PhD Department of Paediatrics and Child Health, University of the Witwatersrand and Charlotte Maxeke Johannesburg Academic Hospital, Johannesburg, South Africa Corresponding author: Z Dadoo (drzdadoo@gmail.com) Background. Retinopathy of prematurity (ROP) is a leading cause of blindness for very-low-birth-weight (VLBW, <1 500 g) babies. ROP screening identifies babies that require treatment to prevent major visual impairment.

Objectives. To evaluate the screening for ROP at Charlotte Maxeke Johannesburg Academic Hospital (CMJAH) by reviewing the number of babies screened according to the CMJAH guidelines, the grades of ROP found and the treatment modality received.

Methods. This was a retrospective record review of VLBW babies born between 1 January 2013 and 31 December 2013 at CMJAH, whether inborn or transferred in. The babies were divided into two groups based on age at final outcome. Final outcome was defined as death, discharge or transfer out of the unit. The ‘early’ outcome group had their final outcome before day 28 of life. The ‘late’ outcome group had their final outcome at day 28 or more of life. The early outcome group qualified for outpatient ROP screening and the late outcome group qualified for inpatient ROP screening.

Results. There were a total of 572 VLBW babies at CMJAH during this time period. The babies had a mean birth weight of 1 127 (standard deviation (SD) 244.75) g and gestational age of 29 (2.743) weeks. The mean duration of stay was 29 (21.66) days and there were 309 female babies. Of these 572 babies, 304 comprised the early outcome group and 268 comprised the late outcome group. In the early outcome group babies who were transferred out of the unit or died were excluded; therefore the remaining 147 babies discharged home qualified for outpatient ROP screening. Inpatient ROP screening was carried out in 36/147 (24.4%) of these babies (not in accordance with ROP screening guidelines). ROP was documented in 4/36 (11.1%). Outpatient ROP screening records were unavailable. Exclusions from the late outcome group included five babies. In the late outcome group 111/263 (42.2%) were screened for ROP. ROP was found in 17%. One baby required treatment with intravitreal antivascular endothelial growth factor (VEGF) and three babies required surgery.

Conclusions. More than half of the babies in the late outcome group were not screened during their stay (57.8%). More than one-third of babies were discharged prior to reaching the current recommended age for screening. Efforts need to be intensified to identify and screen all eligible babies prior to discharge. Outpatient ROP screening is not well documented; therefore prevalence cannot be established.

S Afr J Child Health 2016;10(1):79-82. DOI:10.7196/SAJCH.2016.v10i1.1099

Retinopathy of prematurity (ROP) is known to be an important cause of visual impairment and blindness in the surviving premature population. Over the last 10 - 15 years an estimated 50 000 children are blind as a result of ROP, and it is likely that many more are unilaterally blind or visually impaired.[1] As the disease can be present without any symptoms or clinical signs, it is necessary to screen premature babies for ROP. Most ROP will resolve by itself and only requires continued monitoring until resolution or maturation of retinal vessels occurs. However, severe forms of ROP require treatment to preserve or salvage vision and to improve quality of life.

In developed countries two epidemics of blindness due to ROP have been described. The first occurred predominantly in the USA in the 1940s - 1950s. The principal risk factor was the supply of unmonitored supplemental oxygen to the premature baby. The subsequent restriction in oxygen use led to a decrease in blindness due to ROP. The second epidemic started in the 1970s as a result of the higher survival rates of extremely premature babies secondary to advances in neonatal intensive care units (NICUs).[1] A third epidemic of ROP is currently said to be occurring in middleincome countries[2] such as South Africa (SA). Reasons for this include improved survival of premature babies in these countries, together with a lack of adequate monitoring of oxygen therapy. Countries with infant mortality rates (IMRs) >60/1 000 live births do not usually have NICUs; therefore, premature babies do not survive and these countries have a low incidence of ROP.[2] Countries with IMRs of 9 60/1 000 live births represent the highest burden of blindness caused 79

SAJCH

by ROP, as more premature babies survive in NICUs where oxygen administration may be poorly monitored.[1] SA’s IMR for 2011 was 35/1 000 live births.[3] As we succeed in improving our IMR, strategies need to be in place to target prevention of known risk factors for ROP and screening for ROP that may require treatment.

If screening programmes are not put in place, the incidence of blindness from untreated ROP is likely to increase. It was first reported in 1988 that treatment could improve the outcome for severe ROP.[4] This makes ROP screening a priority. The World Health Organization (WHO)’s Vision 2020 programme has recognised ROP as an important cause of childhood blindness in industrialised and middle-income countries.[5] Their strategies advocate examining premature babies at risk of ROP, treating those premature babies with severe ROP and promoting oxygen monitoring to all premature babies receiving oxygen therapy.

The two important aspects of screening for ROP are who to screen and when to screen them. Knowledge of risk factors for ROP helps to identify who needs to be screened. Risk factors for ROP are divided into two groups – prenatal and postnatal.[6] Prenatal factors include gestational age (GA) and birth weight. Postnatal factors include prolonged exposure to oxygen and other identified markers of neonatal illness severity. Examples of the latter include the need for mechanical ventilation, the presence of sepsis or intraventricular haemorrhage, the administration of blood transfusions and poor postnatal weight gain.[7] Low levels of serum insulin-like growth factor 1 (IGF-1) are found in babies with poor postnatal weight

MARCH 2016 Vol. 10 No. 1

RESEARCH gain.[8] General criteria used in screening programmes are birth weight and GA, combined with sickness criteria. [9] The recommended age for screening is based on the timing of the occurrence of ROP and is related to the maturity of the retinal vessels.

There are concerns that in middle- and low-income countries, compared with highincome countries, a greater number of older and larger babies are presenting with ROP. In a large prospective study done at Chris Hani Baragwanath Academic Hospital to establish the frequency of ROP it was concluded that this was not the case, and the screening weight could safely be lowered to 1 250 g.[10] According to the latest ROP guidelines published in the South African Medical Journal (SAMJ), all very-low-birth-weight (VLBW) babies <1 500 g or 32 weeks’ GA should be screened for ROP.[11] Screening is repeated until retinal vascularisation has reached a stage where the risk of a serious adverse outcome is considered minimal. ROP screening is carried out by an ophthalmologist and by means of indirect ophthalmoscopy. Newer screening techniques involve the use of digital cameras to capture images of the retina. The 2013 SAMJ guidelines recommend screening all VLBW babies at 4 - 6 weeks chronological age

or 31 - 33 weeks corrected GA – whichever comes last. The guidelines detail where and how to screen, as well as how to follow up and manage patients and when to stop screening. [11] These guidelines are in line with the WHO Vision 2020 strategy. Vision 2020 ensures the availability of ophthalmologists experienced in indirect ophthalmoscopy who can identify premature babies who require treatment for ROP, that babies at risk for ROP have their fundi examined starting 4 - 6 weeks after birth, and that those with severe disease are treated immediately.[5] ROP is classified according to the Inter­ national Classification of ROP (ICROP), was standardised in 1984, and updated in 1987 and again in 2005.[12] ROP is characterised by using four components (Table 1): (i) loca­ tion (zones 1 - 3); (ii) severity (stages 1 - 5); (iii) extent (circumferential location of the disease reported as clock hours); and (iv) plus disease (tortuosity of posterior retinal vessels).[7] Two important definitions are those of threshold ROP and prethreshold ROP (Table 2). Threshold ROP carries a risk of blindness of 50%, which can be reduced to 25% with treatment. Pre-threshold ROP can require either treatment or close observation – depending on the type. The various treat­

Table 1. The four components included in the classification of ROP[7] Component

Description

Location

How far the developing retinal blood vessels have progressed The retina is divided into three concentric circles or zones Zone 1: imaginary circle with optic nerve at the centre Zone 2: extends from the edge of zone 1 to the ora serrata on the nasal side of the eye and half the distance to the ora serrata on the temporal side

Severity

ment options available for ROP include cryotherapy, laser ablative therapy, intravitreal antivascular endothelial growth factor (antiVEGF) and retinal reattachment. [7] Not all of these options are available in our setting.

In SA, approximately 1 in 5 of all VLBW babies is affected by ROP. A study at Chris Hani Baragwanath Academic Hospital found the incidence to be ~17%.[10] In another study undertaken at Tygerberg Children’s Hospital in Cape Town, the incidence was found to be 21.8%.[13] A study by Delport et al.[14] at Kalafong Hospital, Pretoria, found the incidence of ROP to be 24.5%. The incidence at Charlotte Maxeke Johannesburg Academic Hospital (CMJAH) is unknown.

The present study aimed to review the screening programme for ROP in VLBW babies at CMJAH.

Methods

This study was undertaken at CMJAH, a tertiary care institution that serves as a referral centre for the primary care clinics and other hospitals in the area. It was a retrospective record review of all the VLBW babies admitted to CMJAH from 1 January 2013 to 31 December 2013, whether inborn or transferred in. Babies who died or were transferred out before day 28 of life were excluded from the present study.

Patient information was obtained from an existing neonatal VLBW database at CMJAH, which is kept for the purpose of clinical audit purposes. The database consists of standard information that is collected upon the discharge of each baby.

Data are managed using REDCap (Research Electronic Data Capture) tools hosted by

Zone 3: the outer crescent-shaped area extending from zone 2 out to the ora serrata temporally

Table 2. Important definitions used in ROP screening[7]

The stage of the disease

Threshold ROP

Zone 1 or 2: 8 cumulative clock hours of stage 3 with plus disease

Prethreshold ROP

Zone 1: Any ROP less than threshold

Stage 1: a demarcation line between normal and avascular retina Stage 2: a ridge of fibrovascular tissue replaces the demarcation line Stage 3: abnormal blood vessels and fibrous tissue develop on the edge of the ridge and extend into the vitreous Stage 4: partial retinal detachment

Zone 2: stage 2 ROP with plus disease stage 3 without plus disease stage 3 with plus disease but less than 8 cumulative clock hours

Stage 5: complete retinal detachment Extent

Circumferential location of the disease and reported as clock hours in the appropriate zone

Plus disease

Refers to the presence of vascular dilatation and tortuosity of the posterior retinal vessels in at least two quadrants

Table 3. ROP grading results in babies screened as inpatients Group

No ROP, n (%)

Grade 1 ROP, n (%)

Grade 2 ROP, n (%)

Grade 3 ROP, n (%)

Grade 4 ROP, n (%)

Grade not recorded, n (%) Total, n

Early outcome group

28 (77.8)

2 (5.6)

2 (5.6)

0 (0)

0 (0)

4 (11.1)

36

Late outcome group

81 (73.0)

11 (9.9)

5 (4.5)

2 (1.8)

1 (0.9)

11 (9.9)

111

80

SAJCH

MARCH 2016 Vol. 10 No. 1

RESEARCH the University of the Witwatersrand.[15] All definitions in the database are according to the Vermont Oxford Network (VON) (www. vtoxford.org).

The ROP screening guidelines for CMJAH were derived from the SAMJ 2013 ROP screening guidelines, and state that all VLBW babies or those born at a GA <32 weeks should be screened at 4 - 6 weeks chronological age. In babies who were screened more than once, the worst grade of ROP recorded was used for the purpose of the study. Intravitreal anti-VEGF and surgery for ROP were available at CMJAH at the time of the study.

Prior to discharge, all VLBW babies at CMJAH were transferred to kangaroo mother care (KMC) once their current weight was >1 000 g, they were tolerating full enteral feeds and they were off supplemental oxygen.

Whenever possible, these babies were transferred to regional hospitals for continuing care. The VLBW babies were discharged from hospital once they had reached a weight >1 600 g, were taking full oral feeds (either by cup or breast) and were maintaining their blood glucose levels. Babies were referred for ROP screening to the ophthalmologist at the discretion of the attending paediatric registrar, in accordance with the abovementioned guidelines. Babies in KMC were included in the screening programme. Results of the ROP screening were recorded on the daily charts for each patient.

IBM (USA) SPSS Statistics version 22 for analysis. Data were described using standard statistical methods. Categorical variables were described using frequencies and percentages, and continuous variables by using measures of central tendency – mean and standard deviation – as the data were normally distributed.

Ethics approval

The study was approved by the Committee for Research on Human Subjects, University of the Witwatersrand, Johannesburg (clearance certificate No. M130947).

Results

A total of 572 (309 female) VLBW babies were admitted to the neonatal unit during the study period. A total of 162 babies were excluded.

There were 128 deaths and 29 transfers to regional hospitals prior to 28 days. Five babies were transferred in to the unit after 28 days of life for surgical procedures; 2 died immediately and 3 were transferred back to their original hospitals within 2 days. The final sample therefore included 410 babies. The mean birth weight was 1 127 g with a standard deviation (SD) of 244.75 g, and the mean (SD) GA was 29 (2.74) weeks. The mean age at admission was 1 day (5.806) and the babies had a mean

duration of stay of 28 (21.66) days. There were 147 babies in the early outcome group and 263 babies in the late outcome group (Fig. 1).

ROP screening was documented in 147/410 (35.9%) VLBW babies. The ROP findings are summarised in Table 3. Plus disease was not found in any of the babies.

Intravitreal anti-VEGF treatment was used in one baby and surgical treatment was documented in three babies.

Although the 147 babies in the early outcome group were not required to be screened as inpatients, ROP screening was carried out in 36 (24.5%) and of these 4 (11.1%) had evidence of ROP. Screening for ROP was undertaken in 111/263 (42.2%) babies in the late outcome group and 19 (17.1%) had evidence of ROP.

Discussion

This study shows that less than half of the VLBW babies at CMJAH eligible for inpatient ROP screening according to the hospital’s guidelines were actually screened for ROP. More than one-third of babies were discharged before they had reached the required age for screening. Despite this, 24.5% of these babies were screened before 4 weeks of life. Of concern is that 11.1% of

572 VLBW babies

Groups

The VLBW babies in the study population were divided into two groups based on the calculated chronological age at final outcome, in accordance with the ROP screening guidelines mentioned above. Final outcome was defined as death, discharge or transfer out of the unit. The final outcome for the ‘early’ outcome group occurred before day 28 of life; in the ‘late’ outcome group final outcome occurred on day 28 of life or later. The early outcome group qualified for outpatient ROP screening and the late outcome group qualified for inpatient ROP screening.

304 early outcome

Excluded

Statistical analysis

The relevant data for the present study were extracted from the neonatal database and exported to a Microsoft Excel (USA) spread­ sheet. Demographic information, out­ come, whether ROP screening had been per­ formed and the grade and treatment of ROP (intravitreal anti-VEGF or surgery) were collected for each patient. Duration of stay and chronological age at final outcome (discharge, death or transfer out to a regional hospital) were calculated.

The data from the Excel spreadsheet were imported to the statistical software

268 late outcome

128 early deaths

Eligible for outpatient ROP SCREENING

29 early transfers

147 early discharges

Excluded 2 late deaths 3 late transfers

Eligible for inpatient ROP SCREENING 237 late discharges 17 late transfers 9 late deaths

36 screened for ROP (as inpatients)

Fig. 1. Number of babies in each group and their final outcome.

81

SAJCH

MARCH 2016 Vol. 10 No. 1

111 screened for ROP (106 late discharges, 4 late transfers out, 1 late death)

RESEARCH these babies had ROP. This group of early discharges is important as they require outpatient ROP follow-up. It is not known whether these babies attended screening for ROP as outpatients, so it is possible that a number of babies with ROP were missed. Education of caregivers in this group is essential, as defaulters to follow-up are at risk of presenting with more severe grades of ROP and increased morbidity.

Although a true prevalence for ROP at CMJAH for 2013 cannot be calculated, as a result of the low level of screening, ROP was found in 15.6% of VLBW babies, which is similar to the 17% rate reported at Chris Hani Baragwanath Academic Hospital.[10] Other SA studies reported slightly higher rates – 21.8% at Tygerberg Children’s Hospital [13] and 24.5% at Kalafong Hospital.[14] This review shows that the inpatient ROP screening at CMJAH is not optimal and needs to be improved. Inpatient ROP screening was not carried out in 57.8%. Babies at risk need to be promptly identified. The attending medical staff (interns, medical officers, registrars and consultants) should to be familiar with the guidelines.

Junior staff will need to be educated on the harms of oxygen therapy and the subsequent complication of ROP and its consequences.

Although it may seem attractive to delay the discharge of VLBW babies until they have achieved the recommended age for ROP screening, this is not feasible because of high patient numbers and extreme pressure for beds.

Adjusting the screening protocol to allow ROP screening at a younger age in those babies who will be discharged before 28 days of age would be a simpler solution and would prevent missed opportunities to identify babies with ROP. No babies were recorded to have plus disease. These data may have not been captured on discharge, were truly not present or may have been under-reported by the ophthalmologist performing the screening. Bigger babies are also at risk of ROP. The 2013 SAMJ ROP guideline suggests that premature babies with weights between 1 500 g and 2 000 g may also be at risk if they have risk factors; if oxygen monitoring in this group of babies has been suboptimal then screening should be considered.[11] This group of babies was not included in the present study, but should not be overlooked in screening programmes for ROP. Ideally, an electronic prospective data capture system needs to be implemented to capture all the results of ROP screening – both inpatient and outpatient. This would only be possible in conjunction with the Department of Ophthalmology, especially regarding the outpatient screening. This will assist greatly with future research and in gauging the incidence of ROP at CMJAH.

Study limitations

One limitation is the design of the study – the retrospective nature of the study means a precollected dataset was used. ROP information is not available for babies on the low-birth-weight (LBW) database who may have a GA of <32 weeks but a weight of >1 500 g. Another potential limitation is that of inter-observer error in classifying the grade of ROP present, as different ophthalmology registrars did the screening, with different levels of skill and experience.

82

SAJCH

Conclusion

More than half of VLBW babies who met criteria for ROP screening according to CMJAH ROP screening guidelines were not screened during their inpatient stay. Efforts need to be intensified to identify these babies and screen them prior to discharge.

Records for outpatient ROP screening are not well organised and not easily accessible at both the neonatal follow-up clinic and the ophthalmology unit. There is a need for a co-ordinated database between the two specialties. In this regard, a true prevalence for ROP at CMJAH cannot be established.

Screening for ROP should include all babies with a GA of <32 weeks (even if their weight is >1 500 g). In addition, babies weighing between 1 500 g and 2 000 g with risk factors for ROP should not be omitted from screening programmes.

References

1. Gilbert C. Retinopathy of prematurity: A global perspective of the epidemics, population of babies at risk and implications for control. Early Hum Dev 2008;84(2):77-82. [http://dx.doi.org/10.1016/j.earlhumdev.2007.11.009] 2. Zin A, Gole G. Retinopathy of prematurity – incidence today. Clin Perinatology 2013;40(2):185-200. [http://dx.doi.org/10.1016/j.clp.2013.02.001] 3. The World Bank. Infant mortality rate (per 1 000 live births). http://data. worldbank.org/indicator/SP.DYN.IMRT.IN (accessed 11 September 2013).

4. Cryotherapy for retinopathy of prematurity cooperative group. Multicentre trial of cryotherapy for retinopathy of prematurity: Preliminary results. Arch Ophthalmol 1998;106(4):471-479. [http://dx.doi.org/10.1001/archopht.1988.01060130517027] 5. World Health Organization and the International Agency for the Prevention of Blindness Joint Initiative. Vision 2020: The Right to Sight Action Plan 2006-2010. http://www.iapb.org/vision-2020/what-is-avoidable-blindness/ childhood-blindness (accessed 11 September 2013).

6. Wikstrand M, Hard A, Niklasson A, Smith L, Lofqvist C, Hellstrom A. Maternal and neonatal factors associated with poor early weight gain and later retinopathy of prematurity. Acta Paediatr 2011;100(12):1528-1533. [http:// dx.doi.org/10.1111/j.1651-2227.2011.02394.x] 7. Cloherty JP, Eichenwald EC, Hansen AR, Stark AR. Manual of Neonatal Care.

2nd ed. Philadelphia: Lippincott Williams & Wilkins, 2012:840-845.

8. Smith L, Hard AL, Hellstrom A. The biology of retinopathy of prematurity. Clin Perinatology 2013;40(2):201-214. [http://dx.doi.org/10.1016/j.clp.2013.02.002] 9. Holmstrom G, Hellstrom A, Jakobsson P, Lundgren P, Tornqvist K, Wallin A. Swedish national register for retinopathy of prematurity (SWEDROP) and the evaluation of screening in Sweden. Arch Ophthalmol 2012;130(11):1418-1424. [http://dx.doi.org/10.1001/archophthalmol.2012.2357] 10. Mayet I, Cockinos C. Retinopathy of prematurity in South Africa at a tertiary hospital: A prospective study. Eye (Lond) 2006;20(1):29-31. [http://dx.doi. org/10.1038/sj.eye.6701779] 11. Visser L, Singh R, Young M, Lewis H, McKerrow N (ROP working group South Africa). Guideline for the prevention, screening and treatment of retinopathy of prematurity (ROP). S Afr Med J 2013;102(2):116-125. [http://dx.doi. org/10.7196/samj.6305] 12. International Committee for the Classification of Retinopathy of Prematurity.

The international classification of retinopathy of prematurity revisited. Arch Ophthalmol 2005;123(7):991-999. [http://dx.doi.org/10.1001/archopht.123.7.991] 13. Van der Merwe S, Freeman N, Bekker A, Harvey J, Smith J. Prevalence of and risk factors for retinopathy of prematurity in a cohort of preterm infants treated exclusively with non-invasive ventilation in the first week after birth. S Afr Med J 2013;103(2):96-100. [http://dx.doi.org/10.7196/samj.6131] 14. Delport SD, Swanepoel JC, Odendall PJL, Roux P. Incidence of retinopathy of prematurity in very low birth weight infants born at Kalafong Hospital, Pretoria. S Afr Med J 2002;92(12):986-990.

15. Harris PA, Taylor R, Thielke R, Payne R, Gonzalez N, Conde JG. Research electronic data capture (REDCap) – a metadata-driven methodology and workflow process for providing translational research informatics support. J Biomed Inform 2009;43(2):377-381. [http://dx.doi.org/10.1016/j. jbi.2008.08.010]

MARCH 2016 Vol. 10 No. 1

RESEARCH

Perceptions of community-based human milk banks before and after training in a resource-limited South African setting H E Goodfellow,1 MB ChB; P Reimers,2 MTech (Nursing); K Israel-Ballard,3 DrPH; A Coutsoudis,2 PhD Department of Paediatrics and Child Health, School of Clinical Medicine, University of KwaZulu-Natal, Durban, South Africa Department of Paediatrics and Child Health, Nelson Mandela School of Medicine, University of KwaZulu-Natal, Durban, South Africa 3 Maternal and Child Health and Nutrition Global Program, Seattle, USA 1 2

Corresponding author: H E Goodfellow (heloise_vdb@hotmail.com)

Background. Human breastmilk provides gold standard nutrition and immunological support to infants. For low birth weight, HIVinfected, HIV-exposed or otherwise vulnerable babies, it can mean the difference between life and death. When a mother’s own milk is not available, safe, donated human breastmilk is an excellent alternative. High rates of under-5 mortality have prompted the South African (SA) Ministry of Health to commit to scaling up human milk banks in key health facilities. Community-based human milk banks (CBHMBs) have the potential to complement these efforts, but there is little research on the feasibility and acceptability of this approach.

Objective. To determine mothers’ perceptions of breastfeeding and CBHMBs, and to ascertain how training could affect those perceptions.

Methods. A total of 40 black mothers in KwaZulu-Natal, SA, participated in a survey on breastfeeding and human milk banks (HMBs) prior to commencing a breastfeeding peer-counselling training course that included information on HMBs. The survey was re-administered following the completion of the module on HMBs. The questionnaire was repeated ~4 months later.

Results. Following training, significant changes were observed in mothers’ knowledge and perceptions around donor milk and safety. No significant changes were observed in reports of what these mothers presumed the community’s perceptions around donor milk banking were.

Conclusion. Education of mothers can play an important role in supporting CBHMBs by improving mothers’ perceptions and acceptance of breastfeeding, donor milk, and milk banking. Changing community concerns around HMBs will require more than just changing the perceptions of mothers.

S Afr J Child Health 2016;10(1):83-86. DOI:10.7196/SAJCH.2016.v10i1.1103

Human breastmilk is the gold standard in infant nutrition. It is particularly crucial for vulnerable, orphaned, or low birth weight infants and those exposed to, or infected with HIV. It provides powerful nutritional and immunological protection and reduces mortality due to infectious diseases.[1-3] It also lowers the risk of necrotising enterocolitis among preterm, low birth weight infants. [4,5] Globally, the promotion of exclusive breastfeeding is seen as an important public health priority.[6] When a mother’s own breastmilk is not available, the World Health Organization recommends donated human breastmilk (DHM) as the best alternative.[7] As a result, countries worldwide are establishing human milk banks (HMBs) to collect, pasteurise, and provide safe DHM to vulnerable infants. In 2011, the South African (SA) Ministry of Health committed to promoting and supporting HMBs in postnatal wards and neonatal intensive care units (NICUs) to reduce neonatal and postnatal morbidity and mortality.[8] The use of DHM in NICUs was seen as a temporary measure while the mothers are supported to establish their own breastmilk supply, and therefore HMBs were seen as part of the strategy to support and promote breastfeeding.

SA’s AIDS epidemic continues to exact a devastating toll on mothers, leaving many infants and children orphaned. According to a 2011 census, 18.8% of SA children between birth and 17 years of age had lost one or both parents to the disease. In KwaZulu-Natal (KZN), one of the poorest provinces in the country, this number rose to more than a quarter (27%).[9] Given the poor socioeconomic conditions in many parts of KZN, the provision of adequate nutrition to infants younger than 6 months of age is a major dilemma. This has stimulated the concept of community-based HMBs (CBHMBs), which operate outside a health facility, in locations such as homes for orphans or other community centres. These CBHMBs have the potential to 83

SAJCH

reduce infant mortality by expanding access to lifesaving human milk to vulnerable infants beyond those only in a neonatal ward. Although research has demonstrated that HMBs in NICUs are acceptable and feasible in SA[10] and beyond,[5,11–14] there is, to our knowledge, no data recording the acceptability or experiences of CBHMBs. Successfully increasing CBHMBs would require an understanding of the benefits of breastfeeding and knowledge and acceptance of donor milk banking. Rigorous quality-control systems would also need to be established to ensure the safety of donor milk. The objective of this study was to determine mothers’ perceptions of CBHMBs and how education and training could affect or improve their opinions.

Methods

The research team interviewed women who were participating in a community nutrition programme being conducted at the Cato Manor Clinic in an informal settlement in Durban in KZN. This province has the third-highest unemployment rate (33%) in SA.[9] The community nutrition programme conducted at the clinic included an interactive breastfeeding peer counselling training course, which was offered to mothers who had recently delivered babies. The course not only assisted them with their own breastfeeding practices but also looked to prepare them to become breastfeeding counsellors in the community. The mothers were enrolled in the course when their infants were approximately 6 weeks of age. One-hour long sessions were held every 2 weeks until the infants reached 5 months of age. Thereafter the sessions were held monthly until the infants were 12 months of age.

The research team verbally administered questionnaires to mothers when they first enrolled in the programme and before they received

MARCH 2016 Vol. 10 No. 1

RESEARCH any training sessions. The same questionnaires were re-administered to these mothers ~22 weeks post delivery. At this stage they had completed the HMB training module and had been part of the training course for about 4 months. The questionnaires, which were administered by a research assistant fluent in English and isiZulu, contained questions on demographics, as well as questions which aimed to ascertain the acceptability and safety of donor breastmilk and the willingness of the mothers to donate breastmilk and to set up a CBHMB. Questions also attempted to elicit how mothers thought the community would react to or accept the concept of mothers donating breastmilk and the fact that infants, including orphans, would be receiving DHM. Most of the questions were not open ended and the questionnaire was developed from scratch since there was no appropriate validated questionnaire available for use. The questionnaire was piloted with five mothers of similar background before it was finalised.

A convenience sample of 40 women was approached for permission to be enrolled into an evaluation of the effect of the training programme on perceptions around human milk banking. These were the last mothers to be enrolled into the peer counsellor training programme (March - May 2013). Only 38 of the women who had completed the training programme were available to complete the post-training questionnaire. Although Cato Manor Clinic serves a predominantly isiZulu-speaking community, ~80% of the women are able to speak English. The training programme was limited to women who could read and speak simple English since many of the training materials were in English. The discussion sessions during the training sessions were, however, conducted in isiZulu.

All mothers signed consent forms before participating in the study and the data were held confidentially. The study was approved by the Biomedical Research Ethics Committee of the University of KwaZulu-Natal (Ref: BE 210/12).

Results

Socioeconomic data

Data from the questionnaire show that most survey respondents were single, unemployed, had little education, and were living in homes with poor sanitation (Table 1).

Mothers’ perceptions

Survey questions about DHM focused on donor milk safety and personal and community views of donor milk use. Some significant changes in knowledge and perceptions were observed following the training. Initially, most mothers were hesitant to become involved in donating, feeding, or banking breastmilk. The data showed that post training, mothers’ perceptions had changed on most, but not all, topics (Fig. 1).

Willingness to donate breastmilk

Data collected before training showed that a narrow majority of mothers, 23 of 40 (58%), were willing to donate breastmilk. After training, this improved significantly (p<0.05) with most mothers, 33 of 38 (86.8%) being willing to be donors.

Willingness to feed own baby donated breastmilk

Initially, mothers largely rejected the possibility of feeding their own baby with milk donated by another mother, with only 12 of the 40 (30%) indicating willingness to do so. Following training, however, this increased significantly (p<0.05), with 28 of the 38 (73.7%) mothers prepared to feed their own baby donated breastmilk.

Acceptability of CBHMBs to the mothers themselves

Before training, some mothers, 23 of 40 (57.5%), considered the idea of them setting up CBHMBs acceptable. Following training this did 84

SAJCH

not change significantly (p=0.21), with 27 of the 38 (71.1%) finding the concept acceptable.

Community perceptions

This question revealed that mothers continued to hold the view that although their perceptions may have changed, the wider community perceptions around DHM remained largely unfavourable. In the initial questionnaire 24 of the 40 (60%) mothers felt the community would view feeding a baby donor milk as unacceptable. In the posteducation questionnaire the findings showed a non-significant difference (p=0.65), with 21 of the 38 (55.3%) mothers reporting this.

Similarly before training, 20 of the 40 (50%) mothers indicated that the community would perceive breastmilk donors as ‘not acceptable’ and after training, 21 of the 38 (55.3%) held this view (p=0.73).

Discussion

To the best of our knowledge this is the first survey measuring perceptions of mothers to CBHMBs, before and after education around HMBs. The findings of pre-training perceptions are similar to early findings from Nigeria, where the major concern expressed by mothers with regard to the feeding of donor breastmilk to their own baby was that it would not be safe.[13] Studies on the Table 1. Sociodemographic characteristics of respondents (N=40) Characteristic

n (%)

Age (years), mean (SD)

25.4 (5.6)

15 - 24

23 (57.5)

25 - 34

14 (35.0)

≥35

3 (7.5)

Marital status Single

26 (65.0)

Married

5 (12.5)

Living with partner

9 (22.5)

Education Primary education incomplete

2 (5.0)

Completed primary education

17 (42.5)

Completed secondary education

16 (40.0)

Some tertiary education

5 (12.5)

Employment Unemployed

30 (75.0)

Employed

2 (5.0)

On maternity leave

8 (20.0)

Dependent on social welfare grant Yes

12 (30.0)

No

28 (70.0)

Water Tap in dwelling

20 (50.0)

Tap in yard ≤10 m away

13 (32.5)

Tap in yard >10 m away

7 (17.5)

Toilet Flush

25 (62.5)

Pit latrine ≤10 m away

13 (32.5)

Pit latrine >10 m away

2 (5.0)

MARCH 2016 Vol. 10 No. 1

RESEARCH 100

80

%

60

40

20

0 Willingness to donate breastmilk for orphans

Willingness to feed own baby donor breastmilk

Acceptability of mothers setting up a small donor milk bank for orphans

Baseline percentage (n=40)

Percentage who presumed the community would not accept a donor breastmilkfed baby

Percentage who presumed the community would not accept a breastmilk donor

Post-training percentage (n=38)

Fig. 1. Mothers' perceptions before and after training.

acceptability of HMBs within hospital settings have shown similar findings. A 2013 qualitative study in South Australia focusing on a middle-class, well-educated group of mothers found that HMBs would be well received provided the safety of donor milk was well established.[14] The results of this study have corroborated the findings of earlier focus group discussions conducted in SA,[15] which reported that obstacles to acceptability of DHM arose from lack of education about DHM and banking practices. The results of this study showed that training improved mothers’ understanding of the uses and value of DHM and the processes used to ensure its safety. Before the training, some mothers showed willingness to donate milk to orphans and to set up HMBs in their own homes. Following training, this number increased, suggesting that training can complement and expand upon existing interest.

The results of this study therefore clearly demonstrate that education around breastfeeding can improve mothers’ perceptions about DHM; the role of donors; and the value of HMBs; including home-based HMBs or CBHMBs. Changing perceptions of DHM among this key target population is a critical first step which could in turn impact the potential of these women to serve as donors in the future.

The results also suggest that some concerns, particularly around acceptability in the community, may persist. This may be due to mothers’ fears that long-standing traditional myths associated with human milk would still be prevalent in the community. Although mothers’ perceptions had changed favourably, it was clear that the mothers felt that even though they had been trained this was unlikely to have a major effect on community perceptions around milk banking.

This highlights the importance, as we have seen with breastfeeding promotion, of community awareness campaigns targeting not only mothers but partners, grandparents and community leaders.[16] In SA this would be particularly important because of the high HIV prevalence among women of reproductive age and the perceived dangers associated with breastmilk. The prevalence of HIV-infected antenatal women in KZN was 37.4% in 2012, the highest in the country and higher than the national prevalence of 29.5%.[17] A limitation of this study was that only women who could speak and read simple English were included. This population may not be representative of all community members, especially those in more 85

SAJCH

rural, less educated settings where traditional views and myths could have an effect on perceptions of donor milk. A further limitation is the ability to generalise the results beyond this population, which was comprised of women with an interest in breastfeeding.

Conclusion

When a mother’s own milk is not available, DHM provides a lifesaving second chance for many vulnerable infants. Continued research, including attention to sustainable, context-appropriate options such as CBHMBs, is key to increasing access to donor milk for those who need it. Simplified, low-cost processes designed for resourcelimited settings are being explored for pasteurisation, and additional work in this area is needed.[18] Integrating HMB education within breastfeeding-promotion campaigns and ensuring rigorous qualitycontrol systems are crucial steps to establishing effective systems.[19] Trust and support varies widely among cultures and communities, so targeted education and engagement remain key. The results of this study demonstrate the critical role of education training to improve awareness and acceptability of this lifesaving intervention and suggest the importance of multifaceted awareness programmes that extend beyond the mother to the wider community.

Funding and conflicts of interest. The authors declare that they have no competing interests. The authors would like to acknowledge the funders who graciously supported this work, including the Bill and Melinda Gates Foundation, through the Grand Challenge Explorations Initiative; the University of Washington Department of Computer Science and Engineering; donations from private foundations and individuals to the PATH Health Innovation Portfolio; and the National Science Foundation Research Grant No. 11S-1111433.

Acknowledgements. The authors wish to thank all the mothers who participated in the survey and Gcinile Maphanga for assisting with the interviews. We would also like to acknowledge support from the eThekwini Health Unit and the Cato Manor Clinic for allowing us to conduct the study at the latter Clinic.

References

1. Duijt L, Jaddoe VW, Hofman A, Moll HA. Prolonged and exclusive breastfeeding reduces the risk of infectious diseases in infancy. Pediatrics 2010;126(1):e18-e25. [http://dx.doi.org/10.1542/peds.2008-3256]

MARCH 2016 Vol. 10 No. 1

RESEARCH 2. WHO Collaborative Study Team on the Role of Breastfeeding on the Prevention of Infant Mortality. Effect of breastfeeding on infant and child mortality due to infectious diseases in less developed countries: A pooled analysis. Lancet 2000;355(9202):451-455. [http://dx.doi.org/10.1016/S0140-6736(00)82011-5] 3. Jones G, Steketee RW, Black RE, Bhutta ZA, Morris SS; and the Bellagio Child Survival Study Group. How many child deaths can we prevent this year? Lancet 2003;362(9377):65-71. [http://dx.doi.org/10.1016/S0140-6736(03)13811-1] 4. Boyd CA, Quigley MA, Brocklehurst P. Donor breast milk versus infant formula for preterm infants: Systematic review and meta-analysis. Arch Dis Child Fetal Neonatal Ed 2007;92(3):F169-F175.[http://dx.doi.org/10.1136/ adc.2005.089490] 5. Arslanoglu S, Ziegler EE, Moro GE. World Association of Perinatal Medicine (WAPM) Working Group on Nutrition. Donor human milk in preterm infant feeding: Evidence and recommendations. J Perinat Med 2010;38(4):347-351. [http://dx.doi.org/10.1515/jpm.2010.064] 6. Imdad A, Yakoob MY, Bhutta ZA. Effect of breastfeeding promotion interventions on breastfeeding rates, with special focus on developing countries.

BMC Public Health 2011;11(Suppl 3):S24. [http://dx.doi.org/10.1186/14712458-11-s3-s24] 7. World Health Organization. Infant and Young Child Feeding: Model Chapter for Textbooks for Medical Students and Allied Health Professionals. Geneva; WHO, 2009.

8. South African Department of Health. The Tshwane Declaration for the Support of Breastfeeding in South Africa. S Afr J Clin Nutr 2011;24(4):214.

9. Statistics South Africa. Stastistical release Census 2011. http://www.statssa.gov. za/publications/P03014/P030142011.pdf (accessed 22 July 2014).

10. Coutsoudis I, Adhikari M, Nair N, Coutsoudis A. Feasibility and safety of setting up a donor breastmilk bank in a neonatal prem unit in a resource

86

SAJCH

limited setting: An observational, longitudinal cohort study. BMC Public Health 2011;11(1):356. [http://dx.doi.org/10.1186/1471-2458-11-356] 11. Geraghty SR, List BA, Morrow GB. Guidelines for establishing a donor human milk depot. J Hum Lact 2010;26(1):49-52. [http://dx.doi. org/10.1177/0890334409347467] 12. Osbaldiston R, Mingle LA. Characterization of human milk donors. J Hum Lact 2007;23(4):350-357. [http://dx.doi.org/10.1177/0890334407307547] 13. Ogala WN. Attitudes of nursing mothers to breast milk banking. Niger J Paediatr 1987;14(3-4):97-101.

14. Mackenzie C, Javanparast S, Newman L. Mothersâ&#x20AC;&#x2122; knowledge of and attitudes toward human milk banking in South Australia: A qualitative study. J Hum Lact 2013;29(2):222-229. [http://dx.doi.org/10.1177/0890334413481106] 15. Coutsoudis I, Petrites A, Coutsoudis A. Acceptability of donated breast milk in a resource limited South African setting. Int Breastfeed J 2011;6:3. [http:// dx.doi.org/10.1186/1746-4358-6-3] 16. Shealy KR, Ruowei L, Benton-Davis S, Grummer-Strawn LM. The CDC Guide to Breastfeeding Interventions. Atlanta: US Department of Health and Human Services, Centers for Disease Control and Prevention, 2005.

17. South African National Department of Health. The 2012 National Antenatal Sentinel HIV and Herpes Simplex Type-Prevalence Survey. Pretoria: National Department of Health, 2013.

18. Chaudhri R, Vlachos D, Kaza J, et al. 2011. A system for safe flash-heat pasteurization of human breast milk. In: Proceedings of the 5th ACM workshop on Networked systems for developing regions (NSDR 2011). New York: ACM, 2011:9-14. [http://dx.doi.org/10.1145/1999927.1999932] 19. PATH. Strengthening Human Milk Banking: A Global Implementation Framework. Version 1. Seattle: Bill and Melinda Gates Foundation Grand Challenges initiative, PATH, 2013.

MARCH 2016 Vol. 10 No. 1

RESEARCH

Prevalence and nature of communication delays in a South African primary healthcare context J van der Linde,1 MCommunication Pathology; D W Swanepoel,1,2 PhD; J Sommerville,3 MSc; F Glascoe,4 PhD; B Vinck,1,5 PhD; E M Louw,3 PhD Department of Speech-Language Pathology and Audiology, Faculty of Humanities, University of Pretoria, South Africa Ear Sciences Centre, School of Surgery, University of Western Australia, Nedlands, Australia; and Ear Science Institute Australia, Subiaco, Australia 3 Department of Statistics, Faculty of Natural and Agricultural Sciences, University of Pretoria, South Africa 4 School of Medicine, Vanderbilt University, Nashville, Tennessee, USA 5 Department of Speech-Language Pathology and Audiology, Ghent University, Belgium 1 2

Corresponding author: J van der Linde (jeannie.vanderlinde@up.ac.za) Background. Communication delays are the most common impairment in early childhood and have a negative effect on long-term academic, psychological and social development. Baseline prevalence of communication delays or disorders enables adequate planning of service delivery and successful implementation of intervention strategies, to reduce disorder prevalence.

Objective. To determine the prevalence and describe the nature of communication delays in infants aged 6 - 12 months in underserved communities in South Africa (SA).

Method. A parent interview and the Rossetti Infant-Toddler Language Scale (RITLS) were used to collect data from the caregivers of 201 infants aged 6 - 12 months by means of convenience sampling at primary healthcare facilities in the Tshwane district, SA. Results. Thirteen percent (n=26) of infants were diagnosed with communication delay. Associations affecting language delays were established for three risk factors (i.e. housing status, age of mother and number of siblings). The effect of combined risk factors on language development revealed that an infant was at greatest risk (27% probability) of developing a language delay when: (i) mothers were between the ages of 19 and 34 years; (ii) parents owned their own home; and (iii) there were three or more children in the household.

Conclusion. The prevalence of communication delays in the sample population was high, possibly because the majority of infants were exposed to risk factors. The implementation of preventive measures such as awareness campaigns and developmental screening and surveillance should be considered in the SA primary healthcare context.

S Afr J Child Health 2016;10(1):87-91. DOI:10.7196/SAJCH.2016.v10i1.1121

Communication delays are the most common impairment in early childhood[1] and have a negative effect on long-term academic, psychological and social development.[2,3] Early identification of, and early intervention for, communication delays in infants minimise the effect of the delay on educational and social outcomes.[2] Reported prevalence of communication delays varies significantly within countries and internationally.[2,4,5] In the UK a communication delay prevalence of 16.3% in children has been reported.[1] Similar findings (prevalence of 16.5% and 11.6%) were reported in school-going children in Sydney, Australia.[6] In contrast, other studies have reported much lower figures, such as 1.4%[7] and 8.0%.[8] More specifically, a systematic review reported that 15% of 2-year-olds presented with expressive language delays.[9,10] Possible reasons for the variability is the presence of risk factors, difficulty assessing infants and toddlers and the limited availability of well-developed assessment tools.[2,4] Risk factors such as poverty, lack of stable residence, limited prenatal care and inadequate healthcare facilities contribute to communication delays in infants.[11] People living in underserved communities, e.g. in informal settlements in South Africa (SA), experience a double burden of poverty and ill health as the environment they live in influences child development.[11] Residential density, living in crowded homes and poor-quality housing lead to parents being less interactive with their children, which in turn has a negative effect on communication development.[12] Also, a gender bias exists, with males more likely to present with communication delays than females.[1,4] Apart from risks, identification of communication delays in infants is difficult as development occurs over time, resulting in varied prevalence rates.[1,4] Most parents only discover their child’s communication delays when he or she fails to meet typical developmental milestones.[2] 87

SAJCH

Ironically, the most important period of communication acquisition and development is between 8 months and 2 years.[2] Studies reporting the prevalence of communication delays in infants younger than 2 years are limited.[2,4] This is problematic as prevalence rates vary across the ages of infants and young children.[1] Establishing the prevalence of communication delays or disorders enables appropriate planning for service delivery and successful implementation of intervention strategies, which may ultimately result in a decline in the prevalence of the disorder.[4] Despite previous efforts[13] to improve early identification of infants with delayed communication development in primary healthcare settings of SA, this practice remains uncommon.[14] Establishing the prevalence of communication delays in infants from underserved communities in SA will, however, advocate for the implementation of early identification and intervention services.

An adequate understanding of the prevalence and nature of comm­ unication delays in a specific population improves classification of communication delays.[1] Previous research has focused only on speech and receptive and expressive language delays and has not evaluated all the aspects of communication development such as pragmatics and interaction-attachment.[4] Most large-scale prevalence studies have used a broad classification of communication delays, and as a result the true nature of these delays has been obscured.[1] Understanding the nature of communication delays allows predictions that are of clinical and research significance, i.e. early use of gestures predicts later vocabulary development and early word use predicts later social-emotional development.[15] Since there is a dearth of information on the prevalence and nature of communication delays in infants 12 months and younger, the

MARCH 2016 Vol. 10 No. 1

RESEARCH

Methods

A prospective cross-sectional study was employed to determine the prevalence and nature of communication delays in infants from underserved SA communities. Prior to data collection, permission and ethical clearance was obtained from the Tshwane district research committee, Department of Health, and the Faculties of Health Sciences and Humanities, University of Pretoria.

Setting

Three clinics (Olievenhoutbosch clinic, Salvokop clinic and Daspoort Polyclinic), situated in underserved communities of the Tshwane district, Gauteng Province, SA, were utilised for data collection. Olievenhoutbosch clinic serves a population of 70 863 indi­ viduals residing in an area of 11.39 km2. [16] Both Salvokop and Daspoort form part of the Pretoria sub-district. The clinic situated in Salvokop area serves a population of 7 123 and Daspoort clinic a population of 6 355 indi­viduals.[16]

Participants

A total of 201 participants were included in the study by means of convenience sampling.

All the parents or caregivers of infants aged between 6 and 12 months, who were proficient in Afrikaans or English, were asked to participate during their visit to the primary healthcare clinic (PHC). Gender distribution of infants was similar (55% male). The home languages spoken most were Sepedi (33%), isiZulu (16%), chiShona (11%) and Ndebele (10%). Ninety-four percent of the participants resided in the Olievenhoutbosch area; the remainder were from other areas such as Salvokop (2%) and Mamelodi (0.5%). The majority of participants (98.5%) were black, with 1.5% from other ethnic groups.

Seven infants (of 201 participants) had been born to teenage mothers. Altogether 62% of parents or caregivers left the educational system at Grade 10 or less and 71% reported a household income of less than ZAR3 000 a month. About one-third (32%) of infants have two or more siblings. In general 17% of South Africans (20 years or older) are functionally illiterate, 34% completed some secondary levels of education and 29% completed Grade 12.[16] Furthermore, 46% of the SA population is deemed poor.[17,18]

The Rossetti Infant-Toddler Language Scale (RITLS) is a comprehensive, easy-to-administer and relevant tool to assess the preverbal and verbal communicative abilities and interaction in infants and young children.[19] Although this is a criterion-referenced tool, it has been used and validated in the past.[20-23] The tool assesses the following domains: pragmatics, gesture, play, language comprehension, language ex­ press­ion and interaction-attachment. When an infant has one (or more) unmet milestone(s) in a specific developmental domain (such as language expression) at a specific age interval, the milestones of the previous interval are evaluated until the infant has met all the milestones at that age interval. The infant’s developmental level is therefore categorised as the interval at which he/she obtained all the milestones within a domain. An infant’s development is classified as delayed when domain-specific developmental levels differed 6 months or more from the chronological age (e.g. when a 12-month-old infant’s language expression scored on a 3 - 6-month-old de­ velop­mental level).[19] As the first items in the gesture subdomain only start at 9 - 12 months, an infant can only present with a delay when he/she is 15 months or older. Since participants in the study were all between 6 and 12 months of age and their development of gestures could not be classified as delayed, this subdomain was excluded from the results.

Procedures

An experienced speech-language pathologist collected all the data. Parental/caregiver in­ formed consent was obtained before data collection commenced. First the parent inter­ view, and then the RITLS, was conducted on each participant. The RITLS was completed by observing and eliciting infant behaviour and also by making use of parental responses.

Data analysis

A statistical software program, SAS (version 9.3), was used to conduct the data analysis.

60

Results

Of the 201 participants, 13.0% (n=26) were diagnosed with communication delay, i.e. a delay in one or more of the communication domains of the RITLS. The delayed infants were evenly distributed in terms of gender (54% male). The majority of delayed in­ fants (58%) presented with a delay in one communication domain (Fig. 1).

58

50 40 27

30 20

12

10

Material

A structured interview schedule was de­ velop­ ed to obtain participant background information, i.e. date of birth, duration of pregnancy, and gender.

Descriptive statistics were used to describe the prevalence and nature of communication delays in a group of infants. To determine the existence of a significant association between risks and the delayed outcome of the receptive and/or expressive language domains of the RITLS, the χ2 and Fisher’s exact test statistics were used with a significance level of p≤0.05 and p≤0.1.

Risk factors significantly associated with receptive and/or expressive language delays (p≤0.05, p≤0.1) were included in the second phase of the statistical analysis where a log linear model analysis was used to model the probabilities of developing language delays, taking into account both single and simultaneous effects of the relevant risks.

Since only three factors were significant at 5% probability, a probability of 10% was used to add additional factors into the model. Maternal education was included as the fourth factor with a 10% probability (p=0.095). Since the data on the age of the mother were too limited in the category ≤18 years (n=7), this category had to be excluded in the log linear analysis.

Although a maternal age of 19 - 34 years is not considered an environmental risk, the effect of age for mothers aged ≥35 had to be explored alongside the low-risk group.

The outcomes of the model were expressed as indices and converted into odds of language delays for a specific combination of categories of risk factors. Based on the odds, the estimated probability to have a language (receptive and/ or expressive) delay for a specific combination of risks was calculated using the following formula: prob=odds/(1+odds).

70

%

objective of this study was to determine the prevalence and nature of communication delays in infants aged 6 - 12 months in under­served communities in SA.

0

0 1

2

3

4

4 5

Delayed communication domains, n Fig. 1. Distribution of the number of delayed communication domains in participants with delays (n=26).

88

SAJCH

MARCH 2016 Vol. 10 No. 1

RESEARCH only two participants (2/26; 7.7%) had a delay in their interaction attachment skills.

Associations between risks and delays in language expression and/or comprehension are presented in Table 2. Three risks were found to be significantly associated with language delays in the study population, namely: • Infants of mothers with three or more child­ ren had significantly higher prevalence of

Only 4% of infants with a positive diagnosis presented with delays in five of the domains.

The prevalence rates for the domain-specific outcomes of the RITLS are presented in Table 1. Most participants with a positive diagnosis (22/26; 84.6%) presented with delayed language expression. Nine of the delayed participants (9/26; 34.6%) presented with delayed language comprehension and

Table 1. Domain-specific outcomes of the RITLS for all participants (N=201) Domains

No delay, n (%)

Delay, n (%)

Pragmatics

197 (98.0)

4 (2.0)

Play

195 (97.0)

6 (3.0)

Interaction-attachment

199 (99.0)

2 (1.0)

Language expression

179 (89.1)

22 (10.9)

Language comprehension

192 (95.5)

9 (4.5)

Table 2. Association of language delays with risk factors Delayed (%)

Significance (p-value)

Test statistic

Grade 10 or less (n=66)

17

0.095*

χ2

Grade 11 - 12, and/or tertiary education (n=134)

10

0.0564†

χ2

0.5385

Fisher’s exact

0.7524

Fisher’s exact

0.0241†

χ2

0.5834

χ2

0.3097

χ2

0.0357†

Fisher’s exact

Risk factors Level of education (n=200)

Number of children (n=201) ≤2 (n=135)

9

≥3 (n=66)

18

Prematurity (n=201) 0 - 2 months premature (n=195)

12

≥3 months premature (n=6)

17

Employment (n=201) Yes (n=173)

12

No (n=28)

14

Housing status (n=201) Home owners (n=47)

21

Informal housing or staying with others (n=154)

10

Gender (n=201) Male (n=111)

11

Female (n=90)

13

Average household income (n=199) <ZAR1 500 (n=80)

9

≥ZAR1 500 (n=119)

13

Age of mother at birth of youngest infant (years) (n=199) ≤18 (n=7)

43

19 - 34 (n=165)

10

≥35 (n=27)

15

*Statistically significant association (p≤0.1). †

Statistically significant association (p≤0.05).

89

SAJCH

MARCH 2016 Vol. 10 No. 1

language delays (sample percentage of 18%) than infants of mothers who have less than three children (9%; χ2, p=0.054).

• Living in informal housing or staying with others showed a significantly lower prevalence in language delays (10%) compared with living with caregiver in their own house (21%; χ2, p=0.024).

• Language delays in infants born to mothers who were 18 years or younger (43%) and mothers aged 35 years or older (15%) were significantly higher than those born to mothers between the ages of 19 and 34 years (10%; Fisher’s exact test, p=0.035).

The outcome of the log linear analysis is shown in Table 3 with the four risk factors with the strongest association with language delay presented as combined risk factors.

The indices were used to calculate the probabilities of combined risk factors by multiplying the overall mean effect (value of the intercept of the log linear model) with the index of the combination of categories of risk factors under consideration.

A probability of 21% was associated with language delay in infants with two or more siblings, born to a mother aged 19 - 34 years with limited education who lives in informal housing or with others. In contrast, infants with two or more siblings born to a mother aged 19 - 34 years with a Grade 11 - 12 and/or tertiary education living in informal housing or with others, only had a 10% risk to present with a language delay (Table 3).

Discussion

Few studies have reported prevalence of comm­ uni­ cation disorders under the age of 2 years. [1,4] The prevalence (13.4%) of communication disorders in the sample population is higher than the prevalence rates (5.6%) reported in previous research conducted in the UK for infants aged 0 2 years.[1] A median prevalence of 5% for speech and language delays in 2-year-olds was also reported in a systematic review conducted in 2000.[4] A study conducted in the UK reported that 20% of referrals (during 1999 - 2000) of children of all ages were for receptive language difficulties, 17% for expressive language difficulties and 29% for speech difficulties.[1] Of the 20% of children with receptive language delays 6% were aged between 0 and 2 years, and of the 17% of all children with expressive language delays 13% were from the same cohort.[1] Similar results were yielded by two other studies, one conducted in the USA, and the other a systematic review where 15% of 2-yearolds presented with expressive language delays.[9,10] Of the current sample population (N=201) 11% of infants, aged 6 - 12 months, presented with delays in expressive language,

RESEARCH Table 3. Associated probability of combined risk factors predisposing language delay Parameter

Combination of categories

Index

Overall mean effect Age of mother, number of children, education level, housing status

Odds

p-value (%)

0.14 (intercept) 19 - 34 years, <3 children, Grade 11/12 and/or tertiary education, informal housing/staying with others

0.42

0.059

0.056 (6)

19 - 34 years, <3 children, Grade 10 or less, informal housing/ staying with others

0.27

0.038

0.037 (4)

19 - 34 years, <3 children, all education levels, home owners

1.46

0.204

0.169 (17)

19 - 34 years, ≥3 children, Grade 10 or less, informal housing/ staying with others

1.89

0.264

0.209 (21)

19 - 34 years, ≥3 children, Grade 11/12 and/or tertiary education, informal housing/staying with others

0.82

0.115

0.103 (10)

19 - 34 years, ≥3 children, all education levels, home owners

2.60

0.364

0.267 (27)

≥35 years, any number of children, all education levels, informal housing/staying with others

0.99

0.139

0.122 (12)

≥35 years, any number of children, all education levels, home owners

1.54

0.216

0.178 (18)

which is similar to previous research findings for a slightly older cohort.[9,10] High prevalence rates for communication and more specifically language delays, reported in the current study, may be ascribed to multiple risks present in the target population that may influence their communication development. Language delay was significantly associated with three risk factors: housing status (p=0.0241), age of the mother (p=0.035) and number of children in the home (p=0.054).

Infants with parents who are home owners were more at risk of language delays than those who stayed with others or in informal housing. The diversity of neighbourhoods in which infants live shape their social learning independent of their caregiver and/or family interaction.[24] The diverse neighbourhood of informal settlements or living with others appears to aid language development in infants.

Consequently what was deemed a risk factor in the past[11] may facilitate more opportunities for communication interactions, and may be conducive to social language learning. Investigation into this complex interaction is however needed.

The effect of combined risk factors on language development revealed that infants in the current study were at greatest risk (27% probability) of developing a language delay when: (i) mothers were between the ages of 19 and 34 years; (ii) parents own their own home; and (iii) there are three or more children in the household.

Although high rates of spontaneous resolution of language delays have been reported in the past,[2,25] association between language outcomes of children with delayed expressive language onset has been established.[5,26] Making a definitive diagnosis of a socialcommunication delay is difficult at young ages.[27] Nevertheless the most important phase of communication acquisition and development takes place between 8 and 24 months.[2,28] As a result, early detection of developmental risks is important regardless of the final diagnosis, especially since a variety of developmental problems can lead to language delays.[27] After the identification of risk factors, collaboration among primary healthcare workers, social services and community early intervention providers is crucial.[29] Clinicians should regularly advise on, and make parents aware of the value of talking frequently with their children, modelling and expanding their child’s utterances and actively teaching new words.[29] However, challenged families, who are exposed to multiple risks, may not respond well to brief advice.[29] Collaboration with community healthcare workers as part of community-orientated primary healthcare in SA may improve 90

SAJCH

responsiveness of these families as continued support will be provided.[30] Future research should evaluate the implementation of preventive measures such as awareness campaigns and develop­ mental screening and surveillance as part of the communityorientated primary healthcare initiative.

Conclusion

In the current study 13% of infants between 6 and 12 months from an underserved primary healthcare context presented with communication delays. Specifically expressive language delays were most commonly detected in these infants. Association between three risk factors (i.e. housing status, age of mother and number of siblings) and language delays was established for this age cohort of infants. Furthermore, the probability of language delay when exposed to these risks in combination has demonstrated that infants were at greatest risk when mothers were between the ages of 19 and 34 years, when the parents own their own home and when there are three or more children in the household. Since many infants are exposed to these risks in SA the implementation of preventive measures such as awareness campaigns and developmental screening and surveillance should be prioritised.

Acknowledgements. The authors would like to acknowledge the Mellon Foundation for funding the vulnerable children programme of the Faculty of Humanities, University of Pretoria, the COPC living laboratory of the University of Pretoria, the National Research Foundation, and Department of Statistics, University of Pretoria for their assistance in the data processing and analysis. We would also like to acknowledge the Vice Chancellor’s Academic Development Grant for funding the research. Mrs Ahmed and Mrs Bogatsu are also acknowledged for their contribution.

References 1. Broomfield J, Dodd B. Children with speech and language disability: Caseload characteristics. Int J Lang Commun Disord 2004;39(3):303-324. [http://dx.doi. org/10.1080/13682820310001625589] 2. Eadie PA, Ukoumunne O, Skeat J, et al. Assessing early communication behaviours: Structure and validity of the Communication and Symbolic Behaviour Scales-Developmental Profile (CSBS-DP) in 12-month-old infants.

Int J Lang Commun Disord R Coll Speech Lang Ther 2010;45(5):572-585. [http://dx.doi.org/10.3109/13682820903277944] 3. Young AR, Beitchman JH, Johnson C, et al. Young adult academic outcomes in a longitudinal sample of early identified language impaired and control children. J Child Psychol Psychiatry 2002;43(5):635-645. [http://dx.doi. org/10.1111/1469-7610.00052]

MARCH 2016 Vol. 10 No. 1

RESEARCH 4. Law J, Boyle J, Harris F, Harkness A, Nye C. Prevalence and natural history of primary speech and language delay: Findings from a systematic review of the literature. Int J Lang Commun Disord 2000;35(2):165-188. [http://dx.doi. org/10.1080/136828200247133] 5. Hawa VV, Spanoudis G. Toddlers with delayed expressive language: An overview of the characteristics, risk factors and language outcomes. Res Dev Disabil 2014;35(2):400-407. [http://dx.doi.org/10.1016/j.ridd.2013.10.027] 6. McLeod S, McKinnon DH. Prevalence of communication disorders compared with other learning needs in 14 500 primary and secondary school students. Int J Lang Commun Disord 2007;42(Suppl 1):37-59. [http://dx.doi. org/10.1080/13682820601173262] 7. Paul TJ, Desai P, Thorburn MJ. The prevalence of childhood disability and related medical diagnoses in Clarendon, Jamaica. West Indian Med J 1992;41(1):8-11.

8. Harasty J, Reed VA. The prevalence of speech and language impairment in two Sydney metropolitan schools. Aust J Hum Commun Disord 1994;22(1):1-23. [http://dx.doi.org/10.3109/asl2.1994.22.issue-1.01] 9. Samuels A, Slemming W, Balton S. Early childhood intervention in South Africa in relation to the developmental systems model: Infants Young Child 2012;25(4):334-345. [http://dx.doi.org/10.1097/IYC.0b013e3182673e12] 10. Evans GW. Child development and the physical environment. Annu Rev Psychol 2006;57(1):423-451. [http://dx.doi.org/10.1146/annurev. psych.57.102904.190057] 11. Van der Linde J, Kritzinger A, Redelinghuys A. The identification process in early communication intervention (ECI) by primary health care personnel in Ditsobotla sub-district. S Afr J Commun Disord 2009;56:48-59.

12. Kathard H, Pillay M. Promoting change through political consciousness: A South African speech-language pathology response to the World Report on Disability. Int J Speech Lang Pathol 2013;15(1):84-89. [http://dx.doi.org/10.3 109/17549507.2012.757803] 13. Greenwood CR, Buzhardt J, Walker D, McCune L, Howard W. Advancing the construct validity of the Early Communication Indicator (ECI) for infants and toddlers: Equivalence of growth trajectories across two early head start samples. Early Child Res Q 2013;28(4):743-758. [http://dx.doi.org/10.1016/j. ecresq.2013.07.002] 14. Statistics South Africa. Census 2011. https://www.statssa.gov.za/Census2011/ default.asp (accessed 20 February 2014).

15. Statistics South Africa. Poverty Trends in South Africa: An examination of absolute poverty between 2006 and 2011. Report No.: 03-10-06. http://beta2. statssa.gov.za/publications/Report-03-10-06/Report-03-10-06March2014.pdf (accessed 5 May 2014).

16. Mayosi BM, Benatar SR. Health and health care in South Africa - 20 years after Mandela. N Engl J Med 2014;371(14):1344-1353. [http://dx.doi.org/10.1056/ NEJMsr1405012] 17. Rossetti L. The Rossetti Infant-Toddler Language Scale: A Measure of Communication and Interaction: Examiner’s Manual. Austin: Linguisystems, 2006.

91

SAJCH

18. Desmarais C, Sylvestre A, Meyer F, Bairati I, Rouleau N. Three profiles of language abilities in toddlers with an expressive vocabulary delay: Variations on a theme. J Speech Lang Hear Res 2010;53(3):699. [http://dx.doi. org/10.1044/1092-4388(2009/07-0245)] 19. Sylvestre A, Mérette C. Language delay in severely neglected children: A cumulative or specific effect of risk factors? Child Abuse Negl 2010;34(6):414428. [http://dx.doi.org/10.1016/j.chiabu.2009.10.003] 20. Dettman SJ, Pinder D, Briggs RJS, Dowell RC, Leigh JR. Communication development in children who receive the cochlear implant younger than 12 months: Risks versus benefits. Ear Hear 2007;28(Suppl):11S-18S. [http://dx.doi. org/10.1097/AUD.0b013e31803153f8] 21. Steiner AM, Goldsmith TR, Snow AV, Chawarska K. Practitioner’s guide to assessment of autism spectrum disorders in infants and toddlers. J Autism Dev Disord 2012;42(6):1183-1196. [http://dx.doi.org/10.1007/s10803-011-1376-9] 22. Desmarais C, Sylvestre A, Meyer F, Bairati I, Rouleau N. Systematic review of the literature on characteristics of late-talking toddlers. Int J Lang Commun Disord R Coll Speech Lang Ther 2008;43(4):361-389. [http://dx.doi. org/10.1080/13682820701546854] 23. Horwitz SM, Irwin JR, Briggs-Gowan MJ, Bosson Heenan JM, Mendoza J, Carter AS. Language delay in a community cohort of young children. J Am Acad Child Adolesc Psychiatry 2003;42(8):932-940. [http://dx.doi.org/10.1097/01.

CHI.0000046889.27264.5E] 24. Howard LH, Carrazza C, Woodward AL. Neighborhood linguistic diversity predicts infants’ social learning. Cognition 2014;133(2):474-479. [http://dx.doi. org/10.1016/j.cognition.2014.08.002] 25. Reilly S, Wake M, Bavin EL, et al. Predicting language at 2 years of age: A prospective community study. Pediatrics 2007;120(6):e1441-e1449. [http:// dx.doi.org/10.1542/peds.2007-0045] 26. Rice ML, Taylor CL, Zubrick SR. Language outcomes of 7-year-old children with or without a history of late language emergence at 24 months. J Speech Lang Hear Res 2008;51(2):394-407. [http://dx.doi.org/10.1044/10924388(2008/029)] 27. Ben-Sasson A, Habib S, Tirosh E. Feasibility and validity of early screening for identifying infants with poor social-communication development in a well-baby clinic system. J Pediatr Nurs 2014;29(3):238-247. [http://dx.doi. org/10.1016/j.pedn.2013.11.001] 28. Reilly S, Eadie P, Bavin EL, et al. Growth of infant communication between 8 and 12 months: A population study. J Paediatr Child Health 2006;42(12):764770. [http://dx.doi.org/10.1111/j.1440-1754.2006.00974.x] 29. Glascoe FP, Leew S. Parenting behaviors, perceptions, and psychosocial risk: Impacts on young children’s development. Pediatrics 2010;125:313-319. [http:// dx.doi.org/10.1542/peds.2008-3129] 30. Bam N, Marcus T, Hugo J, Kinkel H-F. Conceptualizing Community Oriented Primary Care (COPC) – the Tshwane, South Africa, health post model. Afr J Prim Health Care Fam Med 2013;5(1):1-3. [http://dx.doi.org/10.4102/phcfm. v5i1.423]

MARCH 2016 Vol. 10 No. 1

RESEARCH

An analysis of national data on care-seeking behaviour by parents of children with suspected pneumonia in Nigeria M B Abdulkadir,1 MBBS, FWACP (Paed); Z A Abdulkadir,2 MBBS, MWACS; W B R Johnson,1 MBBS, FWACP (Paed) 1 2

Department of Paediatrics and Child Health, University of Ilorin/University of Ilorin Teaching Hospital, Ilorin, Kwara State, Nigeria Department of Obstetrics and Gynaecology, University of Ilorin Teaching Hospital, Ilorin, Kwara State, Nigeria

Corresponding author: M B Abdulkadir (docmohng@gmail.com) Background. Pneumonia is responsible for 940 000 under-5 deaths annually. Most of these deaths result from delays in instituting effective treatment.

Objectives. To determine care-seeking behaviour by parents of children with pneumonia and sociodemographic factors that influence decisions to seek appropriate care.

Methods. The study was an analysis of the Nigeria Demographic and Health Survey 2013, which was a nationwide cross-sectional survey using a stratified cluster design of 40 680 households. All children under-5 living in the surveyed households with suspected pneumonia in the preceding 2 weeks were recruited along with their mothers. Sociodemographic characteristics of the parents and where they sought care for their child were obtained. Binomial logistic regression analysis was used to determine the contribution of various sociodemographic variables to the decision on seeking appropriate medical care.

Results. Of the 28 950 children surveyed, 565 had suspected pneumonia, which equates to an occurrence rate of 19.5 per 1 000 children. About 36% of parents were judged to have sought appropriate care when their children had pneumonia. High paternal education, health decision-making by both husband and wife, and belonging to the higher quintiles on a wealth index were factors that positively influenced care-seeking behaviour.

Conclusion. Care-seeking behaviour for pneumonia is poor. Paternal education and joint decision-making are key determinants of parents seeking appropriate care for their children with suspected pneumonia in Nigeria.

S Afr J Child Health 2016;10(1):92-95. DOI:10.7196/SAJCH.2016.v10i1.1076

Acute respiratory infections are a leading cause of death in children <5 years old, responsible for ~940 000 (15%) of all deaths in 2013. [1] There has been a 44% decline in pneumonia-related deaths since 2000. [1] This decline has been more pronounced in developed countries and south Asia compared with sub-Saharan Africa. [2,3] About half the world’s pneumonia-related deaths occur in Nigeria, the Democratic Republic of the Congo, Ethiopia, India and Pakistan.[2] Several strategies have been developed globally to address pneumonia-related mortality.

These strategies have been encapsulated in the Global Action Plan for the Prevention and Control of Pneumonia (GAPP).[4,5] Early diagnosis and prompt treatment with appropriate antibiotics have been critical in addressing such mortality.[2] This may be deployed as community or facility-based interventions. The case management approach formulated by the World Health Organization (WHO) has as its main foundations: appropriate use and choice of antibiotics against the major causes of bacterial pneumonia (pneumococcus and Hib); training health workers to utilise simple clinical signs to assess severity and guide appropriate treatment; and appropriate and effective use of oxygen.[5] Launched in 2009, GAPP recognised the importance of timely treatment and set an ambitious target of ensuring that 90% of children with pneumonia have access to appropriate pneumonia case management by 2015.[6] Unfortunately, in most sub-Saharan African countries, parents often seek healthcare from sources that lack the requisite knowledge, skills or facilities to provide appropriate care.[7-9] This has a significant effect on pneumonia morbidity and mortality.[10] Studies have shown that the care-seeking behaviour of parents is influenced by household size, age, education of the mother, past experiences, traditional beliefs, cost and location of health facilities.[7,11-13] West and Central African regional data have demonstrated that appropriate care-seeking behaviour for pneumonia has improved from 34 to 39% between 2000 and 2013. [10] 92

SAJCH

Despite this modest increase, there have been wide disparities within regions, countries and individual states.[10] It thus becomes crucial to generate local data regarding appropriate care-seeking behaviour to guide researchers and policy makers on effective pneumonia case management. The study aimed to determine care-seeking behaviours of parents of children with a symptom complex of pneumonia and factors that may influence these behaviours.

Methods

The study utilised the publicly available dataset of the Nigeria Demo­ graphic and Health Survey (NDHS) 2013, conducted by the National Population Commission (NPC). Permission was obtained for the use of the data from the Demographic and Health Surveys (DHS) Programme and Inner City Fund International. A detailed description of the study design, participants and study instruments has been previously published by the NPC.[14] Ethical approval for the survey was obtained from the National Health Research Ethics Committee.

Informed consent was obtained from study participants before they were allowed to participate in the survey. The data contained no information that could be used to trace participants.

A nationwide cross-sectional survey was carried out and a stratified three-stage cluster design was adopted.[14] The first stage involved the selection of 893 localities with probability of selection being proportional to size. At least one enumeration area (defined from the last population census) was randomly selected from each of the localities. This resulted in 904 enumeration areas being selected. In a few of the larger localities, more than one enumeration area was selected.[14] A household listing was made for each enumeration area. Forty-five households were selected in urban and rural clusters using systematic probability sampling with the sampling frame being the list of households in each enumeration area.[14]

MARCH 2016 Vol. 10 No. 1

RESEARCH A total of 40 680 households were surveyed for the demographic and health survey. Trained interviewers administered a questionnaire to each household with questions pertaining to household characteristics, health of children and care-seeking behaviour.[14] Children aged 0 59 months who had a cough and chest-related fast or difficult breathing in the preceding two weeks were recruited into the study alongside their parents. These symptoms constituted ‘suspected pneumonia’ as defined by the WHO’s Integrated Management of Childhood Illnesses.[2,15] Appropriate care-seeking was defined as seeking care from all public or private hospitals, health centres or posts, private doctors and community health workers.

It excludes no medical care, pharmacies, shops and traditional practitioners.[2,10] Data analysis was carried out using SPSS version 20 software (IBM, USA). Frequencies of missing variables were indicated, where present, if they constituted more than 1% of the overall proportion of the variable. However, these missing variables were excluded from χ2 and regression analyses as recommended by the DHS Programme.[16] Data were weighted using sampling weights provided within the database. Sampling weights are adjustment factors applied to each case in tabulations to adjust for differences in probability of selection and interview between cases in a sample, either due to design or chance.[16] Often in DHS the samples are selected to ensure the maximum representation of certain variables, such as urban/rural setting, which introduces bias into the sample selection process. To eliminate this bias, the frequency of occurrence of each event is weighted to indicate the number of individuals that should have been selected had bias not been present in the sample.[16] Wealth index was generated by principal component analysis of household assets. The sum of scores obtained for each household, standardised to a normal distribution and appropriate cut-offs, was used to define poorest, poorer, middle, higher and highest quintiles.[17] Frequency distribution tables and cross-tabulation of variables were generated.

Mean and standard deviation for continuous variables were provided and the proportion for categorical variables was also determined.

Basic tests of statistical significance such as χ2 and Student t-tests were utilised as required.

Maternal, household, socioeconomic, education and child factors were compared individually to care-seeking categories through univariate analysis. Significant factors were built into a multiple logistic regression model that was used to identify significant predictors of appropriate care-seeking behaviour.

Unweighted data were used for the regression analysis as recommended by the DHS Programme.[16] Odds ratio and p-value were used to determine significance of associations.

P<0.05 was considered significant.

Table 1. Univariate analysis of weighted numbers and percentages of selected determinants of appropriate care-seeking behaviour Variables

Appropriate, n (%)*

Inappropriate, n (%)*

p-value

Age of child (months), mean (SD)

24.82 (14.78)

22.62 (15.39)

0.110

Age of mother (years), mean (SD)

28.88 (7.02)

28.81 (7.08)

0.920

Age of household head (years), mean (SD)

39.70 (10.36)

38.75 (10.14)

0.310

Number of children under five in household

2.20 (1.33)

2.32 (1.25)

0.290 0.360

n=194

n=350

Male

101 (52.1)

168 (48.0)

Female

93 (47.9)

182 (52.0)

n=195

n=350

Urban

72 (36.9)

77 (22.0)

Rural

123 (63.1)

273 (78.0)

n=195

n=349

None

86 (44.1)

183 (52.4)

Primary

32 (16.4)

81 (23.2)

Secondary

64 (32.8)

79 (22.6)

Tertiary

13 (6.7)

6 (1.7)

Duration of mother education (years), mean (SD)

5.32 (5.52)

3.57 (4.41)

0.0001†

Highest paternal educational level

n=187

n=328

0.0001†

None

68 (36.4)

149 (45.4)

Primary

29 (15.5)

72 (22.0)

Secondary

57 (30.5)

90 (27.4)

Tertiary

33 (17.6)

17 (5.2)

Sex of child

Place of residence

Highest maternal educational level

0.0001†

0.001†

* Weighted data to account for probability of selection. May not add up to total numbers seeking appropriate or inappropriate care because of non-response. † Statistically significant.

Results

There were 28 950 children who were younger than 60 months of age in the 40 680 households surveyed. Of these, 14 509 (50.1%) were males and 14 440 (49.9%) were females. The mean (standard deviation (SD)) age of the children was 28.1 (17.3) months with a range of 0 - 59 months.

Of the 28 950 under-5 children, 565 (2.0%; 95% confidence interval (CI) 1.8 - 2.1) fulfilled the criteria for suspected pneumonia in the preceding 2 weeks. This meant the pneumonia prevalence rate was 19.5 per 1 000 children.

Among the children with suspected pneumonia, there were 279 (49.4%) males. The mean (SD) age of children with suspected pneumonia was 23.5 (15.2) months with a range of 0 - 59 months.

Regarding the age distribution of the children with pneumonia, of the total number of children: 137 (24.3%) were aged between 0 and 11 months; 183 (32.4%) were aged between 12 and 23 months; 115 (20.4%) were aged between 24 and 35 months; 57 (10.0%) were aged between 36 and 47 months; and 64 (11.3%) were aged between 48 and 59 months.

There were 544 valid responses to care seeking for the children (21 cases had missing records regarding care for the child and were 93

SAJCH

MARCH 2016 Vol. 10 No. 1

excluded from subsequent analyses of careseeking behaviour). A total of 194 (35.7%) parents sought appropriate care when their child had suspected pneumonia. Selected sociodemographic factors were compared with care-seeking behaviour for pneumonia in a univariate analysis using weighted data to account for probability of sampling (Tables 1 and 2). Factors that were significantly associated with care-seeking behaviour for suspected pneumonia were: living in an urban setting (p=0.001); maternal educational level (p=0.001); total duration of maternal education (p=0.001); increasing paternal educational level (p=0.001); wealth index (p=0.001); and who the decision maker was regarding the mother’s health (p=0.014) (Tables 1 and 2).

Binomial logistic regression was used to identify predictors of appropriate careseeking for suspected pneumonia using the variables listed above that were significant on a univariate analysis. Tertiary level education for the husband/partner (odds ratio (OR) 2.16, 95% CI 1.03 - 4.52; p=0.040) increased the odds of seeking appropriate care by 116% (p=0.040).

Also belonging to the middle quintile on the wealth index (OR 2.06, 95% CI 1.12 - 3.80;

RESEARCH Table 2. Univariate analysis of weighted numbers and percentages of selected determinants of appropriate care-seeking behaviour Appropriate, n (%)*

Variables

Inappropriate, n (%)*

p-value 0.230

n=193

n=346

Islam

126 (65.3)

203 (58.7)

Christianity

63 (32.6)

138 (39.9)

Traditionalist

4 (2.1)

5 (1.4)

n=192

n=350

Mother’s religion

Covered by health insurance Yes

4 (2.1)

2 (0.6)

No

188 (97.9)

348 (99.4)

n=195

n=351

38 (19.5)

98 (27.9)

Household wealth index quintiles Poorest Poorer

52 (26.7)

126 (35.9)

Middle

46 (23.6)

66 (18.8)

Richer

27 (13.8)

44 (12.5)

Richest

32 (16.4)

17 (4.8)

n=185

n=326

Decision maker regarding health of mother Mother alone

8 (4.3)

27 (8.3)

Mother and husband/partner

65 (35.1)

81 (24.8)

Husband/partner alone

112 (60.5)

216 (66.3)

0 (0.0)

2 (0.6)

Others Current marital status

n=194

n=350

Never married

2 (1.0)

13 (3.7)

Married

183 (94.3)

318 (90.9)

Living with partner

2 (1.0)

8 (2.3)

Widowed/separated/divorced

7 (3.6)

11 (3.1)

0.190

0.0001†

0.032†

0.355

* Weighted data to account for probability of selection. May not add up to total numbers seeking appropriate or inappropriate care because of non-response. †

Statistically significant.

p=0.020) and families where decision-making regarding the health of the mother was made jointly by husband/partner and mother (OR 3.64, 95% CI 1.54 - 8.61; p=0.003) are all associated with increased odds of appropriate care seeking for suspected pneumonia. Other parameters failed to achieve significance on the regression model (Tables 3 and 4).

Discussion

Pneumonia remains the leading infectious cause of death among children under-5, killing ~2 600 children a day.[10] Poor or delayed care accounts for up to 70% of child deaths in developing countries.[18,19] The occurrence rate for suspected pneumonia among children under 5 years of age in the current study was 19.5 per 1 000 children (95% CI 18.0 - 21.1).

Appropriate case management is a major focus of both the 2009 GAPP and the subsequent integrated Global Action Plan for Pneumonia and Diarrhoea.[6,20] Children with pneumonia should receive a full course of effective antibiotics because most severe cases

have a bacterial cause.[2,15,21] Timely institution of appropriate treatment is dependent on early recognition of the symptomatology by parents, decision to seek care from appropriate sources, availability, accessibility and affordability of the necessary therapy.[2] Studies have shown that parents are able to recognise the common symptoms and signs of potentially severe ill health, but that this does not always translate into appropriate health-seeking action.[18,22,23] In the current study, care-seeking behaviour was poor with only 35.7% of parents of children with suspected pneumonia seeking care from appropriate sources. This is much lower than the sub-Saharan African average of 48% and the average for developing countries of 60%. [2] It is, however slightly higher than the 32% reported in NDHS 2008.[14] It suggests that care-seeking behaviour for pneumonia in Nigeria has remained almost the same over the 5-year period from 2008 to 2013. This is in contrast to reports from most other countries that have shown significant improvement in care-seeking behaviour.[2,10] 94

SAJCH

MARCH 2016 Vol. 10 No. 1

The decision-making process regarding med­ ical care for ill health by parents is complex and often influenced by several factors.[7,10-13] Cultural factors, parental education, previous experience, knowledge and socioeconomic status often influence health-related decisionmaking in develop­ing countries.[7,8,11,13,18] These factors would contribute in differing magni­ tudes depend­ing on location and culture. The educational level of the husband or partner contributed significantly to appropriate care seeking in this study. A significantly higher proportion (48%) of male partners who had sought for appropriate care had secondary school education or more compared with only 32.6% of those who had inappropriate care.

While maternal education was significant on a univariate analysis but did not reach statistical significance on the regression analysis, a cursory look at the data suggests it may still play a role. Possible interactions between maternal and paternal education may have eliminated the role of the mother’s education in the regression analysis. Also, considering the husband or partner was responsible for decision-making (alone or in combination with the mother) regarding the health of the mother in 92.8% of the respondents, it is not surprising that maternal education did not significantly influence appropriate care seeking. This is further strengthened by the finding in this study that where decisions regarding health of the mother are taken by both partners, the odds of seeking appropriate care for the child with suspected pneumonia increase almost three fold. Taffa and Chepngeno[9] demonstrated that maternal education played an important role in determining appropriate care-seeking behaviour for children with diarrhoea, pneumonia and malaria in Kenya.

Therefore, we postulate that unlike the findings in our study, maternal education may influence appropriate health decisions in societies where women play a key role in decision-making in the household. The finding that parents in the higher wealth quintile were more likely to seek appropriate care independent of education suggests that the cost of medical care plays a significant role in the decision-making process.

This is especially important as the majority of parents in the study relied on out of pocket expenses for care. Obviously parents in the higher wealth quintiles are able to afford these expenses readily and as such would more appropriately seek proper care. Similar findings regarding cost and willingness/ability to pay for appropriate care have been reported by other authors.[24]

Study limitation

A limitation of the study was the use of suspected pneumonia as the factor defining the need for care seeking. Caution must be taken in the use of suspected pneumonia as a defining factor for determining true burden of pneumonia as the

RESEARCH Table 3. Determinants of appropriate care seeking for children with suspected pneumonia using unweighted data

Table 4. Determinants of appropriate care seeking for children with suspected pneumonia using unweighted data

Variable

OR (95% CI)*

p-value

Variable

Mean (SD) duration of education (yr)

1.11 (0.84 - 1.48)

0.47

Household wealth index quintiles

Place of residence Urban

1.00

Rural

1.50 (0.88 - 2.57)

0.136

Highest maternal educational level None

1.00

Incomplete primary

0.47 (0.14 - 1.60)

0.23

Complete primary

0.50 (0.08 - 2.99)

0.45

Incomplete secondary

0.30 (0.02 - 3.95)

0.36

Complete secondary

0.59 (0.02 - 18.08)

0.76

Tertiary

0.36 (0.00 - 29.87)

0.65

Poorest

1.00

Poorer

1.14 (0.67 - 1.93)

0.63

Middle

2.06 (1.12 - 3.80)†

0.02†

Richer

1.48 (0.67 - 3.26)

0.34

Richest

1.13 (0.37 - 3.43)

0.83

Decision maker regarding health of mother

Highest paternal educational level None

p-value†

OR (95% CI)*

*

1.00

Mother alone

1.00

Mother and husband/partner

3.64 (1.54 - 8.61)†

0.003†

†

0.020†

Husband/partner alone

2.71 (1.17 - 6.27)

Others

0.00 (0.00 - ∞)

1.00

Unweighted data used for regression.

Statistically significant.

†

Incomplete primary

1.05 (0.40 - 2.76)

0.93

Complete primary

1.04 (0.55 - 1.96)

0.91

Incomplete secondary

1.06 (0.51 - 2.20)

0.87

Complete secondary

1.20 (0.64 - 2.24)

0.58

Tertiary

2.16 (1.03 - 4.52)†

0.04†

*Unweighted data used for regression. † Statistically significant.

term includes several false positives and as such overestimates pneumonia two to seven fold.[25] However, its use as a denominator for evaluating careseeking behaviour in large-scale studies (≥20 000 children) is appropriate and has been validated.[25]

Conclusion

The study has strengthened the global recommendation that pneu­ monia case management efforts must take care-seeking behaviour into account. It has also demonstrated that interventions focusing on improving education, socioeconomic status of parents, empowering women in decision-making and increased coverage of health insurance may play a crucial role in improving pneumonia-related morbidity and mortality in Nigeria.

References

1. UNICEF. Committing to Child Survival: A Promise Renewed Progress Report 2014. New York: UNICEF, 2014.

2. Gupta G. Tackling pneumonia and diarrhoea: The deadliest diseases for the world’s poorest children. Lancet 2012;379(9832):2123-2124. [http://dx.doi. org/10.1016/S0140-6736(12)60907-6] 3. Wardlaw T, You D, Hug L, Amouzou A, Newby H. UNICEF Report: Enormous progress in child survival but greater focus on newborns urgently needed.

Reprod Health 2014;11(1):82. [http://dx.doi.org/10.1186/1742-4755-11-82] 4. World Health Organization. Global Action Plan for the Prevention and Control of Pneumonia in children aged under 5 years. Wkly Epidemiol Rec 2009;84(43):451-452. [http://dx.doi.org/10.2471/BLT.08.053348] 5. Qazi S, Weber M, Boschi-Pinto C, Cherian T. Global action plan for the prevention and control of pneumonia (GAPP): Report of an informal consultation: La Mainaz, Gex, France, 5 - 7 March 2007. Geneva: World Health Organization, 2008:1-24.

6. Qazi S, Aboubaker S, MacLean R, et al. Ending preventable child deaths from pneumonia and diarrhoea by 2025. Development of the integrated Global Action Plan for the Prevention and Control of Pneumonia and Diarrhoea. Arch Dis Child 2015;100(Suppl 1):S23-S28. [http://dx.doi.org/10.1136/archdischild-2013-305429] 7. Abdulkadir MB, Ibraheem RM, Johnson WBR. Sociodemographic and clinical determinants of time to care-seeking among febrile children under-five in North-Central Nigeria. Oman Med J 2015;30(5):331-335. [http://dx.doi. org/10.5001/omj.2015.68]

95

SAJCH

8. Abdulraheem I, Parakoyi D. Factors affecting mothers’ healthcare‐seeking behaviour for childhood illnesses in a rural Nigerian setting. Early Child Dev Care 2009;179(5):671-683. [http://dx.doi.org/10.1080/03004430701500885] 9. Taffa N, Chepngeno G. Determinants of health care seeking for childhood illnesses in Nairobi slums. Trop Med Int Health 2005;10(3):240-245. [http:// dx.doi.org/10.1111/j.1365-3156.2004.01381.x] 10. UNICEF. Despite steady progress, pneumonia remains one of the single largest killer of young children worldwide. 2013. http://data.unicef.org/child-health/ pneumonia.html (accessed 11 January 2015).

11. Sreeramareddy C, Shankar R, Sreekumaran B, Subba S, Joshi H, Ramachandran U. Care seeking behaviour for childhood illness: A questionnaire survey in western Nepal. BMC Int Health Hum Rights 2006;6(1):7. [http://dx.doi. org/10.1186/1472-698X-6-7] 12. D’Souza R. Care-seeking behavior. Clin Infect Dis 1999;28(2):234. [http:// dx.doi.org/10.1086/515120] 13. Goldman N, Heuveline P. Health-seeking behaviour for child illness in Guatemala. Trop Med Int Health 2000;5(2):145-155. [http://dx.doi.org/10.1046/ j.1365-3156.2000.00527.x] 14. National Population Commission (Nigeria), ICF International. Nigeria Demographic and Health Survey 2013. Abuja: National Population Commission, 2014.

15. Gove S. Integrated management of childhood illness by outpatient health workers: Technical basis and overview. Bull World Health Organ 1997;75(1):7-16.

16. Rutstein SO, Rojas G. Guide to DHS Statistics. Calverton: ORC Macro, 2006:1-160.

17. Rutstein SO, Johnson K. The DHS Wealth Index. Calverton: ORC Macro, 2004.

18. Hill Z, Kendall C, Arthur P, Kirkwood B, Adjei E. Recognizing childhood illnesses and their traditional explanations: Exploring options for care-seeking interventions in the context of the IMCI strategy in rural Ghana. Trop Med Int Health 2003;8(7):668-676. [http://dx.doi.org/10.1046/j.1365-3156.2003.01058.x] 19. de Souza ACT, Peterson KE, Andrade FMO, Gardner J, Ascherio A. Circumstances of post-neonatal deaths in Ceara, Northeast Brazil: Mothers’ care-seeking behaviors during their infants’ fatal illness. Soc Sci Med 2000;51(11):1675-1693. [http://dx.doi.org/10.1016/S0277-9536(00)00100-3] 20. Greenwood B. A global action plan for the prevention and control of pneumonia. Bull World Health Organ 2008;86(5):322-322A. [http://dx.doi. org/10.1590/S0042-96862008000500002] 21. Sazawal S, Black RE. Effect of pneumonia case management on mortality in neonates, infants, and preschool children: A meta-analysis of communitybased trials. Lancet Infect Dis 2003;3(9):547-556. [http://dx.doi.org/10.1016/ S1473-3099(03)00737-0] 22. Herman E, Black RE, Wahba S, Khallaf N. Developing strategies to encourage appropriate care-seeking for children with acute respiratory infections: An example from Egypt. Int J Health Plann Manage 1994;9(3):235-243.

23. Hildenwall H, Nantanda R, Tumwine JK, et al. Care-seeking in the development of severe community acquired pneumonia in Ugandan children. Ann Trop Paediatr 2009;29(4):281-289. [http://dx.doi.org/10.1179/027249309X12547917869005] 24. Rutebemberwa E, Kallander K, Tomson G, Peterson S, Pariyo G. Determinants of delay in care-seeking for febrile children in eastern Uganda. Trop Med Int Health 2009;14(4):472-479. [http://dx.doi.org/10.1111/j.1365-3156.2009.02237.x] 25. Campbell H, el Arifeen S, Hazir T, et al. Measuring coverage in MNCH: Challenges in monitoring the proportion of young children with pneumonia who receive antibiotic treatment. PLoS Med 2013;10(5):e1001421.[http:// dx.doi.org/10.1371/journal.pmed.1001421]

MARCH 2016 Vol. 10 No. 1

CASE REPORT

Too much diarrhoea, too many infections, and too few neutrophils T Padayachi, MB ChB, DCH (SA), FCP (Paed) (SA); I E Haffejee, MB BCh, MD, FCP (Paed) (SA), FCFP (SA) Department of Paediatrics and Child Health, Nelson R Mandela School of Medicine, University of KwaZulu-Natal, and King Edward VIII Hospital, Durban, South Africa Corresponding author: I E Haffejee (iehaffejee@hotmail.com)

Neutropenia is a common finding in ill paediatric patients and can be conveniently subdivided into that due to decreased production or increased destruction. Intrinsic defects in granulocytes or their progenitors or extrinsic factors such as infection, drugs and autoimmune phenomena are the main causes. Our case focuses on cyclic neutropenia presenting with severe chronic diarrhoea lasting 17 weeks, multiple recurrent bacterial infections and pulmonary tuberculosis (TB). To our knowledge this is the first case of pulmonary TB described in association with cyclic neutropenia and chronic diarrhoea in Africa.

S Afr J Child Health 201610(1):96-98. DOI:10.7196/SAJCH.2016.v10i1.1050

A 7-month-old boy was admitted to the paediatric gastroenteritis ward in King Edward VIII Hospital, Durban, with acute gastroenteritis of 3 days’ duration. Birth history was significant for prematurity with a low birth weight of 1.55 kg secondary to premature rupture of membranes. He developed neonatal respiratory distress syndrome requiring surfactant replacement therapy as well as intermittent positive pressure ventilation for the first 6 days of life. He was also HIV-exposed and had received perinatal antiretroviral prophylaxis; the HIV polymerase chain reaction (PCR) was negative at 6 weeks of age. Previous medical history included an admission at another hospital for acute gastroenteritis, and lower respiratory tract infection at 3 months of age. The baby had been exclusively formula fed from birth, with complementary foods introduced from 5 months of age.

There was no family history of recurrent infections or unexplained deaths.

Clinical examination revealed an acutely malnourished patient (weight 3.36 kg, length 57 cm), whose weight-for-length z-score was <–3. He was moderately dehydrated, and not in hypovolaemic shock. He had oral candidiasis, generalised lymphadenopathy, and hepatomegaly. Blood tests showed a mild metabolic acidosis with no electrolyte derangements (sodium 139 mmol/L, potassium 3.4 mmol/L, chloride 112 mmol/L, sodium bicarbonate 17 mmol/L, anion gap 13 mmol/L, urea 1.6 mmol/L, creatinine 22 µmol/L), and a slightly low serum albumin of 27 g/L. The full blood count showed a normocytic normochromic anaemia of 8 g/dL, a platelet count of 284 × 109/L, and remarkably, a leucopenia of 1.11 × 109/L with a severe neutropenia of 0.3 × 109/L, a normal lymphocyte count (0.65 × 109/L) and monocytes of 0.14 × 109/L. Blood levels of folate, iron and vitamin B12 were normal, and the direct Coombs' test was negative. There was no reticulocytosis.

Blood culture was positive for Escherichia coli sensitive to amoxicillin-clavulanic acid and gentamicin, which were administered for 1 week. Two weeks later he developed culturepositive nosocomial Enterobacter aerogenes septicaemia, which was treated with pipericillin-tazobactam and amikacin. The neutrophil count at this stage was 0.37 × 109/L. Stool cultures were negative for bacteria, viruses and parasites, and stools were not fatty or foul smelling. Faecal elastase levels were adequate (>500 µg/g). The diarrhoea continued despite introducing a lactose-free formula after 4 days in hospital, necessitating semi-elemental formula feeds 9 days after admission. The gastroenteritis resolved 3 weeks after admission, only to recur soon after.

96

SAJCH

After resolution of the bacterial sepsis with two subsequent negative blood cultures over a period of 2 weeks, the patient remained persis­ tently neutropaenic, and continued to have diarrhoea. Viral studies (cytomegalovirus, Epstein-Barr virus, toxoplasma, parvovirus, rotavirus and adenovirus) were all negative, including a repeat HIV PCR. Thyroid function was normal (thyroid-stimulating hormone 1.56 mIU/L, T4 9.9 pmol/L), and serum complement levels (C3 0.67 g/L, C4 0.2 g/L) and immunoglobulin concentrations (IgG <0.25 g/L, IgA 0.95 g/L, IgM <0.15 g/L) were within normal limits.

Due to the high prevalence of tuberculosis (TB) in South Africa (SA), a presumptive diagnosis of TB was considered as a cause of the chronic nature of the diarrhoea. Purified protein derivative (Mantoux) test was negative but this could have been a false negative owing to severe malnutrition. Radiography of the chest revealed hilar lymphadenopathy; there were no other features suggestive of TB. Abdominal ultrasonography showed no granulomas or enlarged lymph nodes. Microscopy of gastric washings for acid-fast bacilli using the auramine stain was negative. The patient was empirically started on anti-TB treatment 6 weeks after admission. Sputum culture was later found to be positive for acid-fast bacilli, and anti-TB treatment was continued for 6 months. It should be noted that although the patient had a radiological finding of hilar adenopathy together with culture-proven pulmonary TB, a diagnosis of abdominal TB was not confirmed.

Eventual resolution of the chronic diarrhoea only took place 17 weeks after admission.

Differential diagnosis

The aetiology of neutropenia can be broadly divided into defects of myelopoiesis and extrinsic factors like malnutrition, drugs or autoimmune phenomena.[1] Diarrhoea is a frequent complication of neutropenia.[2] In our patient, a differential diagnosis of (i) cyclic neutropenia, (ii) chronic congenital neutropenia, (iii) chronic idio­ pathic neutropenia and (iv) Schwachman-Diamond syndrome (exocrine pancreatic insufficiency with neutropenia) was considered.

Other conditions were excluded on clinical and laboratory grounds (Table 1). Combined immune deficiency is very unlikely in our patient since the immunoglobulin levels were within normal limits.

Oral candidiasis as well as lymphadenopathy has been described in cyclic neutropenia by previous workers.[1,3] Bone marrow aspirate and trephine biopsy revealed normocellular marrow in which infection could not be definitively excluded.

MARCH 2016 Vol. 10 No. 1

CASE REPORT No granulomas or infiltrates were observed. Myelopoiesis was hypoplastic with a left shift; however, it was adequate and sequential; maturation arrest was not present, more mature forms being present.

This is consistent with a diagnosis of cyclic neutropenia.[3] While hospitalised our patient had a series of infections, including a lower respiratory infection, cellulitis following intravenous line insertion, septicaemia with associated thrombocytopenia, and elevated C-reactive protein. Most infections followed procedures such as intravenous line insertion and bone marrow aspiration. Apart from the pathogens mentioned above, Pseudomonas and Staphylococcus spp. were also cultured sequentially at 2 - 3-week intervals, the neutrophil count ranging from 0.3 × 109/L to 0.37 × 109/L (Table 2). Antibiotics used included meropenem and ciprofloxacillin together with antifungals.

There was no hepatic dysfunction aside from a decreasing albumin level from 27 g/L on admission to 14 g/L 5 months later. The Table 1. Differential diagnosis of neutropenia (adapted)[1] Neutropenia caused by intrinsic defects in granulocytes or their progenitors Cyclic neutropenia Severe congenital neutropenia Shwachman-Diamond syndrome Reticular dysgenesis (associated with severe combined immunodeficiency) Albinism/neutropenia syndromes (including Chėdiak-Higashi) Familial benign neutropenia Bone marrow failure syndromes (congenital and acquired)

persistence of neutropenia and diarrhoea together with recurrent infections prompted us to administer granulocyte colony-stimulatingfactor (G-CSF) in an attempt to raise the neutrophil count.[3] Severe congenital neutropenia appeared unlikely as the bone marrow did not show ‘maturation arrest’ at the myelocyte stage of development, the promyelocytes were not increased nor showed dysplastic morphology such as large size, atypical nuclei or vacuolated cytoplasm. The clinical hallmarks of chronic idiopathic neutropenia are an acquired neutropenia with a relatively stable, suppressed neutrophil count without recurrent infections.[1] Our patient did not seem to fit into either of these two categories. Chronic diarrhoea with neutropenia is a feature of Schwachman-Diamond syndrome but this was ruled out by the non-fatty stools and normal stool elastase; there was also no evidence of skeletal dysplasia.[4] We suspect that our patient had cyclic neutropenia despite persistently (non-cyclic) low counts of <2 × 109/L for the first 3 months. Fig. 1 shows that from 12 weeks onwards there were regular cyclic periods when the neutrophil count dropped with a concomitant rise in the monocyte count; this is consistent with a diagnosis of cyclic neutropenia, which is characterised by periodic oscillations with approximately 3-weekly nadirs. Oscillations can be subtle in some patients.[1] There was a reciprocal monocytosis at each low point of the neutrophil count in our patient, who responded to three courses of G-CSF but still experienced neutropaenic cycles; this has been described in the literature.[1] The amplitude of the cycling may increase, but with shortened cycle periods and increased neutrophil counts at the nadir.[1] The baby was finally discharged 35 weeks after admission with a weight of 5 kg. When seen at followup 9 weeks after discharge, his neutrophil count was 1.17 × 109/L with a monocyte count of 0.9 × 109/L and he was thriving and well.

Discussion

Neutropenia caused by extrinsic factors Infection Drug-induced neutropenia (e.g. indomethacin, valproate, phenytoin, penicillin) Autoimmune neutropenia Nutritional deficiencies Bone marrow infiltration Reticuloendothelial sequestration

Cyclic neutropenia is an autosomal dominant genetic disorder with full penetrance but varying degrees of clinical manifestations; sporadic cases can arise from new germ line mutations. The disorder is secondary to a mutation in the neutrophil elastase gene that leads to increased apoptosis in neutrophil precursors. The genetic diagnosis can be confirmed by sequencing of the ELA2 gene.[1] Unfortunately, this genetic test is not available for use in SA. The primary cellular abnormality in cyclic neutropenia is apoptosis of neutrophil precursors and their removal by marrow macrophages.

This can be shown by flow cytometry which demonstrates increased

Table 2. Sequence of WCC, B/C results and clinical infections during admission Total WCC

Absolute neutrophils

Absolute lymphocytes

Absolute monocytes

Microbiology results

Concurrent clinical infection

On admission

1.11

0.30

0.65

0.14

B/C E. coli followed

by Enterobacter spp.

Septicaemia, gastroenteritis

Week 4

0.87

0.30

0.31

0.22

B/C Pseudomonas

Septicaemia, gastroenteritis

Week 6

1.36

0.57

0.44

0.33

B/C S. aureus

Pneumonia

Week 14

1.23

0.31

0.45

0.46

-

Gastroenteritis

Week 19

2.79

1.56

0.25

0.96

-

Thrombophlebitis

Week 23

2.15

0.80

0.37

0.96

Pus swab Pseudomonas

Cellulitis

Week 27

3.30

1.58

0.76

0.95

-

Thrombophlebitis

Week 29

12.91

6.75

1.36

4.54

-

-

Week 33

3.00

0.90

1.02

1.08

-

-

Week 45

5.83

2.17

1.34

2.28

-

-

WCC = white cell count; B/C = blood culture.

97

SAJCH

MARCH 2016 Vol. 10 No. 1

CASE REPORT neutropenia was based on the clinical presentation, the cyclical nature of the low neutrophil counts with reciprocal monocytosis, and the bone marrow abnormalities. It should be noted that the ‘typical’ bone marrow abnormality seen in cyclic neutropenia, namely hypoplasia of the myeloid series with maturation arrest at the myelocyte stage, was not seen in our patient; however, this occurs during the declining phase of the neutrophil oscillation, and myelopoiesis can be normal or even hyperplastic during neutrophil recovery.

This case illustrates a number of learning points: (i) ongoing diarrhoea of very long duration should be thoroughly investigated to look for a definite underlying cause and not be dismissed as being ‘post-enteritic’; (ii) a low leucocyte count, especially a low neutrophil count, must never be ignored and an attempt at establishing its aetiology should always be made; and (iii) apart from bacterial pathogens causing sequential severe infections, especially following invasive procedures, TB should always be searched for in communities where it is endemic. To our knowledge, this is the first case of pulmonary TB described in association with cyclic neutropenia and chronic diarrhoea in the African continent.

14

12

Number of cells × 109/L

10

8

6

4

2

week 30 week 35

week 28

week 24 week 26

week 20 week 22

week 16 week 17

week 14 week 15

week 13 week 13

week 7

week 10 week 12

week 5 week 7

week 4 week 5

week 4

week 2

Admission

0

Acknowledgements. The authors would like to thank Drs N Mathenjwa and K Naidoo for assisting with patient management.

Week

References WCC

Neutrophils

Monocytes

Bacteraemia

Gastroenteritis

Fig. 1. Response to courses of G-CSF.

numbers of annexin-V stained cells. Furthermore, a lymphocyte subset defect (which can give a similar clinical presentation) can also be diagnosed by flow cytometry. This was not carried out in our patient and is thus a limitation of our study. The diagnosis of cyclic

98

SAJCH

1. Dinauer MC, Newburger PE. The phagocyte system and disorders of granulopoiesis and granulocyte function. In: Orkin SH, Nathan DG, Ginsburg D, Look AT, Fisher DE, Lux SE, eds. Nathan and Oski’s Hematology of Infancy and Childhood. 7th ed. Philadelphia: Saunders, 2009:1109-1220.

2. Aksoy DY, Tanriover MD, Uzun O, et al. Diarrhoea in neutropenic patients: A prospective cohort study with emphasis on neutropenic enterocolitis. Ann Oncol 2007;18(1):183-189. [http://dx.doi.org/10.1093/annonc/mdl337] 3. Dale DC, Bolyaard AA, Aprikyan A. Cyclic neutropenia. Semin Hematol 2002;39(2):89-94.

4. Burroughs L, Woolfrey A, Shimamura A. Shwachman-Diamond syndrome: A review of the clinical presentation, molecular pathogenesis, diagnosis, and treatment. Hematol Oncol Clin N Amer 2009;23(2):233-248. [http://dx.doi. org/10.1016/j.hoc.2009.01.007]

MARCH 2016 Vol. 10 No. 1

CASE REPORT

Ecthyma gangrenosum caused by Stenotrophomonas maltophilia in a neutropenic leukaemic infant: A case report D K Das,1 MD (Paed), DCH; S Shukla,2 MD (Paed) 1 2

Department of Pediatrics, Hi Tech Medical College, Bhubaneswar, Odisha, India Department of Pediatrics, SCB Medical College, Cuttack, Odisha, India

Corresponding author: D K Das (dr.dillipdas@gmail.com)

Ecthyma gangrenosum (EG) is a cutaneous lesion, mostly caused by pseudomonas in immunocompromised patients. Other bacterial and fungal pathogens have also been reported. It can occasionally affect previously healthy children. The cutaneous findings are characterised by small indurated papulovesicles, progressing rapidly to necrotic ulcers with surrounding erythema and a central black eschar. Sites most commonly involved are the buttocks, perineum, limbs and axillae; the face is less commonly involved. We are presenting a rare case of EG in a neutropenic infant who had just completed the induction phase of chemotherapy for acute lymphoblastic leukaemia. The gangrenous lesion was on the face involving the tip of the nose, which is an uncommon location.

Blood and pus cultures grew Stenotrophomonas maltophilia, which is a rare cause of EG. The patient was treated with IV antibiotics (colistin for 14 days) and improved.

S Afr J Child Health 2016;10(1):99-100. DOI:10.7196/SAJCH.2016.v10i1.957

Ecthyma gangrenosum (EG) is a welldescribed skin lesion, mostly seen in immunocompromised patients. The lesion characteristically progresses rapidly from small indurated papulovesicles to necrotic ulcers with surrounding erythema and a central black eschar.[1] Although rare, the presence of EG is indicative of severe systemic infection with a potentially fatal prognosis. It typically occurs on the extremities or in gluteal and perineal regions. Although Pseudomonas aeruginosa is the pathogen most often associated with EG, other organisms have also been reported.[2] We report a case of EG caused

by Stenotrophomonas maltophilia, presenting as an aggressive necrotic facial skin lesion involving the tip of nose in a 9-month-old boy who was receiving treatment for acute lymphoblastic leukaemia (ALL).

Case

A 9-month-old boy with ALL who had recently completed the induction phase of chemotherapy, presented to Hi Tech Medical College, Bhubaneswar – a tertiary care hospital in the eastern part of India – with complaints of fever for 3 days and a gangrenous lesion (which started as vesicle and progressed rapidly to gangrene) on the nose for 2 days. Prior to the diagnosis of ALL, he had been well, without any major illness. Examination revealed mild anaemia, and no lymphadenopathy.

Examination of the nose showed a brownish black gangrenous ulcer surrounded by an erythematous halo (Fig. 1). Haematological investigation showed a leucocyte count of

Fig. 1. Tip of the nose showing brownish-black gangrenous ulcer surrounded by an erythematous halo suggestive of EG.

1 200/mm3 with an absolute neutrophil count of 380/mm3. C-reactive protein was 80 mg/L. Blood and pus cultured from the site showed growth of S. maltophilia. A diagnosis of EG caused by S. maltophilia was made. Empiric antibiotic therapy consisted of ceftazidime and amikacin intravenously, but this was changed to colistin based on sensitivity results conducted on the isolate. Colistin was continued for 14 days and the lesion healed completely. He was well at discharge 21 days after admission.

Discussion

EG generally develops in patients with under­ lying immunodeficiency, but occasionally can occur in immunocompetent indivi­ duals. [2] Factors that are associated with higher 99

SAJCH

MARCH 2016 Vol. 10 No. 1

mortality include neutropenia, septic shock, inappropriate or delayed antibiotic therapy and resistant microorganisms.[2,3] EG has also been described in infants and young children with transient risk factors, such as concurrent viral infection and recent antibiotic therapy.[4] EG is caused by invasion of microorganisms into the media and adventitia of subcuta­ neous vasculature, precipitating a haemor­ rhagic occlusive vasculitis.[5] The skin lesions of EG are the manifestation of this necrotising vasculitis. The lesions characteristically begin as an erythematous nodule, macule, vesicle or bulla and evolve into gangrenous ulcerations with a black eschar and a surrounding rim of erythema.[6] The vesicles, initially filled with serous fluid, appear on the surface of the oedematous skin, and then coalesce to form large bullae that slough away, leaving ulcerated, necrotic centres with erythematous halos.[6,7] Lesions progress very rapidly. The gluteal region is most commonly involved (57% cases), followed by extremities in­clud­ ing the axillae (30%) and the trunk (6%). The face is also affected in ~6% of cases.[7] Fever and other constitutional symptoms may be present, depending on the extent of the underlying infection and the patient’s immune status. Gastrointestinal and respiratory complaints are also commonly described.[8] Suspicion for EG warrants a prompt diagnosis with cultures and sensitivities performed on blood and pus swabs or tissue specimens.

P. aeroginosa is the most common offend­ ing organism. Other organisms which have been isolated in patients with EG are shown in Table 1.

CASE REPORT Table 1. Organisms isolated in patients with EG Gram-negatives

Gram-positives

Fungi

Escherichia coli

Staphylococcus aureus

Aspergillus fumigatus

Klebsiella pneumoniae

Streptococcus pyogenes

Candida albicans

Morganella morganii

Fusarium solani

Neisseria gonorrhoea

Meterhisium anisopliae

Serratia marcescens

Mucor pusilus

S. maltophilia

S. maltophilia is an aerobic, Gram-negative baci­ llus and mostly causes opportunistic nosocomial infection.[9] Colonisation with S. malto­­philia is most commonly encountered in patients with immunosuppression. Skin and soft-tissue infection in the form of cellulitis, infected mucocutaneous ulcers, EG and paronychia have been associated with S. maltophilia infection.[9] The route of transmission is unknown; it is speculated that invasion may take place via defects in mucous membranes and by colonisation of central venous catheters.

Infections with S. maltophilia are often lifethreatening because of intrinsic resistance to many antibiotics and the general condition of the affected patients.[9] Early diagnosis and prompt treatment of EG are crucial for decreasing mortality and preventing complications associated with longterm sequelae. The choice of antimicrobial treatment depends on the site, severity of infection and antimicrobial sensitivity tests.

Initial antibiotics should cover the Gramnegative organisms. Common­ ly used anti­ biotics are ceftazidime, carbenicillin indanyl sodium, gentamicin sulfate, imipenem, mezlocillin and piperacillin sodium. A

combination of an antipseudomonal β-lactam agent and either an aminoglycoside or a quinolone is used for empiric therapy. Once culture reports are available, antibiotics should be modified accordingly.[10] Trimethoprim-sulfamethoxazole (TMPSMX) is recommended as the agent of choice for treatment of S. maltophilia infection.

Ticarcillin-clavulanate has good activity against S. maltophilia and is the agent of choice in individuals intolerant of TMPSMX, or if the organism is resistant to TMPSMX. Alternative antibiotic agents that may be used to treat isolates resistant to first-line agents include colistin and polymyxin B. Surgical drainage of localised abscesses and debridement of all necrotising tissues may be needed to prevent the spread of infection and septicaemia. Large tissue defects may require reconstructive surgery.

Conclusion

Our case illustrates an unusual presentation of EG, both in terms of the site involved and organism isolated. Patients presenting with EG warrant hospitalisation, thorough micro­ biological investigation and initiation of

100

SAJCH

MARCH 2016 Vol. 10 No. 1

empirical broad-spectrum antipseudomonal antibiotic therapy, with rationalisation of antibiotic choice once susceptibility results of the isolate become available.

References 1. Greene SL, Su WP, Muller SA. Ecthyma gangrenosum: Report of clinical, histopathologic, and bacteriologic aspects of eight cases. J Am Acad Dermatol 1984;11(5 Pt 1):781-787.

2. Patel JK, Perez OA, Viera MH, Halem M, Berman B. Ecthyma gangrenosum caused by Escherichia coli bacteremia: A case report and review of the literature. Cutis 2009;84(5):261-267.

3. Peña C, Suarez C, Gozalo M, et al. Prospective multicenter study of the impact of carbapenem resistance on mortality in Pseudomonas aeruginosa bloodstream infections. Antimicrob Agents Chemother 2011;56(3):1265-1272. [http:// dx.doi.org/10.1128/AAC.05991-11] 4. Pechter PM, Marchione R, Milikowski C, Berman B. Ecthyma gangrenosum secondary to Staphylococcus aureus in an infant with transient neutropenia.

Pediatr Dermatol 2012;29(3):320-323. [http:// dx.doi.org/10.1111/j.1525-1470.2011.01427.x] 5. Goolamali SI, Fogo A, Killian L, et al. Ecthyma gangrenosum: An important feature of pseudomonal sepsis in a previously well child.

Clin Exp Dermatol 2009;34(5):e180-e182. [http:// dx.doi.org/10.1111/j.1365-2230.2008.03020.x] 6. Pandit AM, Siddaramappa B, Choudhary SV, Manjunathswamy BS. Ecthyma gangrenosum in a new born child. Indian J Dermatol Venereol Leprol 2003;69(7):52-53.

7. Downey DM, O’Bryan MC, Burdette SD, Michael JR, Saxe JM. Ecthyma gangrenosum in a patient with toxic epidermal necrolysis. J Burn Care Res 2007;28(1):198-202. [http://dx.doi.org/10.1097/ BCR.0B013E31802CA481] 8. Fink M, Conrad D, Matthews M, Browning JC. Primary ecthyma gangrenosum as a presenting sign in a child. Dermatol Online J 2012;18(3):3.

9. Smeets JG, Lowe SH, Veraart JC. Cutaneous infections with Stenotrophomonas maltophilia in patients using immunosuppressive medication. J Eur Acad Dermatol Venereol 2007;21(9):1298-1300. [http:// dx.doi.org/10.1111/j.1468-3083.2007.02201.x] 10. Marra AR, Pereira CA, Gales AC, et al. Bloodstream infections with metallo-beta-lactamase-producing Pseudomonas aeruginosa: Epidemiology, microbiology, and clinical outcomes. Antimicrob Agents Chemother 2006;50:388-390.

CPD March 2016 CPD questionnaires must be completed online at www.mpconsulting.co.za

True (T) or False (F): Regarding HIV testing in children whose status is unknown 1. The majority of mothers (~60%) do not give consent for the testing of their children to establish their HIV status when they are admitted to hospital.

2. The major reason for not giving consent is that the mothers do not want to be stressed by receiving a positive result.

Regarding poisoning in children 10. Infants comprise the largest proportion of children with poisoning.

11. Most suspected poisonings in infants are related to pharmaceutical agents.

12. Paraffin ingestion makes up over one-third of the admissions for poisoning in children in a Ghanaian hospital.

Regarding coeliac disease and type 1 diabetes mellitus 3. Coeliac disease is five times more common in subjects with type 1 diabetes mellitus than in the general population.

4. Only those patients with type 1 diabetes who are symptomatic of suggested coeliac disease should be screened for coeliac disease.

Regarding medico-legal deaths in Pretoria 13. Children and adolescents comprise ~9% of the cases seen.

14. Pedestrian deaths make up less than one-third of deaths due to road accidents.

Regarding hearing loss in children 5. It is estimated that between 3 and 6/1 000 children in South Africa have hearing loss.

Regarding haemophilia A in South Africa 6.

There are ~5 carrier females for every affected male in the population.

7. Carrier females do not have an increase in the propensity to bleed excessively.

Regarding nutrition in the first 1 000 days of life 8. The pattern of growth in the first 1 000 days of life has a significant effect on the risk of non-communicable diseases in later life.

9. Childhood obesity coexists with childhood undernutrition in the first 2 years of life in the Western Cape.

Regarding children admitted to a paediatric intensive care unit (PICU) 15. Severity of malnutrition on admission is not associated with the risk of mortality of a child admitted to a PICU. 16. International clinical severity scores underpredict mortality rates in a South African PICU. Regarding malaria in Ethiopia 17. Malarial presentations of children in hospital are associated with chronic undernutrition.

18. Sleeping under impregnated bed nets reduces the risk of malarial infection by about two-thirds.

Regarding retinopathy of prematurity (ROP) 19. Approximately 1 in 5 very-low-birth-weight infants in South Africa is affected by ROP. 20.

Screening for ROP should include all infants born before 32â&#x20AC;&#x201A; weeksâ&#x20AC;&#x2122; gestation.

A maximum of 3 CEUs will be awarded per correctly completed test.

CPD questionnaires must be completed online via www.mpconsulting.co.za. After submission you can check the answers and print your certificate. Accreditation number: MDB015/177/02/2016 (Clinical)

101

SAJCH

MARCH 2016 Vol. 10 No. 1