SAJOG Vol 22, No 2 (2016)

ISSN 2305-8862

December 2016 Vol. 22 No. 2

• Validation of the prolapse quality-of-life questionaire (P-QOL): An Afrikaans version in a South African population • Birth preparedness and complication readiness among pregnant women in a rural community in southern Nigeria • Sepsis: Primary indication for peripartum hysterectomies in a South African setting • Three-dimensional colour Doppler of ductus venosus agenesis in the first trimester

SAJOG December 2016 Volume 22 No. 2

THE SOUTH AFRICAN JOURNAL OF OBSTETRICS AND GYNAECOLOGY

Editor William Edridge Editorial Board: SAJOG Alan Alperstein (Cape Town) Geoffrey Buga (Walter Sisulu) Hennie Cronje (Free State) Franco Guidozzi (Witwatersrand) Justus Hofmeyr (East London) Thinus Kruger (Stellenbosch) Gerhard Lindeque (Pretoria) Eddie Mhlanga (KwaZulu-Natal) Sam Monokoane (Limpopo) Jack Moodley (KwaZulu-Natal) Dan Ncayiyana (Durban) Hein Odendaal (Stellenbosch) Zephne van der Spuy (Cape Town) HEALTH & MEDICAL PUBLISHING GROUP (HMPG):

CONTENTS

CEO and Publisher Hannah Kikaya Email: hannahk@hmpg.co.za

Editorial

Executive Editor Bridget Farham

37

The language of communication

W Edridge

Research 38 Validation of the prolapse quality-of-life questionaire (P-QOL): An Afrikaans version in a South African population

C Brandt, C van Rooyen, H S Cronje

42 Patient knowledge about and intention to use the intrauterine contraceptive device (IUCD) at a tertiary-level hospital

N van der Westhuizen, G Hanekom

47 Birth preparedness and complication readiness among pregnant women in a rural community in southern Nigeria

S H Ibadin, V Y Adam, O A Adeleye, O H Okojie

52 Sepsis: Primary indication for peripartum hysterectomies in a South African setting L Jansen van Vuuren, C A Cluver

57 Factors predictive of adnormal semen parameters in male partners of couples attending the infertility clinic of a tertiary hospital in south-western Nigeria

Managing Editors Ingrid Nye Claudia Naidu Technical Editors Emma Buchanan Paula van der Bijl Production Manager Emma Jane Couzens DTP and Design Travis Arendse Clinton Griffin Chief Operating Officer Diane Smith | Tel: 012 481 2069 Email: dianes@hmpg.co.za Online Support Gertrude Fani | Tel: 072 463 2159 Email: publishing@hmpg.co.za ISSN 2305-8862 Journal website: www.sajog.org.za Use of editorial material is subject to the Creative Commons Attribution – Non-commercial Works Licence. https://creativecommons.org/licenses/ by-nc/4.0

O P Aduloju, P T Adegun

Case Reports 62 Skene's gland duct cysts: The utility of vaginal/transperineal imaging in diagnosis and mapping for surgery

P F Kruger, R Kung, F Hamidinia, R Rahmani

65 Three-dimensional colour Doppler of ductus venosus agenesis in the first trimester D Singh, L Kaur

67 Respiratory arrest caused by a large uterine myoma

M S Kim, G H Lee, M C Choi, H Park, S G Jung, K A Kim

69

CPD Questions

Listed in Excerpta Medica (EMBASE), Biological Abstracts (BIOSIS), Science Citation Index (SciSearch), Current Contents/Clinical Medicine Published by the Health and Medical Publishing Group, a subsidiary of the South African Medical Association HEAD OFFICE Block F, Castle Walk Corporate Park, Nossob Street, Erasmuskloof Ext. 3, Pretoria, 0181 Tel: 012 481 2069 EDITORIAL OFFICE Suite 11, Lonsdale Building, Lonsdale Way, Pinelands, 7405 Tel: 021 532 1281 Cell: 072 635 9825 E-mail: publishing@hmpg.co.za All letters and articles for publication must be submitted online at www.sajog.org.za ©Copyright: Health and Medical Publishing Group (Pty) Ltd.

This open-access article is distributed under Creative Commons licence CC-BY-NC 4.0.

EDITORIAL

The language of communication Speaking to a person in their own language provides an opportu­ nity to place that person at their ease, an essential requirement in medicine, and makes retrieving information from that person easier.

An article in this edition discusses the conversion of an English uterovaginal pelvic organ prolapse questionnaire into the language of Afrikaans. This South African (SA) language is based on European Dutch brought by settlers in a migration from Europe in the 17th century. It is closer to modern Flemish, spoken in Belgium. Afrikaans is the first language of approximately 13% of the SA population.

SA has 11 official languages. Zulu (or more correctly isiZulu) is the first language of 22% of the population. Several languages are spoken by only 5%. English is the first language of only 9%. There is clearly a great need to translate questionnaires into many languages and this is acknowledged by the article. Any person in SA wishing to perform research translates questionnaires into the predomi­ nant languages of that region. This facility in language should really be for everyday usage and not just for writing papers or for giving validity to academic research.

The problem of language is a worldwide phenomenon. It affects all populations and medical professionals from the Amazon to the Hindu Kush, from Timbuktu to Kamchatka and into the Pacific.

India has 22 official languages, though dialects would multiply this number even further. These include Hindi in the north, Bengali in the east, Tamil in the south, and Gujarati in the west. Any Indian would see this as an insulting oversimplification. Hindi is the langu­age of government and films with the international language of family conflict. Hindi is spoken by 40% of India’s population; English is spoken with ease by 9%.

Any large or populous area suffers from the same problem. The Beijing language of Mandarin may be spoken theoretically by over 900 million of China’s 1.2 billlion people. But China is divided into at least 56 different ethnic groups. The number of spoken languages is large, including the Turkic languages in the west incorporated into China by several thousand years of migration along western trade routes.

Communicating in medicine in India, China, South America, Africa, or the Middle East may require many university-trained doctors to speak and investigate medical problems via interpreters.

This must limit the accuracy and reliability of information. In English there is a children’s game (broken telephone) in which each child whispers a message to a neighbour, and as it passes down a line of children a very different message emerges. Although in a medical interaction where language is an issue this process is limited to the message of one interpreter, the same problem can occur.

It is not just the actual accuracy of medical information that may be lost in translation, but also its completeness. Much of medicine lies in the detail. We have all sat next to a patient who communicates with an interpreter at some length, with a series of sentences exchanged, to be told simply on inquiring that ‘the pain has gone.’

Accuracy and completeness of information are important, not just

37

in acquiring information for the detective element of medicine, but also in the information and reassurance that is given by the doctor to the patient. Reassurance is an important element of medicine.

This can be lost by an interpreter. As language skill of the doctor improves this can become very clear, as reassurance given to the patient in several sentences is omitted and is translated by an interpreter as, ‘you can get your pills from the pharmacy.’

Communicating through an interpreter may also limit the pri­ vacy of the moment, which must inhibit disclosure of information thought to be sensitive by the patient; this encompasses almost everything that might pertain to obstetrics or gynaecology. The patient may not experience loss of trust in the interpreter, but it is possible that sensitive information such as urine leakage may not be volunteered if a patient does not see the interpreter as sensitive or sympathetic.

Ethnic diversity breeds confusion if a dominant centralised train­ing sys­tem requires centralised language usage. Even when a common language is used by the doctor and patient, misinformation and miscommunication can prevail.

A dominant language, such as Mandarin, English, or Hindi, may engender a dominant pattern of behaviour from the doctor when talking to patients of diverse ethnic backgrounds for whom the preferred medical language is a second, third or fourth language. Doctors may often be better educated than their patients (though not necessarily clearer thinking). A tendency to dictate in a doctor-patient interaction rather than to listen easily emerges. A British training film attempted to emphasise this point by showing a senior doctor celebrating communication skills on a large ward round as the patient repeatedly replies, ‘Yes, doctor.’

When the ward round is completed the tea lady approaches, offering to pour tea or coffee from the trolley. ‘Yes, doctor,’ replies the patient, who clearly has not been part of the previous ward round conversation.

The doctor-patient interaction may easily benefit the doctor more than the patient. Doctors may satisfy themselves that commu­ nication has been good even when it has not. Much depends on the attitude and humility of the doctor and much depends on language. As we pass through a world where technology and machinery replace the fundamentals of medical investigation, where notes contain phrases such as ‘not sonographically or clinically appendicitis’, basic skills are being lost. The use of language, the use of another language, and in a common language the careful choice of routine words where complex words have no place are at the core of maintaining this process of medicine, the ability to investigate and the ability to heal.

Will Edridge Editor S Afr J Obstet Gynaecol 2016;22(2):37. DOI:10.7196/SAJOG.2016.v22i2.1143

SAJOG • December 2016, Vol. 22, No. 2

This open-access article is distributed under Creative Commons licence CC-BY-NC 4.0.

RESEARCH

Validation of the prolapse quality-of-life questionnaire (P-QOL): An Afrikaans version in a South African population C Brandt,1 PhD (Physiotherapy); C van Rooyen,2 M Comm; H S Cronjé,3 MB ChB, MD Department of Physiotherapy, University of the Free State, Bloemfontein, South Africa Department of Biostatistics, University of the Free State, Bloemfontein, South Africa 3 Private practice, Centurion, Pretoria, South Africa 1 2

Corresponding author: C Brandt (gnftcb@ufs.ac.za)

Background. The prolapse quality-of-life questionnaire (P-QOL) has been validated and translated into eight languages. The lack of an Afrikaans version of the P-QOL limits studies in Afrikaans-speaking patients with pelvic organ prolapse (POP).

Objective. To validate an Afrikaans version of the P-QOL in a South African population.

Methods. The P-QOL was translated into Afrikaans by a medical translator and three gynaecologists. This descriptive study determined construct validity comparing 25 symptomatic (64.1%, n=39) and 14 asymptomatic (35.9%, n=39) participants' median domain scores. The POP stage was determined according to the POP quantification (POP-Q) scale and compared with their domain scores by means of percentages.

A second P-QOL was completed and the stability determined by the test-retest method. The Cronbach alpha was used to determine internal consistency and the kappa value to determine measure of agreement.

Results. Symptomatic participants had higher median domain scores than asymptomatic participants. All asymptomatic participants had stage 0 POP and 33.3% of symptomatic participants had stage III POP. Stability was good, with an average of above 50%. The mean Cronbach alpha value was 0.94 and the kappa value indicated moderate to good strength of agreement between items (к=0.41 - 0.80).

Conclusion. The Afrikaans P-QOL was found to be valid and reliable to determine quality of life in women with POP, correlating with the findings of other validation studies and supporting the evidence that the P-QOL is a high-quality disease-specific quality-of-life questionnaire.

S Afr J Obstet Gynaecol 2016;22(2):38-41. DOI:10.7196/SAJOG.2016.v22i2.1077

Pelvic organ prolapse (POP) occurs in 46 - 73% of women in South Africa (SA) and may also be associated with other pelvic floor dysfunctions.[1] The assessment of quality of life (QOL) is becoming increasingly important in determining the outcome of pelvic floor reconstructive surgery, as well as other pelvic floor disorders.[2] POP is mostly benign, but it is distressing and disabling, with a large effect on the patient's QOL.[3] The multifactorial patho­ physiology may be the cause of associated symptoms such as bladder, bowel, sexual, and even pain symptoms.[4] The symptoms may lead to physical, social, psychological, domestic, and/or sexual limitations in the patient's activities of daily living.[5] A survey done

by Muller[6] in the USA found that women with POP experienced compromised bladder and bowel control as most limiting their QOL. Ranked second was the inability to enjoy physical activities such as sport. They also emphasised satisfaction with conservative and surgical management as an important factor in determining QOL.[6] It is important to validly and reliably determine the QOL from the patient’s perspective, because it has been indicated that the validity of QOL outcomes based on physicians’ perspectives should be interpreted with caution.[7] Changes in QOL from the patient’s per­spective should therefore be seriously considered when treatment and treatment outcomes of POP are determined.[8] The prolapse quality-of-life questionnaire (P-QOL) is one of only a few validated and reliable condition-specific questionnaires developed to assess the impact of urogenital prolapse on the QOL of patients. The questionnaire covers various domains of life which

include general health, prolapse impact, role limitations, physical limitations, social limitations, personal relationships, emotional problems, sleep/energy disturbances and prolapse severity.[9] The P-QOL has been successfully translated into eight lang­ uages, and includes versions in Italian,[5] Turkish,[10] German,[11] Portuguese,[12] Dutch,[2] Thai,[8] and most recently Persian.[13] The lack of a validated Afrikaans questionnaire investigating QOL in patients experiencing prolapse limits studies and effective outcome measurement in Afrikaans-speaking patients in SA.[15] A second problem is that the P-QOL was originally developed for a European population; this might raise the question as to its validity in a multicultural SA population.

The purpose of the study was therefore to validate the P-QOL in an Afrikaans-speaking SA population.

Methods

The observational, descriptive study was approved by the Ethics Committee of the Faculty of Health Sciences of the University of the Free State, Bloemfontein, SA. Written consent and permission were obtained from the participants and institutions where the study was conducted.

Translation of the P-QOL The original English P-QOL questionnaire was translated into Afrikaans by an independent medical translator. The translated Afrikaans version was reviewed and translated back into English by

SAJOG • December 2016, Vol. 22, No. 2

38

RESEARCH three independent urogynaecologists. The urogynaecologists agreed that the original content had been retained and that no ambiguity was present. No changes were made in the translated version.

Pilot study The translation was followed by a pilot study (n=5) to confirm the readability and participants comprehension of the questions.

Participants were asked to complete the questionnaire and were then interviewed by the researchers. No problems were identified and no changes were made to the translated Afrikaans version of the P-QOL. The same methodological procedures were followed as described for the main study in the following sections.

Sampling A convenience sample was used, consisting of 40 women meeting the inclusion criteria (Table 1). The eligibility criteria were aligned with the validated versions of the P-QOL, and applied to the demo­ graphics of a SA population.

Procedures The eligible participants had to complete an informed consent document, a P-QOL and a demographic data form after the study procedures had been explained to them by the researchers. The same urogynaecologist determined the POP-Q score for all participants according to clinical and ultrasonography findings. The completed questionnaires and forms were checked by the researchers to ensure that all the information was gathered.

Following consultation, each participant was given a second blank P-QOL in an addressed envelope, to complete and mail back after 2 weeks in order to determine the stability of the questionnaire by a test-retest analysis. The date on which the questionnaire had to be completed was indicated by a note on the envelope. Reminders were sent to all participants to complete the second questionnaire.

Data analysis The SAS software package (SAS, USA) and Excel (Version 2010) (Microsoft, USA) software were used for statistical analysis. The two questionnaires’ construct validity was determined by assessing the domain score of symptomatic and asymptomatic participants, and then comparing it with the POP-Q score to determine criterion validity by means of percentages. The test-retest method was used to indicate stability of the P-QOL, the kappa value to calculate the measure of agreement, and the Cronbach alpha to measure the internal consistency.

Descriptive statistics were used to explain the demographic data.

Medians and percentiles were calculated for continuous data, and fre­quencies and percentages were calculated to describe categorical data.

Results

A total of 40 women were enrolled into the study. Twenty-five women (64.1%, n=39) were symptomatic and 14 women (35.9%, Table 1. Eligibility criteria Inclusion criteria Women attending private practices and provincial outpatient clinics Women literate in Afrikaans Exclusion criteria Women <18 years Women >90 years Current pregnancy Childbirth or pelvic surgery in the past 6 months Active urinary tract infections Cognitive impairment

Table 2. Summary of demographic data of participants Symptomatic (n=25)

Asymptomatic (n=14)

Age (years), median

60

45.5

Body mass (kg) , median

67.5

80

Parity, n

(n=12)*

0

0

0 participants

1

1

3 participants

2

7

4 participants

≥3

17

5 participants

Method of delivery, n

(n=12)*

NVD

20

9 participants

Caesarean section

1

2 participants

Both

4

1 participant

Stage 0

0

14 (35.9)

Stage I

4 (10.3)

0

Stage II

7 (18.0)

0

Stage III

13 (33.33)

0

Stage IV

1 (2. 6)

0

POP-Q findings, n/N (%)

NVD = normal vaginal delivery.. *Two asymptomatic participants did not complete the parity question (including ‘Method of delivery’)

39

SAJOG • December 2016, Vol. 22, No. 2

RESEARCH Table 3. P-QOL domain scores of symptomatic and asymptomatic participants Symptomatic (median)

Asymptomatic (median)

Prolapse QOL domain

Assessment 1 score (n)

Assessment 2 score (n)

Assessment 1 score (n)

Assessment 2 score (n)

General health perceptions

25 (25)

25 (18)

25 (14)

25 (7)

Prolapse impact

33.33 (25)

33.33 (18)

33.33 (14)

66.67 (7)

Role limitations

33.33 (25)

33.33 (18)

0 (14)

0 (7)

Physical limitations

25 (24)

33.33 (18)

0 (13)

33.33 (7)

Social limitations

22.22 (13)

22.22 (12)

0 (9)

22.22 (4)

Personal relationships

16.67 (13)

33.33 (11)

0 (10)

0 (3)

Emotions

22.22 (25)

16.67 (18)

0 (14)

0 (7)

Sleep/energy

50 (25)

33.33 (18)

16.67 (14)

0 (7)

Severity measures

16.67 (24)

16.67 (18)

12.5 (14)

0 (7)

n=39) were asymptomatic, with one data form incomplete with regard to the POP-Q score. Table 2 shows a summary of the demographic and clinical characteristics of symptomatic and asymp­ tomatic participants..

Construct validity was determined by assessing the domain score of symptomatic and asymptomatic participants, and then comparing it with the POP-Q findings to determine criterion validity. The P-QOL domain scores of symptomatic participants were mostly higher compared with asymptomatic participants, indicating a poorer QOL (Table 3). Fourteen (35.9%) asymptomatic participants were classified as stage 0 on the POP-Q system and 64.1% of the symptomatic participants (n=25) had POP, ranging from stage II to IV. All asymptomatic participants (100%, n=14) were stage 0 on the POP-Q system and the majority of the symptomatic participants (52.0%, n=13/25) were at stage III. Stability was measured by means of the test-retest method to de­ter­ mine the consistency of the responses by participants (response rate 65%, n=40. All the reliability percentages were above 50% (Fig. 1).

According to the interpretation of Birkimer and Brown,[15] the median percentage of 69.2%, the sample size of 40, and the number of disagreements equalling <3 indicate a non-chance agreement.

The measure of agreement between specific questions was determined by means of calculating the kappa value (Fig. 2). The lowest kappa value was 0.12 for question 38, indicating poor strength in agreement. The highest kappa value was 0.65 for question 29, indicating good strength in agreement (κ range 0.61 - 0.80). The strength of agreement was fair in 4 questions (κ range 0.21 - 0.40), moderate in 11 (κ range 0.41 - 0.60) and good in 2 (κ range 0.6 1 0.80). The majority of questions, 13 of 20, therefore had a moderate to good strength of agreement.

The internal consistency was determined by means of the Cronbach alpha. A Cronbach alpha >0.7 is interpreted as acceptable.

Table 4 shows that assessments 1 and 2 had a mean Cronbach alpha score of 0.94, indicating very high reliability.

Discussion

Approximately half of the population in SA is female,[14]and POP may occur in up to three-quarters of them.[1] Pelvic floor disorders have been described as mainly being a QOL disorder.[16] Subjective improvement and improvement in QOL are main goals of manage­ ment of patients with POP (pre- and/or postoperatively) and need to be clinically evaluated by means of a valid outcome measure.[6,16]

Table 4. Internal consistency (Cronbach α coefficient) for P-QOL domains between assessment 1 and 2 Cronbach α Prolapse QOL domain

Assessment 1 score

Assessment 2 score

General health perceptions

*

*

Prolapse impact

*

*

Role limitations

0.89

0.88

Physical limitations

0.90

0.76

Social limitations

0.64

0.67

Personal relationships

0.61

0.40

Emotions

0.91

0.91

Sleep/energy

0.83

0.58

Severity measures

0.36

0.46

Mean score

0.94

0.94

*The domains ‘general health perceptions’ and ‘prolapse impact’ had only one item and the internal consistency could not be calculated.

This study determined validity and reliability aspects of an Afrikaans version of the P-QOL in a province in SA where Afrikaans is widely spoken. Content validity was determined in a similar way as described by previous validation studies of the P-QOL.[5] The content was found to be valid after review by a medical translator, panel of experts and pilot study.

The construct validity was indicated by the median domain scores of the symptomatic patients being higher than the asymp­tomatic patients’ scores, except for ‘general health perceptions’ and ‘prolapse impact’ scores. The results correlate with the findings of other validation studies which indicated statistical significant differences between the scores of symptomatic and asymptomatic patients. Lenz et al.[11] and De Oliveira et al.[12] also found the domain scores of ‘general health perceptions’ to be similar for both groups, because of the fact that this category consists of only one question and can be affected by symptoms or diseases not related to POP. A finding significant to this study was the lack in differences in scores relating to ‘prolapse impact’. The difference in the sample size for which each of these medians was calculated may affect the interpretation of these results. Another possibility might be that POP is multifactorial and can include an interaction and coexistence of several pelvic floor disorders affecting the experience of symptoms in even the minor (‘asymptomatic’) stages of POP.[2]

SAJOG • December 2016, Vol. 22, No. 2

40

RESEARCH 100 90 80

Reliabilty, %

70

69.2% 65.4% 65.4%

60

73.1%

73.1%

81.8%

76.9%

84.6% 76.9%

69.2%

69.2% 69.2%

65.4%

62.5% 53.8%

57.7%

53.8% 53.8%

50 40 30 20

Conclusion

10 0

Q1

Q2 Q21 Q22 Q23 Q25 Q26 Q27 Q28 Q29 Q30 Q31 Q32 Q33 Q34 Q35 Q37 Q37

Question number

Fig. 1. Test-retest reliability of the nine P-QOL domains. (*Test retest reliability could not be calculated for questions 24 and 36.)

0.7 0.6

Kappa value

The limited population also raised some concern as to the interpretation of the content validity found in this study. It must be taken into consideration that the descent of the Afrikaans-speaking participants included in this study may differ from the descent of Afrikaans-speaking patients in other demographical areas in SA. Clinical and cultural differences may affect the validity of an Afrikaans questionnaire, and it is therefore recommended that it should be tested in different regions in SA.

0.5

This study found the translated Afrikaans version of the P-QOL to have good content, construct and criterion validity, as well as very high stability, strength of agreement and internal consistency. This correlates with the validity and reliability of other translated versions of the P-QOL, supporting the evidence that the P-QOL is a high-quality disease-specific QOL questionnaire. It can be recommended that the P-QOL be translated into other African languages, and especially to determine content validity in the different African cultures.

0.4 0.3

Acknowledgements. Departments of Urogynaecology and Biostatistics of the University of the Free State, L Taute, Y Sutherland, P Ackerman, H Luckhoff, F Dry and C van Dyk for their field work.

0.2 0.1 Q38

Q37

Q34 Q35 Q36

Q31 Q32 Q33

Q26 Q27 Q28 Q29 Q30

Q21 Q22 Q23 Q24 Q25

Q1 Q2

0

Question number

Fig. 2. Kappa values indicating strength of agreement. (*Kappa values could not be calculated for questions 24 and 36.)

Lower domain scores also showed a relationship with less severe stages of POP, as assessed by the POP-Q system, indicating good criterion validity similar to the findings of previous studies.

The methods used in this study to calculate construct and criterion validity differed from most other studies which used the MannWhitney U test and the Spearman’s rho correlation, respectively. This difference in methodology limits specific comparison of some of the validity findings of this study with those of the previous studies, though the conclusive findings were similar. The use of different statis­ tical methods, all finding the same results, can however be seen as a strength to substantiate the findings from different validation studies.

Reliability measures correlate with previous findings on translated versions of the P-QOL.[2,5,8,10,12,13] Stability measures in this study indicated a non-chance agreement with the test-retest method, while the kappa value indicated moderate to good strength of agreement between questions, for the majority of the questions. A very high internal consistency was indicated by a mean Cronbach alpha of 0.94 during the first and second assessments.

Unfortunately the response rate for the second assessment was low as a result of poor patient compliance, even though patients were reminded about completing and sending back their questionnaires.

Despite this limitation, a statistical analysis was still possible for this relatively small population from which the sample was drawn.

41

Author contributions. CB was involved with protocol, project development, data collection and management, and manuscript writing, HSC with protocol, project development, and manuscript writing, and CvR with data analysis. The Afrikaans version of the P-QOL can be obtained from the corresponding author.

References 1. Cronjé HS. Pelvic Organ Prolapse. Bloemfontein: University of the Free State. http://www. topicsinmedicine.org/pelvic%20organ%20prolapse.html (accessed 16 April 2011).

2. Claerhout F, Moons P, Ghesquiere S, Verguts J, de Ridder D, Deprest J. Validity, reliability and responsiveness of a Dutch version of the prolapse quality-of-life (P-QoL) questionnaire. Int Urogynecol J 2012;21:569-578. http://dx.doi.org/10.1007/s00192-009-1081-9 3. Cronjé HS. Pelvic organ prolapse. In: Kruger TF, Botha MH, eds. Clinical Gynaecology, 4th ed. Cape Town: Juta & Co., 2011:487-515.

4. Fritel X, Varnoux N, Zins M, Breart G, Ringa V. Symptomatic pelvic organ prolapse at midlife, quality of life, and risk factors. Obstet Gynecol 2009;113(3):609-616. http://dx.doi.org/10.1097/ AOG.0b013e3181985312 5. Digesu GA, Santamato S, Khullar V, et al. Validation of an Italian version of the prolapse quality of life questionnaire. Eur J Obstet Gynecol Reprod Biol 2003;106(2):184-192.

6. Muller N. Pelvic organ prolapse: A patient-centred perspective on what women encounter seeking diagnosis and treatment. Aust N Z Cont J 2010;16(3):70-80.

7. Srikrishna S, Robinson D, Cardozo L, Gonzalez J. Is there a difference in patient and physician quality of life evaluation in pelvic organ prolapse? Int Urogynecol J 2008;19:517-520. http://dx.doi. org/10.1007/s00192-007-0477-7 8. Manchana T, Bunyavejchevin S. Validation of the prolapse quality of life (P-QOL) questionnaire in Thai version. Int Urogynecol J 2010;21:985-993. http://dx.doi.org/10.1007/s00192-010-1107-x 9. Digesu GA, Khullar V, Cardozo L, Robinson D, Salvatore S. P-QOL: A validated questionnaire to assess the symptoms and quality of life of women with urogenital prolapse. Int Urogynecol J 2005;16:176-181. http://dx.doi.org/10.1007/s00192-004-1225-x 10. Cam C, Sakalli M, Ay P, Aran T, Cam M, Karateke A. Validation of the prolapse quality of life questionnaire (P-QOL) in a Turkish population. Eur J Obstet Gynecol Reprod Biol 2007;135(1):132135. http://dx.doi.org/10.1016/j.ejogrb.2007.06.009 11. Lenz H, Stammer H, Brocker K, Rak M, Scherg H, Sohn C. Validation of a German version of the P-QOL questionnaire. Int Urogynecol J 2009;20:641-649. http://dx.doi.org/10.1007/s00192-009-0809-x 12. De Oliveira MS, Tamanini JTN, de Aguiar Cavalcanti G. Validation of the prolapse quality-of-life questionnaire (P-QoL) in Portuguese version in Brazilian woman. Int Urogynecol J 2009;20:11911202. http://dx.doi.org/10.1007/s00192-009-0934-6 13. Nojomi M, Digesu GA, Khullar V, et al. Validation of Persian version of the prolapse quality of life questionnaire (P-QOL). Int Urogynecol J 2012;23:229-233. http://dx.doi.org/10.1007/s00192-011-1529-6 14. Statistics South Africa. http://beta2.statssa.gov.za/?page_id=593 (accessed 27 February 2015).

15. Birkimer JC, Brown JH. Back to basics: Percentage agreement measures are adequate, but there are easier ways. J Appl Behav Anal 1979;12:535-543.

16. Lowenstein L, FitzGerald MP, Kenton K, et al. Patient-selected goals: The fourth dimension in assessment of pelvic floor disorders. Int Urogynecol J Pelvic Floor Dysfunct 2008;19(1):81-84.

SAJOG • December 2016, Vol. 22, No. 2

This open-access article is distributed under Creative Commons licence CC-BY-NC 4.0.

RESEARCH

Patient knowledge about and intention to use the intrauterine contraceptive device (IUCD) at a tertiary-level hospital N van der Westhuizen, MB ChB, Dip Obst (SA); G Hanekom, MMed (O&G), FCOG (SA) Department of Obstetrics and Gynaecology, Faculty of Health Sciences, University of the Free State, Bloemfontein, South Africa Corresponding author: N van der Westhuizen (nadiavdwest@gmail.com)

Background. The intrauterine contraceptive device (IUCD) is a highly effective and safe method of contraception. Prevention of unwanted pregnancies has made its use a matter of national priority in certain countries. Despite numerous advantages and suitability, the uptake of the IUCD is poor. Patients in South Africa (SA) seem to lack knowledge regarding this contraceptive.

Objectives. To determine the quantity and quality of knowledge about the IUCD, and to evaluate its acceptability for future use.

Methods. A prospective cross-sectional study was conducted at Pelonomi Tertiary Hospital. A total of 201 consecutive patients were interviewed using structured questionnaires: of these, 193 formed the final study group.

Results. Almost half (49.2%, n=95) of our patients were aware of the existence of the IUCD. Its use was very low, with only one patient having used it before. Overall qualitative knowledge was poor, even among those aware of the existence of the IUCD. There was a significant association between level of education and knowledge, with patients having passed grade 12 or higher significantly more likely to have knowledge of the IUCD than those at lower levels (relative risk 1.57, 95% confidence interval 1.18 - 2.08). Forty-five percent (n=86) of patients indicated a desire for future IUCD use.

Conclusion. Despite the availability of the IUCD in SA clinics and hospitals, its uptake is poor. Awareness of this method seems to have improved over the past few years, but the qualitative knowledge is still considerably lacking. Education plays a major role in the knowledge of contraception. Better educational aids at all facilities will increase its use and reduce unwanted pregnancies.

S Afr J Obstet Gynaecol 2016;22(2):42-46. DOI:10.7196/SAJOG.2016.v22i2.1048

The intrauterine contraceptive device (IUCD), a long-acting re­versible contraceptive (LARC), has been shown to be one of the most reliable contraceptive methods, which can be considered as effective as tubal ligation.[1] Advantages of the IUCD include reversibility, long-term efficacy and confidentiality.[2] It is also considered safe and effective for use in selected HIV-infected patients.[3] Worldwide there is a high prevalence of teenage and unwanted pregnancies with an increasing incidence every year,[4] and abortion rates in SA remaining virtually unchanged.[5] In a study performed in the USA, 42% of adolescents reported having been sexually active at least once in their lives. The methods of contraception used by this group, however, were mostly those with a relatively high failure rate with typical use, such as withdrawal, oral contraception or condom use.[6] Unintended pregnancies have been reported as being a result of low use of LARCs. These methods have been suggested to lower the rate of unwanted pregnancies, and the use of the IUCD has been made a national priority in the USA since 2009.[7] Currently the IUCD is the best method of contraception for high-risk obste­ tric patients. This group includes patients with previous venous thrombo-embolism, ovarian cancer, valvular heart disease, and those with chronic diseases like rheumatoid arthritis, as well as any other autoimmune disease.[3] It can also be offered as emergency contraception and is suitable for postpartum insertion, 10 minutes after delivery or even during a caesarean section, eliminating the risk to loss of follow-up for contraception compliance.[2]

A survey done in primary-care family-planning clinics in Cape Town[8] concluded that the knowledge of the IUCD as a contra­ ceptive method was very poor. Despite its availability, it was underused and not a preferred method to prevent pregnancy. It was shown that 41% of patients had heard about this method, but that only 4% had ever used it.[8] A national survey[9] in SA concluded that 66% of young women fell pregnant unintentionally as a result of not using any contraception. This was proposed to be due to gaps in the know­ ledge of how to use contraception correctly rather than a total absence of knowledge. School-based sex education in SA plays a significant role in the comprehensive strategy to influence adolescents toward positive sexual behaviour with regard to sexually transmitted diseases, HIV and pregnancy.[9] The primary objective of the study was to determine the knowledge, in terms of quantity and quality, about the IUCD as a method of contraception among pregnant patients attending the High Risk Obstetric Clinic at Pelonomi Tertiary Hospital in Bloemfontein, SA. The secondary objective was also to determine how many of these patients would be interested in using this device in future, after being given a short description of the advantages as well as disadvantages of the IUCD.

Methods

This was a prospective cross-sectional study performed on patients attending the High Risk Obstetric Clinic at Pelonomi Tertiary Hospital for the first time. Data were collected from January 2014 to

SAJOG • December 2016, Vol. 22, No. 2

42

RESEARCH November 2014 by interviewing consecutive patients. The only two inclusion criteria were that patients had to be pregnant and attending the clinic for the first time. Only patients refusing to be interviewed were excluded. Ethical approval for this study was obtained from the Ethics Committee of the Faculty of Health Sciences of the University of the Free State (ECUFS 207/2013).

Investigators conducted interviews with patients by means of a structured questionnaire in the language of their choice, English, Afrikaans or Sesotho. A pilot study including 20 patients was used to finalise the questionnaire. The interviews were done daily on patients attending the clinic for the first time. These patients were referred from local clinics, as well as district hospitals for secondary level antenatal care.

Informed consent was obtained, and the interview was conducted in a confidential consulting room. The data collection tool gathered demographic information, basic obstetric and gynaecological his­ tory and also established the baseline knowledge of different contraception methods. At this point in the interview an IUCD was shown to the patient without any description or explanation.

The patient was only informed that this was an IUCD or ‘loop’, as it is known colloquially. The interview then continued, focusing on collecting information about their general knowledge of the IUCD. After these questions all patients were given the same basic information about the IUCD, including some advantages and disadvantages. A different interviewer conducted the final part of the interview regarding the acceptability of the IUCD for future use. This was done to exclude bias in the form of false favourable responses to impress the original interviewer.

Sociodemographic and reproductive characteristics, and know­ ledge about and acceptability of the IUCD were described by frequencies and percentages. Bivariate analysis of specific variables of interest (e.g. education and knowledge of the IUCD), and associations between sociodemographic and reproductive charac­ teristics were performed using SAS version 9.3 (SAS Institute, USA) (by the Department of Biostatistics, University of the Free State) Table 1. Sociodemographic characteristics (n=193) n (%)

Characteristics Age (years)

and VassarStats (www.vassarastats.net) (by the researchers). Relative risks (RR) with 95% confidence intervals (CI) and χ2 or Fisher’s exact tests were performed as appropriate. P-values <0.05 were considered statistically significant.

Results

A total of 201 women were interviewed, of whom 8 were excluded because of consent issues. Data from 193 interviews were thus inclu­ ded and analysed.

Table 1 shows the sociodemographic characteristics of the participants. Their ages ranged from 18 to 49, with the majority between 20 and 39 (91.6%, n=175), and a mean of 31 years. Just under two-thirds (63.2%, n=122) of the women were unmarried; including those divorced and cohabiting. With regard to education only 4.1% (n=8) had no schooling, with almost a third (31.1%, n=60) having completed grade 12, and 20 patients (10.4%) having some form of tertiary education. The majority (68.9%, n=133) of patients were unemployed at the time of the interview.

The general gynaecological profile of the participants is illustrated in Table 2. About 86% (n=165) of patients reported having a regular menstrual cycle in the 6 months prior to their pregnancy, and 3.1% Table 2. General gynaecological profile n (%)

Characteristics Frequency of menstruation (n=193) Amenorrhoea (contraception-induced or other)

6 (3.1)

Regular monthly cycle

165 (85.5)

Irregular/unexpected vaginal bleeding

22 (11.4)

Nature of menstruation (n=187) Heavy

33 (17.6)

Normal

132 (70.6)

Light

22 (11.4)

Problems with menstruation (n=190) No

160 (84.2)

Yes (irregular, heavy, painful, long)

30 (15.8)

Pregnancies (n=193) 1

33 (17.1) 54 (28.0)

<20

6 (3.1)

2

20 - 29

76 (39.4)

3

51 (26.4)

30 - 39

99 (51.3)

4

26 (13.5)

40 - 49

12 (6.2)

≥5

29 (15.0)

Miscarriages (n=193)

Marital status Single (includes divorced, widow, lives with partner) Married

122 (63.2)

Yes

71 (36.8)

Number (n=50)

Education No education

8 (4.1)

50 (26.0)

1-2

48 (96.0)

>2

2 (4.0)

Termination of pregnancy (n=193)

Primary

22 (11.4)

Secondary school

83 (43.0)

Yes

4 (2.1)

Grade 12 passed

60 (31.0)

No

189 (97.9)

Tertiary

20 (10.4)

Unemployed

43

Yes

37 (19.2)

60 (31.1)

No

134 (69.4)

133 (68.9)

Don’t know

22 (11.4)

Occupation Employed

Intends to have future pregnancies (n=193)

SAJOG • December 2016, Vol. 22, No. 2

RESEARCH (n=6) had amenorrhoea. Eighty-two percent (n=154) had normal to light menstrual flow, with 15.8% regarding their menstrual pattern as problematic, citing heavy, irregular, painful or long cycles as their concern. Twenty-eight percent (n=54) of patients were pregnant with their second child, while 26% (n=51) were in their third pregnancy.

Fifteen percent (n=29) were in their fifth pregnancy or higher.

Previous spontaneous miscarriages were reported by a quarter (25.9%, n=50) of patients, and only four patients admitted to having had a previous termination of pregnancy. Future pregnancies were mostly not wanted (69.4%, n=134), but 19.2% (n=37) of patients desired more children.

The contraceptive most patients were familiar with was the male condom (99.5%, n=192) followed by injectable contraception (97.9%, n=189), with 83.9% and 80.8%, respectively, having used Table 3. Knowledge about and use of contraception (n=193) Type of contraception

Heard about, n (%)

Used before, n (%)

Female sterilisation (tubal ligation)

142 (73.5)

0 (0)

Male sterilisation (vasectomy)

45 (23.3)

0 (0)

IUCD (loop)

95 (49.2)

1 (0.5)

Oral contraceptive

177 (91.7)

71 (36.8)

Progesterone-only pill

3 (1.6)

1 (0.5)

Emergency contraception

92 (47.7)

31 (16.1)

Injection (Depo Provera/Nur-Isterate)

189 (97.9)

156 (80.8)

Male condom

192 (99.5)

162 (83.9)

Female condom

148 (76.7)

20 (10.4)

Spermicides/jelly

3 (1.6)

0 (0)

Diaphragm/cap

2 (1.0)

0 (0)

Hormone implants

21 (10.9)

0 (0)

Natural methods

12 (6.2)

3 (1.6)

Table 4. Quantitative knowledge of the IUCD related to level of education and gravidity Patients, n (%)

Heard about IUCD, n (% in that group)

o education + N primary school

30 (15.5)

8 (26.7)

Secondary school

83 (43.0)

37 (44.6)

Grade 12 passed

60 (31.1)

37 (61.7)

Tertiary education

20 (10.4)

13 (65.0)

193

95 (49.2)

1

33 (17.1)

12 (36.4)

2

54 (28.0)

23 (42.6)

3

51 (26.4)

29 (56.9)

â&#x2030;Ľ4

55 (28.5)

31 (56.4)

193

95 (49.2)

Level of education

Total Gravidity

Total

these methods before. Knowledge about less common contraceptive methods like vasectomy, progesterone-only pill, spermicides, diaphragm cap, hormonal implant and natural methods ranged between 2% and 23% (Table 3).

With regard to the IUCD, 95 patients (49.2%) reported having heard about it, but only a single patient (0.5%) was found to have used it before. Twenty-five percent (n=49) of patients claimed to know how the IUCD works, and 23.3% (n=45) could give an explanation. Qualitative data analysis revealed that most patients knew it was a device that prevents pregnancy, but overall the correct method could not be explained.

Table 4 illustrates the association between the number of patients having knowledge of the IUCD and their level of education and gravidity, respectively. When the patients were divided into those with grade 12 or a higher level of education and those without, a statistically significantly higher percentage of patients with knowledge were observed with a higher level of education (63% v. 40%, p<0.01). Patients with grade 12 or a tertiary qualification were more likely to have knowledge regarding the IUCD compared with patients with a lower level or no education (RR 1.57, 95% CI 1.18 - 2.08). Higher gravidity, 3 or more, was associated with more knowledge about the IUCD (57% v. 40%, p=0.02). Patients with gravidity of 2 or less were statistically less likely to have knowledge about the IUCD compared with those with a gravidity of 3 or more (RR 0.71, 95% CI 0.52 - 0.96).

The qualitative assessment of knowledge among those participants who claimed to be familiar with the IUCD (n=95) revealed that their overall knowledge of the IUCD was poor. Noteworthy findings include that more than a third of participants (35.7%, n=34) felt that unmarried women may not use, or were unsure if they could use, this method of contraception. More than half of the patients (58.9%, n=56) were of the opinion that women without children cannot use the IUCD. Seventy-three percent (n=69) were confident that it is safe to use the IUCD while having many sexual partners.

Five patients were convinced that pregnant women can also use this method of contraception. Two-thirds (66,3%, n=63) were aware that it is possible for HIV-positive women use the IUCD. As mentioned, the results listed above indicated the quality of knowledge of patients who claimed to be familiar with the IUCD. Interpreting these findings as part of the whole study group shows an even poorer overall knowledge.

Multiple true or false questions revealed poor understanding, as well as the myths surrounding the IUCD. A third (33.1%, n=64) of participants believed the IUCD causes cancer and 38.3% (n=74) that it moves around in the body. Forty-one percent of patients were unaware of its duration of action.

At the conclusion of the interview 44.6% (n=86) of the patients were keen on using the IUCD in future with 51.3% (n=99) not interested.

Main reasons for lack of interest were cited as a desire to be sterilised after delivery or wanting more information to make an informed decision. The most appealing factors of the IUCD mentioned were its efficacy, duration of action and convenience (Table 5).

Discussion

Numerous surveys and cross-sectional studies have been done in SA evaluating the knowledge about, attitudes to and acceptability of the IUCD. To our knowledge none of these studies was done in the Free State, nor have any of these investigated high-risk pregnant women.

SAJOG â&#x20AC;˘ December 2016, Vol. 22, No. 2

44

RESEARCH Table 5. Interest in future IUCD use n (%) IUCD in future? (n=193) Yes

86 (44.6)

No

99 (51.3)

Don’t know

8 (4.1)

Appealing factors (n=86) Efficacy

74 (86.0)

Duration of efficacy

71 (82.6)

Convenience

60 (69.8)

Not permanent

20 (23.3)

Confidentiality

12 (14.0)

Minimal effect on hormones

12 (14.0)

Other

10 (11.6)

Unappealing factors (n=99) Not permanent

70 (70.7)

Unfavourable side-effects

35 (35.4)

No protection against STDs and HIV

22 (22.2)

Wants another child soon

7 (7.1)

Wants monthly cycles

6 (6.1)

Other

43 (43.4)

According to the Sexual and Reproductive Health report[10] of October 2014, free contraception should be available to all public healthcare users. Rates of unplanned and unwanted pregnancies are still very high and associated with limited access to and lack of knowledge about contraception among the youth. Currently the IUCD is not provided at many health facilities because of insufficient training of healthcare providers.[10] This could explain the very low rate of use (0.5%) in our study population, even though according to the World Health Organization contraception eligibility criteria, most of our high-risk obstetric patients qualify for its use.[3] Understandably, our study population fits the age group of repro­ ductive women. The majority were unmarried and their level of education was similar to that of the general population of SA. When we compared our findings with those of the 2011 census, 4.1% in the study group v. 8.6% in the general population had no formal education, while 31.0% compared with 28.5% had passed grade 12, whereas 10.4% v. 12.1% had tertiary education.[11] As school education plays a major role in sexual development and reproductive health, one would expect the participants to have more knowledge about contraception. Lack of knowledge was clearly identified in our study, highlighting the vital role of reproductive health education in our schools. The high unemployment rate among study participants could be explained by their low level of education.

The general gynaecological profile of our participants was nor­ mal. This finding is expected in a pregnant study population, indi­ cating previous normal ovulatory cycles and an absence of gross reproductive and gynaecological pathology. The rate of termination of pregnancy was unexpectedly low (2.1%) compared with the 9.9% of the provincial statistics for the Free State of 2010.[12] This could possibly indicate selection bias as our study population was pregnant,

45

and most participants wanted to have children, and were possibly less likely to have had a previous termination.

The overall awareness about contraception appears to be accep­ table. As expected, the male condom was well known and most commonly used, followed by injectable contraception. Awareness of the IUCD, however, was less impressive, with less than half (49.2%) of the study population having heard about this method. This is higher than a similar study done in the Western and Eastern Cape, with awareness only 26% combined.[13] A possible explanation for this is the timeframe of data collection of the latter study. The study was conducted in 2006, which indicates that awareness of the IUCD could possibly have increased in the past 6 - 7 years, with better school education and overall awareness in our public service. This supports the finding of our study that a higher level of education is associated with better knowledge about the IUCD. Qualitative knowledge, however, was poor, and the majority of patients were ignorant of contraceptive methods, as well as the eligibility criteria for their use. The existence of various myths surrounding IUCD use was evident and significantly higher com­ pared with studies conducted in the Western and Eastern Cape.

Nine percent of the 53 women who had heard about the IUCD in the above-mentioned study had misconceptions or incorrect information that influenced them regarding IUCD use, compared with our high percentages that were convinced that IUCDs caused cancer or moved around in the body.[13] This indicates that even if women are aware of the method of contraception, the quality of knowledge is poor and is a matter of concern.

Forty-five percent (n=86) of patients expressed an interest in using the IUCD in future following minimal education during the interview, compared with 74% in a Cape Town survey.[13] This suggests that with more education, especially focusing on finer details and starting at school level, the use of this method might increase and help decrease the number of unwanted as well as adolescent pregnancies.

Study limitations The study was performed in a tertiary hospital on high-risk obstetric patients and thus cannot be regarded as representative of the general female public. These patients had high-risk pregnancies and came into contact with healthcare providers more often than those at lower risk. Therefore they might have better knowledge about contraception compared with other women, indicating potential sampling bias.

Conclusion

Even though the IUCD is clinically regarded as an excellent method of contraception, the overall use in our setting is very low. One of the many problems is certainly a lack of education and, more importantly, lack of detailed knowledge among our patients. This demonstrates significant shortcomings in the reproductive health education of our population. Possible solutions could include estab­ lishing proper guidelines and women’s health information aids for schools, clinics, hospitals and reproductive health centres. Healthcare providers need to familiarise themselves with and be trained in this method to ensure its uptake. This will certainly decrease the rate of unwanted as well as adolescent pregnancies.

Acknowledgements. The authors would like to thank Prof. Gina Joubert of the Department of Biostatistics and the School of Medicine of the University of the Free State for her help with the statistical analysis. Thanks are also due to

SAJOG • December 2016, Vol. 22, No. 2

RESEARCH the staff of Pelonomi High Risk Obstetric Clinic for their assistance during this study. There was no funding for this study.

References 1. Mishell DR. Intrauterine devices: Mechanism of action, safety, and efficacy. Contraception 1998;58:45S-53S. http://dx.doi.org/10.1016/S0010-7824(98)00082-1 2. Espey E, Singh RH. Practice bulletin: Long-acting reversible contraception: Implants and intrauterine devices. The American College of Obstetricians and Gynecologists Women’s Health Care Physicians.

Obstet Gynecol 2011;118(1):184-196. http://dx.doi.org/10.1097/AOG.0b013e318227f05e 3. Frieden TR, Briss PA, Stephens JW, et al. U.S. Medical Eligibility Criteria for Contraceptive Use, 2010. Mor Mortal Wkly Rep CDC Surveill Summ 2010;59:e1-88. http://www.cdc.gov/mmwr/pdf/ rr/rr59e0528.pdf (accessed 8 April 2015).

4. Stanwood NL, Bradley K. Young pregnant women’s knowledge of modern intrauterine devices. The American College of Obstetricians and Gynecologists Women’s Health Care Physicians. Obstet Gynecol 2006;188(6):1417-1422. http:/dx.doi.org/10.1097/01.AOG.0000245447.56585.a0 5. Johnston R. South Africa: Abortion rates by province, 1997 - 2012 (updated 23 January 2015). http:// www.statssa.gov.za/publications/P0305/P03052012.pdf (accessed 12 May 2015).

6. American College of Obstetricians and Gynecologists Committee on Adolescent Health Care LongActing Reversible Contraception Working Group. ACOG Committee opinion no. 539: Adolescents and long-acting reversible contraception: Implants and intrauterine devices. Obstet Gynecol 2012;120(4):983-988. http://dx.doi.org/10.1097/AOG.0b013e3182723b7d

7. American College of Obstetricians and Gynecologists Committee on Gynecologic Practice; LongActing Reversible Contraception Working Group. ACOG Committee opinion no. 450: Increasing use of contraceptive implants and intrauterine devices to reduce unintended pregnancy. Obstet Gynecol 2009;114(6):1434-1438. http://dx.doi.org/10.1097/AOG.0b013e3181c6f965 8. Van Zijl S, Morroni C, van der Spuy ZM. A survey to assess knowledge and acceptability of the intrauterine device in the Family Planning Services in Cape Town, South Africa. J Fam Plann Reprod Health Care 2010;36(2):73-78. http://dx.doi.org/10.1783/147118910791069367 9. Henry J. Kaiser Family Foundation. Young South Africans, broadcast media and HIV/AIDS awareness: Results of a National Survey. March 2007;e24-30. https://kaiserfamilyfoundation.files. wordpress.com/2013/01/7614.pdf (accessed 8 April 2015).

10. National Department of Health, South Africa. Sexual and Reproductive Health. UNFPA South Africa. 2014. http://countryoffice.unfpa.org/southafrica/2011/11/24/4255/reproductive_health_ and_hiv/ (accessed 8 April 2015).

11. South African Educational Statistics. http://www.southafrica.info/about/facts.htm#education (accessed 8 April 2015).

12. Johnston WM. South Africa: Abortion rates by province, 1997 - 2012. http://www.johnstonsarchive. net/policy/abortion/southafrica/ab-sap2.html (accessed 8 April 2015).

13. Gutin SA, Mlobeli R, Moss M, et al. Survey of knowledge, attitudes and practices surrounding the intrauterine device in South Africa. Contraception 2011;83(2):145-150. http://dx.doi.org/10.1016/j. contraception.2010.07.009 14. National Department of Health, South Africa. National Contraception Clinical Guidelines.

2012. http://www.health-e.org.za/wp-content/uploads/2014/05/National-contraception-familyplanning-policy.pdf (accessed 8 April 2015).

SAJOG • December 2016, Vol. 22, No. 2

46

This open-access article is distributed under Creative Commons licence CC-BY-NC 4.0.

RESEARCH

Birth preparedness and complication readiness among pregnant women in a rural community in southern Nigeria S H Ibadin, MBBS; V Y Adam, MBBS, MPH, MWACP, FMCPH; O A Adeleye, MBBS, MPH, FWACP; O H Okojie, MBBS, FWACP, FMCPH Department of Community Health, University of Benin Teaching Hospital, Benin City, Nigeria Corresponding author: S H Ibadin (serbalogun@gmail.com)

Background. Birth preparedness and complication readiness (BPACR) has been advocated as a strategy to overcome costly delays in decision-making to seek skilled attendance at delivery, which in turn contribute significantly to maternal mortality from obstetric causes.

Objective. To assess the determinants of BPACR among pregnant women in a rural community in Edo State, Nigeria.

Methods. A descriptive cross-sectional study was done in Anegbette, a rural community in Etsako Central Local Government Area of Edo State. A house-to-house survey was carried out to identify pregnant women and all eligible women in the study area were included in the study.

Results. A total of 277 pregnant women participated in the study. The mean age (standard deviation) of respondents was 28.7 (5.8) years. Less than half (134, 48.4%) of the respondents were well prepared while 143 (51.6%) were poorly prepared. After adjustment for confounding effect using binary logistic regression analysis, educational level (odds ratio (OR) 0.653, 95% confidence interval (CI) 0.330 - 0.956), occupation (OR 0.384, 95% CI 0.148 - 0.990) and utilisation of antenatal care (OR 3.407, 95% CI 1.830 - 5.074) were significant predictors of BPACR. Conclusion. BPACR was poor among women in the rural community. In order to improve maternal health among rural women in Nigeria, government and donor agency funding for safe motherhood programmes should focus on female empowerment and encourage community participation towards promotion of maternal health.

S Afr J Obstet Gynaecol 2016;22(2):47-51. DOI:10.7196/SAJOG.2016.v22i2.1088

Birth preparedness and complication readiness (BPACR) has been advocated as a strategy to overcome costly delays in decision-making to seek skilled attendance at delivery, which in turn contribute significantly to maternal mortality from obstetric causes.[1] BPACR entails identifying a skilled provider and a birth location, learning to recognise the danger signs that may indicate life-threatening com­ plications for the mother and baby, saving money and arranging for transportation, identifying a blood donor, identifying the nearest emer­gency obstetric services should pregnant women, their families and communities need to seek assistance in case of emergencies.[2] In addition, BPACR requires health providers and facilities to be prepared to attend births and treat complications.[2] BPACR among pregnant women is significantly influenced by their socioeconomic characteristics, among other personal factors. A study conducted among women attending antenatal care in south eastern Nigeria found that although 70.6% of women were aware of BPACR, knowledge of key danger signs in pregnancy was low; educational status was identified to be the best predictor of BPACR.[3] Similarly, having a higher education, upper socioeconomic status and being married were identified as factors associated with good BPACR among women receiving antenatal care in Benin City[4] and Ile-Ife,[5] Nigeria.

A community-based survey in northern Nigeria among 5 083 rural women recruited from three states found that BPACR practices were generally poor, with 33.2% having made no preparation for delivery; fewer than 2.5% had made provisions for a trip to the health facility at delivery and only 32.0% knew any critical danger sign relating to pregnancy and delivery.[6]

47

Most BPACR studies in southern Nigeria have been carried out among antenatal clinic attendees and may not necessarily provide a true representation of the actual BPACR practices of women living in rural communities where antenatal clinic attendance is suboptimal. Community-based surveys that reflect the BPACR practices of rural women appear to be lacking. Studying BPACR among rural women is necessary in order to highlight barriers that may impede good BPACR practices and provide a basis for specific interventions targeted at increasing utilisation of skilled birth attendants and improving maternal outcomes in rural communities.

This study assessed the determinants of BPACR among pregnant women in a rural community in Edo State, Nigeria.

Methods

Study setting This study was conducted in Anegbette, a rural community in Etsako Central Local Government Area (LGA) of Edo State, Nigeria.

Anegbette had a total population of 8 180, of whom 1 799 women were of child-bearing age, in 2011.[7] The community has a functional primary healthcare centre and is also served by a comprehensive health centre in a nearby town, Ekperi, and a general hospital located in the LGA headquarters, Fugar.

Study design and selection of subjects A descriptive cross-sectional study was carried out. The study population consisted of all consenting pregnant women who were permanent residents of Anegbette. A minimum sample size of 170

SAJOG â&#x20AC;˘ December 2016, Vol. 22, No. 2

RESEARCH was calculated using a prevalence of 87.4% (the proportion of women with good BPACR from a similar study carried out in Edo State, Nigeria),[4] standard normal deviation of 1.96 at 95% confidence level and power set at 80%. A house-to-house survey was carried out to identify pregnant women and all eligible women in the study area were included in the study.

Measures and data management A structured questionnaire was designed in line with the study objectives and administered by trained interviewers to pregnant women in their homes. The questionnaire was used to collect infor­ mation on sociodemographic characteristics (age, educational level, marital status, religion, and occupation), utilisation of antenatal care (ANC) and BPACR practices. Utilisation of antenatal clinics was assessed solely by self-reporting. BPACR was assessed using questions adapted from the JHIPEGO BPACR matrix.[2] Good knowledge of danger signs in pregnancy and labour was assessed by the respondents’ awareness of at least three of the six symptoms in the matrix, namely: bleeding, liquor drainage before term, foulsmelling vaginal discharge, severe abdominal pain before term, headaches and fits. These symptoms signal potential complications of pregnancy. Women who met at least four of the eight BPACR criteria, which include: good knowledge of danger signs, saving money towards delivery, purchased items for the baby, identified a health facility for delivery, arranged for accompaniment in case of emergencies, made plans for transportation, identified a blood donor or blood bank, and identified the nearest facility for emergency obstetric care, were classified as being well prepared.

Data were analysed using IBM SPSS version 20.0 software (IBM, USA). Sociodemographic characteristics and utilisation of ANC were the independent variables and the main outcome measure was BPACR. Chi-squared test was used to determine the significance of associations between the independent variables and BPACR. All variables were included in a binary logistic regression model to determine their independent effects. Odds ratios (ORs) and their respective 95% confidence intervals (CIs) were obtained. Statistical significance was set at p<0.05.

quarters (232, 83.7%), of the women were aged between 15 and 34 years, and 260 (93.9%) were married. Among the participants, 118 (42.6%) had no formal education while 116 (41.9%), 41 (14.8%) and 2 (0.7%) had attained primary, secondary and tertiary education, respectively. Most (228, 82.3%) of the women were unskilled workers, 247 (89.2%) were Etsako (the predominant ethnic group), and 224 (80.9%) were Christians (the most practised religion).

Birth preparedness and complication readiness As shown in Table 2, 134 (48.4%) pregnant women were well pre­ pared while 143 (51.6%) were poorly prepared. More than half the respondents were saving money for delivery (158 (57.0%)); 147 (49.5%) had identified a health facility and skilled birth attendants for delivery, 135 (48.7%) had arranged for accompaniment in case of emergencies and 95 (34.3%) had made arrangements for transportation. Identification of blood donors or blood banks and facilities for emergency obstetric care was poor, with less than 15% of respondents having carried out these actions.

Knowledge of danger signs in pregnancy was relatively high, with 173 (62.5%) of women knowing at least three symptoms that signal Table 1. Sociodemographic characteristics of respondents (N=277) Variables

Categories

n (%)

Age (years)

15 - 24

75 (27.1)

25 - 34

157 (56.7)

35 - 44

42 (15.2)

45 - 49

3 (1.1)

No formal education

118 (42.6)

Primary

116 (41.9)

Secondary

41 (14.8)

Tertiary

2 (0.7)

Single

6 (2.2)

Married

260 (93.9)

Divorced

3 (1.1)

Cohabiting

8 (2.9)

Unemployed

22 (7.9)

Unskilled

228 (82.3)

Semi-skilled

2 (0.7)

Skilled

25 (9.0)

Christianity

224 (80.9)

Islam

48 (17.3)

ATR

5 (1.8)

Etsako

247 (89.2)

Esan

14 (5.1)

Others*

16(5.7)

Educational level

Marital status

Ethical consideration Permission to carry out the study was granted by the chairman of Etsako Central LGA and the village head of Anegbette community.

Verbal informed consent to participate in the study was obtained from each respondent. At the end of the interview, the participants were health educated on BPACR and the need to utilise maternal health services, especially skilled birth attendants.

Occupational classification

Limitation The terms ‘good knowledge’ and ‘well prepared’ as used to express the authors’ assessment of knowledge of danger signs in pregnancy and BPACR practices are discretionary because of lack of uniform standards. The meaning of these terms may therefore vary signi­ ficantly from previous usage in other similar literature.

Results

Religion

Ethnicity

Sociodemographic characteristics A total of 277 pregnant women participated in the study. The socio­demographic characteristics of the respondents are shown in Table 1. The age of respondents ranged from 17 to 46 years, with a mean age (standard deviation (SD)) of 28.7 (5.8) years. Over three-

ATR = African traditional religion. *Others include Benin, Esan, Fulani, Isoko and Ijaw.

SAJOG • December 2016, Vol. 22, No. 2

48

RESEARCH potential complications; however, 59 (21.3%) of the women did not know any danger signs (Table 3). The most frequently recognised symptoms among all respondents were: bleeding 173 (62.5%), liquor drainage before term 167 (60.3%) and abdominal pain before term 166 (59.9%). Only 86 (31.0%) knew that severe headache was a danger sign in pregnancy.

A total of 165 (59.6%) of women had registered for ANC at a health facility at the time of the study.

BPACR was significantly associated with educational level (χ2= 7.262, p=0.007) and occupation (χ2=7.911, p=0.005); with BPACR being better among women who had attained at least primary education (54.1%) and women who were semi-skilled or skilled workers (74.1%), as compared with those who had no formal education (32.2%) and were unskilled or unemployed (45.6%). A higher proportion of women who had registered for ANC were well prepared (61.8%) as against women who were unbooked (28.6%); this association was also statistically significant (χ2=29.528, p<0.001). Women who were married also had better BPACR practices than unmarried and separated women; however, this association was not statistically significant (Table 4).

Table 2. Birth preparedness and complication readiness among respondents (N=277) BPACR criteria

n (%)*

Had good knowledge of danger signs

173 (62.5)

Saving money towards delivery

158 (57.0)

Purchased items for the baby

147 (53.1)

Identified a health facility for delivery

137 (49.5)

Arranged for accompaniment in case of emergencies†

135 (48.7)

Made plans for transportation

95 (34.3)

Identified blood donor or blood bank

34 (12.3)

Identified nearest facility for EmOC‡

30 (10.8)

Well prepared

134 (48.4)

Poorly prepared

143 (51.6)

*Multiple responses allowed. † Among the 135 women who had arranged for accompaniment, 127 (94.1%) reported that their husbands would accompany them if they had an emergency. ‡ Emergency obstetric care.

Table 3. Knowledge of danger signs that may signal complications among respondents (N=277) Variable

n (%)*

Knowledge of symptoms Bleeding

187 (67.5)

Liquor drainage before term

167 (60.3)

Abdominal pain before term

166 (59.9)

Fits

156 (56.3)

Foul-smelling vaginal discharge

139 (50.2)

Severe headaches

86 (31.0)

Overall knowledge Good knowledge

173 (62.5)

Poor knowledge†

104 (37.5)

Multiple responses allowed.

Fifty-nine women (21.3%) did not know any danger sign in pregnancy.

*

†

49

After adjustment for confounding effect using binary logistic re­gress­ ion analysis, educational level (OR 0.653, 95% CI 0.330 - 0.956), occupation (OR 0.384, 95% CI 0.148 - 0.990) and utilisation of ANC (OR 3.407, 95% CI 1.830 - 5.074) were significant predictors of BPACR, with women who had no formal education or skills being less likely to be well prepared and women who had registered for ANC being more likely to be well prepared (Table 5).

Discussion

This study revealed that approximately two-thirds of the study popu­ lation had good knowledge of key danger signs in pregnancy, while one-fifth had no knowledge of danger signs. Although bleeding was the most frequently mentioned sign, less than 10% of women had identified potential blood donors/blood bank services or a facility where they could readily receive emergency obstetric care. In northern Nigeria,[6] Ethiopia[8] and Uganda[9] researchers also found that a significant proportion of rural women did not have adequate knowledge of signs that may signal complications and the majority of pregnant women failed to take the aforementioned BPACR actions. This suggests that many pregnant women may not understand the implications or severity of these symptoms, should they occur, or the extent to which they need to make advance preparations to avoid delays that may arise if they need special care during pregnancy and delivery.

Furthermore, over half the pregnant women in Anegbette com­ munity had poor BPACR. In contrast, another study among antenatal clinic attendees in Edo State found that the majority of the women interviewed (87.4%) were well prepared.[4] This discrepancy may be accounted for by the difference in the study population. Data from this study also suggest that BPACR practices in Anegbette were remarkably better than those observed among rural women in northern Nigeria, where less than 2.5% of those surveyed had made any provision for a trip to the health facility, saved money or made transport arrangements, compared with at least 34.3% of women in this study.[6] BPACR practices have also been documented to be poor among rural women in Ethiopia[8] and Tanzania.[10] Poor BPACR has the potential to negatively influence the maternal and child health outcomes, as women who are not well prepared are less likely to utilise skilled birth attendants at delivery or respond to complications in an appropriate or timely manner.

Utilisation of ANC was found to be a significant determinant of BPACR. This may be as a result of the inclusion of BPACR messages in routine service delivery by skilled providers, and underscores the need for the improvement of antenatal coverage in rural communities. It was also observed that women who were uneducated or unskilled were disadvantaged compared with their counterparts who had attained at least primary education or were engaged in occupations requiring some level of skill. This observation is in consonance with findings of similar studies carried out in Edo[4] and Osun[5] states in Nigeria, where educational level was also found to be significantly associated with BPACR. The role of education in the improvement of maternal outcomes has been well documented globally. Female education and empowerment have been proven to positively influence women to become more knowledgeable about their reproductive health, improve their decision-making capabilities, engage in beneficial health practices and increase utilisation of maternal health services.[11] Providing educational and skill acquisition opportunities for rural women is thus imperative in order to improve BPACR and promote utilisation of skilled attendants at every delivery.

SAJOG • December 2016, Vol. 22, No. 2

RESEARCH Table 4. Birth preparedness and complication readiness by sociodemographic characteristics and utilisation of antenatal care BPCR Well prepared (n=124), n (%)

Poorly prepared (n=153), n (%)

Total (N=277), n (%)

Ď&#x2021;2

p-value

15 - 34

109 (47.0)

123 (53.0)

232 (83.7)

1.109

0.292

35 and above

25 (44.4)

20 (55.6)

45 ( 16.3)

No formal education

46 (39.0)

72 (61.0)

118 (42.6)

7.262

0.007

At least primary education

88 (55.3)

71 (44.7)

159 (57.4)

Unemployed and unskilled

114 (45.6)

136 (54..4)

250 (90.3)

7.911

0.005

Semi-skilled and skilled

20 (74.1)

7 (25.9)

27 (9.7)

Married

129 (49.6)

131 (50.4)

260 (93.9)

2.608

0.106

Not married/Separated

5 (29.4)

12 (70.6)

17 (6.1)

Christianity

116 (51.8)

108 (48.2)

224 (80.9)

5.452

0.020

Others

18 (34.0)

35 (66.0)

53 (19.1)

Yes

102 (61.8)

63 (38.2)

165 ( 59.6)

29.528

<0.001

No

32 (28.6)

80 (71.4)

139 (40.4)

Variables Age

Educational level

Occupational classification

Marital status

Religion

Registered for ANC

Although marital status was not significantly associated with BPACR, married women were observed to have better practices than unmarried or separated women. Spousal support may have contributed to this finding as the majority of the women reported that they would be accompanied by their husbands if they had an emergency. While this response does not necessarily reflect a high level of male participation, it does imply that married women receive additional support as compared with single respondents.

Studies in Edo State[12,13] and Northern Nigeria[14] have however highlighted that male participation in maternal health issues is undesirably low and requires improvement. This highlights an opportunity that exists for improving BPACR and maternal health in rural communities through the education of men on their role in safe motherhood.

In addition to spousal and family support, the community also has a crucial part to play in ensuring pregnant women are well prepared and ready for any complications that may occur as BPACR is a shared responsibility. The majority of women in this study population were unaware of any community support mechanism/programme for pregnant women. This indicates that Anegbette community may not have an organised support system to cater for the special needs of pregnant women. This phenomenon was equally observed in other studies conducted in Edo State[4,12] and suggests that community participation towards improving maternal health has been overlooked. Lack of community participation is likely to hinder improvements in safe motherhood, especially in resource-poor countries like Nigeria

Table 5. Predictors of birth preparedness and complication readiness OR

95% CI

p-value

1.104 Reference

0.968 - 1.063

0.548

Educational level No formal education 0.561 At least primary education Reference

0.330 - 0.956

0.033

Social class by occupation Unemployed/unskilled Semi-skilled/skilled

0.378 Reference

0.144 - 0.990

0.048

Marital status Married Not married/separated

1.558 Reference

0.501 - 4.844

0.443

Religion Christianity Others

1.720 Reference

0.867 - 3.412

0.121

Registered for ANC Yes No

3.599 Reference

2.109 - 6.144

<0.001

Variables Age 15 - 34 35 and above

SAJOG â&#x20AC;˘ December 2016, Vol. 22, No. 2

50

RESEARCH where women are often socially and economically disadvantaged.

Community participation plays a key role in health promotion, has been instrumental in bringing about appreciable gains in maternal health indices among poor people living in rural areas in India, Nepal and Bangladesh,[15] and may be an effective way of improving BPACR and utilisation of skilled birth attendants among rural women in Nigeria.

Conclusion

This study found that BPACR was poor among women in Anegbette community. Educational level, occupation and utilisation of antenatal clinics were the determinants of BPACR In order to improve maternal health among rural women in Nigeria, government and donor agency funding for safe motherhood programmes should be geared towards capacity building at the local level by encouraging the formation of community support groups and co-operative societies to educate and empower rural women economically and encourage utilisation of maternal health services.

Maternal health campaigns in Nigeria should also focus on the participatory role of the community as a whole in ensuring safe motherhood.

Acknowledgement. The authors are grateful to the people of Anegbette community and also wish to express appreciation to the final-year medical students who assisted with data collection.

51

References 1. World Health Organization. Birth and Emergency Preparedness in Antenatal Care: Integrated Management of Pregnancy and Childbirth. Geneva: WHO Department of Making Pregnancy Safer 2006. http://www.who.int/reproductivehealth/publications/maternal_perinatal_health/emerggency_ preparedness_antenatal_care.pdf (accessed 24 April 2015).

2. Maternal and Neonatal Health (MNH) Program. Birth Preparedness and Complication Readiness: A Matrix of Shared Responsibilities. Baltimore, MD: MNH Program, 2001;1 - 7.

3. Ekabua JE, Ekabua KJ, Odusola P, Agan TU, Iklaki CU, Etokidem AJ. Awareness of birth preparedness and complication readiness in South Eastern Nigeria. ISRN Obstet Gynaecol 2011;e1-6. http://dx.doi. org/10.5402/2011/560641 4. Tobin EA, Ofili AN, Nkemka E, Enueze O. Assessment of birth preparedness and complication readiness among women attending primary health care centres in Edo State, Nigeria. Ann Nigerian Med 2014;8:7681. http://dx.doi.org/10.4103/0331-3131.153358 5. Abioye Kuteyi EA, Kuku JO, Lateef IC, Ogundipe JA, Mogbeyteren E, Banjo MA. Birth preparedness and complication readiness among women attending three levels of health facilities in Ife Central Local Government, Nigeria. J Community Med Primary Health Care 2011;23(1):41-54. http://www.ajol.info/ index.php/jcmphc/article/view/84664/74652 6. Doctor HV, Findley SE, Cometto G, Afenyadu GY. Awareness of critical danger signs of pregnancy and delivery, and preparations for delivery, and utilization of skilled birth attendants in Nigeria. J Health Care Poor Underserved 2013 24(1):152-170. http://dx.doi.org/10.1353/hpu.2013.0032] 7. Etsako Central LGA. Projected population and work plan for routine immunization, 2011. Department of Primary Health Care, Etsako Central LGA, Fugar.

8. Hailu M, Gebremariam A, Alemseged F, Deribe K. Birth preparedness and complication readiness among pregnant women in Southern Ethiopia. PLoS ONE 2011;6(6):e21432. http://dx.doi.org/10.1371/ journal.pone.0021432 9. Kabakyenga JK, Östergren P, Turyakira E, Pettersson KO. Knowledge of obstetric danger signs and birth preparedness practices among women in rural Uganda. Reprod Health 2011;8(1):33. http://dx.doi. org/10.1186%2F1742-4755-8-33 10. Urassa DP, Pembe AB, Mganga F. Birth preparedness and complication readiness among women in Mpwapwa District Tanzania. Tanzan J Health Res 2012;14(1):42-47. http://dx.doi.org/10.4314/thrb.v14i1.8 11. The World Bank. Successful Approaches to Improving Maternal Health Outcomes. http://go.worldbank. org/0K8ELL1EU0 (accessed 26 April 2015).

12. Obi AI, Abe E, Okojie OH. Male and community involvement in birth preparedness and complication readiness in Benin City, Southern Nigeria. IOSR J Dental Med Sci 2013;10(6):27-32.

13. Adeleye OA, Chiwuzie J. ‘He does his own and walks away’: Perception about male attitudes and practices regarding safe motherhood in Ekiadolor, Southern Nigeria. Afr J Reprod Health 2007;11(1):76-89.

14. Iliyasu Z, Abubakar IS, Galadanci HS, Aliyu MH. Birth preparedness, complication readiness and fathers’ participation in maternity care in a northern Nigerian community. Afr J Reprod Health 2010;14(1):21-32.

15. Towards 4+5 Research Programme Consortium. Improving birth outcomes in poor rural communities: The role of women’s groups in Nepal, Bangladesh and India. Towards 4+5 Briefing Paper 5, April 2010. http://r4d.dfid.uk/Output/186923/Default.aspx (accesssed 26 April 2015).

SAJOG • December 2016, Vol. 22, No. 2

This open-access article is distributed under Creative Commons licence CC-BY-NC 4.0.

RESEARCH

Sepsis: Primary indication for peripartum hysterectomies in a South African setting L Jansen van Vuuren, MB ChB; C A Cluver, MMed, FCOG Department of Obstetrics and Gynaecology, Stellenbosch University and Tygerberg Academic Hospital, Cape Town, South Africa Corresponding author: C A Cluver (cathycluver@hotmail.com)

Background. Peripartum hysterectomies are lifesaving procedures but definitions vary. Indications are variable and dependant on resources and geographical factors.

Objectives. To evaluate the incidence, aetiology and complications associated with peripartum hysterectomies in a tertiary hospital in South Africa.

Methods. A retrospective audit at an academic referral centre over a 5-year period from February 2009 to March 2014 was performed.

Procedures from a gestation of 24 weeks until 6 weeks postpartum were included.

Results. One hundred and sixty cases met inclusion criteria. Nine case records were unavailable. The incidence was 2.77 per 1 000 deliveries. Main indications were sepsis (60, 39.7%), atony (24, 15.9%), morbidly adherent placenta (21, 13.9%), tears (14, 9.3%), uterine rupture (8, 5.3%), placenta praevia (7, 4.6%) and unclassified bleeding (6, 4.0%). There were 6 maternal deaths. Five related to sepsis and one to hypovolaemic shock. One hundred and thirty-eight (91.4%) women required high or intensive care admission.

Conclusion. Sepsis is an important aetiology for peripartum hysterectomies, particularly in southern Africa. The high rate of sepsis may be due to HIV infection, low socioeconomic standards, late diagnosis, limited access to healthcare, sterility issues and differences in the definition and inclusion criteria used for a peripartum hysterectomy.

S Afr J Obstet Gynaecol 2016;22(2):52-56. DOI:10.7196/SAJOG.2016.v22i2.1068

A peripartum hysterectomy is a lifesaving procedure usually per­ formed as an emergency that can be associated with significant morbidity and mortality.[1] Despite many publications, the defi­ nition of a peripartum hysterectomy remains vague. Definitions include a hysterectomy performed within 24 hours of delivery,[2-5] a hysterectomy performed within the same hospitalisation,[6-10] a hysterectomy performed within 72 hours of delivery,[11] a hys­ terectomy performed within 1 month of delivery[12,13] to a hys­ terectomy performed within 6 weeks of delivery.[14] Some even limit the definition to a hysterectomy performed for uncontrolled haemorrhage only[4,15-22] and exclude cases of infection.[12] The vary­ ing definitions make it difficult to compare incidences and aetiology.

If a short time period after delivery is used, complications related to sepsis and delayed haemorrhage may be underestimated.

Peripartum hysterectomies complicate about 1 in 1 000 preg­ nancies.[23] The incidence is lower in higher-resource settings. In a large review from the USA the incidence was 0.77 per 1 000 deliveries[10] while African countries have reported higher incidences ranging from 4.34 to 9.5 per 1 000 deliveries.[9,14,17,19] Lack of antenatal and peripartum care due to limited resources, a high burden of HIV infection and a delay in recognising complications may be reasons for the higher incidences.

In high-resource settings complications of placentation are the most common indication for hysterectomy.[5-8,11,12,22,24,25] In middle-income countries the aetiology is variable, with studies from Turkey, India and Thailand showing similar indications to high-resource settings but with higher rates of uterine rupture.[4,15,16,21,26-28] Sepsis has been reported as an indication in low-resource settings. In Africa the most common aetiologies include uterine rupture, atonic uteri and sepsis.[14,19,29-31] Three studies performed in tertiary care settings in South Africa (SA) have shown particularly high rates of sepsis. A study performed in Mthatha showed that uterine atony, puerperal sepsis

and secondary postpartum haemorrhage made up 57% of the indications. The morbidly adherent placenta was the least common indication.[9] In a review of cases from Durban, uterine rupture and sepsis made up 56% of the indications,[29] and a study from Pretoria reported that puerperal sepsis was the second most common indication after ruptured uteri and accounted for 33% of peripartum hysterectomies.[31] Observational data in our department at Tygerberg Hospital, a tertiary referral centre in Cape Town, suggested that sepsis was a common indication. This study was therefore designed to sys­ tematically evaluate the incidence, aetiology and complications associated with peripartum hysterectomy in the unit.

Methods

A retrospective audit was performed at Tygerberg Hospital, a state academic referral centre over a 5-year period from February 2009 to March 2014. The study was approved by the Stellenbosch University Health Sciences Ethics Committee (S13/08/155).

Cases were identified in theatre record books and individual case records were reviewed. Data were extracted and transferred to an MS Excel spreadsheet. The inclusion criteria were all peripar­ tum hysterectomies performed from a gestation of 24 weeks until 6 weeks postpartum. Information on patients’ age, gravidity, parity, gestation, HIV status, CD4 count, antiretroviral (ARV) use, medical history, past obstetric history, antenatal care and complications, intrapartum course, mode of delivery, intrapartum complications, postpartum complications, neonatal outcome, indication for hysterectomy, type of hysterectomy, surgical complications, esti­ mated blood loss, high-care admissions, blood products required, morbidity, mortality and length of hospital admission was collected.

Statistica version 9 (StatSoft, USA) was used to analyse the data.

Descriptive statistics were used to describe the data. Frequencies

SAJOG • December 2016, Vol. 22, No. 2

52

RESEARCH (counts and percentages), measures of location (mean and median) and spread (standard deviations and percentiles) were used, depending on the distribution of the data.

Results

In the 5-year period 172 hysterectomies were identified. Two emergency theatre books were unobtainable. These contained information on peripartum hysterectomies that may have been performed from 25 October to 9 December 2011 and from 4 July to 15 August 2013. It is estimated that information on 0 to 12 cases may be missing. Twelve cases were excluded as the surgery was performed before 24 weeks. Three hysterectomies were performed for gestational trophoblastic disease, 1 was performed for an extrauterine pregnancy and 8 were performed for miscarriages. Six of the hysterectomies performed for a miscarriage were for septic retained products and 2 were for morbidly adherent placentas.

Of the 160 cases that met the inclusion criteria, 9 individual case records were unavailable. Two of these women died and 7 were discharged. One hundred and fifty-one cases were therefore included in the analysis. Ninety-six of these women delivered at Tygerberg Hospital and 55 delivered at a referral centre and were transferred because of complications. During this time period there were 34 558 births at Tygerberg Hospital, giving an overall rate for peripartum hysterectomies at our institution of 2.77 per 1 000 deliveries.

The youngest patient was 14 and the eldest 46 (mean age was 30 years). The majority were multigravid with a range of 1 - 7 pregnancies. The mean gestation at delivery was 36 weeks with a range from 24 to 42 weeks. One hundred and twenty-nine women (85.4%) had one or more antenatal visits, with 22 women (14.6%) presenting for the first time in labour. One hundred and forty women had singleton pregnancies, 10 patients had twin gestations and 1 patient had a triplet pregnancy. Of the parous women, 57 had had a previous caesarean section, with 28 (18.5%) having had one previous caesarean section and 29 (19.2%) having had two previous caesarean sections.

Fifty women (33.1%) were HIV-positive. Of these, 33 (66.0%) were on ARV treatment (ART) and 18 (36.0%) had a CD4 count <350 when they first presented for antenatal care. During the time period of this study women were initially only started on lifelong ART if their CD4 count was <350. In July 2013 our national guidelines regarding ART in pregnant women changed. Since then all HIVpositive pregnant women are started on lifelong ART at their first Table 1. Antenatal complications (N=151) n (%) Hypertensive-related disorders

49 (32.5)

Hypertension

17 (11.3)

Pre-eclampsia

19 (12.6)

Pre-eclampsia with HELLP syndrome

13 (8.6)

Diabetes mellitus

5 (3.3)

BMI >40

14 (9.3)

PROM

6 (4.0)

APH

15 (9.9)

Antenatal sepsis

0 (0)

HIV

50 (33.1)

HELLP = haemolysis, elevated liver enzyme levels, and low platelet levels; PROM = premature rupture of membranes; APH = antepartum haemorrhage.

53

Table 2. Indications for induction of labour (N=32) n (%) Hypertensive-related disorders

17 (53.1)

Hypertension

3 (9.4)

Pre-eclampsia

11 (34.4)

Eclampsia

3 (9.4)

Intra-uterine fetal demise

5 (15.6)

Postdate pregnancy

4 (12.5)

Prolonged rupture of membranes

3 (9.4)

Previous abruptio placentae

2 (6.3)

Previous intrauterine fetal demise

1 (3.1)

Table 3. Reasons for unsuccessful inductions (N=25) n (%) Fetal distress

8 (32)

Not progressing to active labour

7 (28)

Poor progress during labour

4 (16)

Cephalopelvic disproportion

3 (12)

Uterine rupture

2 (8)

Extrauterine pregnancy

1 (4)

antenatal visit. Other antenatal complications are documented in Table 1.

One hundred and two patients attempted a vaginal delivery.

Twenty-eight delivered vaginally (18.5%) with 1 requiring an assisted ventouse delivery. Seventy women (46.4%) went into spontaneous labour and 32 (21.2%) were induced. Indications for induction of labour are documented in Table 2. Only 7 patients had a successful induction that resulted in a normal vertex delivery. The reasons for unsuccessful induction are detailed in Table 3.

One hundred and twenty-three women had a caesarean section.

Twenty-seven (22.0%) were elective and 96 (78.0%) were emergency procedures. The indications for caesarean section are documented in Table 4.

The main indication for a peripartum hysterectomy was uterine sepsis (60 cases, 39.7%). Fifty-two (86.7%) of these women were delivered by caesarean section; the indications included fetal distress in 16 cases (30.8%), poor progress in 13 cases (25.0%) and failed induction in 7 cases (13.5%). Fifteen of the patients (25.0%) who developed uterine sepsis needing a hysterectomy were induced. Of these, 3 (20.0%) received misoprostol, 4 (26.7%) received dinoprostone (Prepidil), 3 (20.0%) received a balloon catheter, 9 (60.0%) had an artificial rupture of the membranes and 8 (53.3%) received oxytocin. Eight patients (53.3%) needed more than one induction method. Thirty-five patients (58.3%) went into spontaneous labour and 4 patients (6.7%) were not in labour.

None of these women was diabetic. Four were morbidly obese.

Twenty-one (35.0%) were HIV-positive and all of these women had either a CD4 count <350 (52.0%) or were not on ART (48.0%). The shortest time interval from delivery to hysterectomy for sepsis was 3 days. The majority were performed 7 - 14 days post-delivery, with the longest time interval being 41 days.

Uterine atony was the second most common aetiology. Of the 24 (15.9%) peripartum hysterectomies performed for uterine atony,

SAJOG â&#x20AC;˘ December 2016, Vol. 22, No. 2

RESEARCH Table 4. Indications for caesarean section (N=123) n (%) Emergency caesarean section Fetal distress

31 (25.2)

Poor progress

17 (13.8)

Two previous caesarean sections in labour

9 (7.3)

Failed induction of labour

7 (5.7)

Cephalopelvic disproportion

5 (4.1)

Elective caesarean section Morbidly adherent placenta

14 (11.4)

Placenta praevia

5 (4.1)

Two or more previous caesarean sections

4 (3.3)

Multiple pregnancy

2 (1.6)

Breech presentation

2 (1.6)

6 patients (25.0%) were multigravidas, 4 (16.7%) had a multiple pregnancy, 2 (8.3%) had macrosomic fetuses (estimated fetal weight more than 4 kg) and 8 (33.3%) had had a previous caesarean section.

Eight (33.3%) had a vertex vaginal delivery and 16 (66.7%) had a caesarean section, with 8 (33.3%) being elective because of pre­ vious caesarean section(s), multiple pregnancy, breech presentation and suspected fetal macrosomia. Eight were emergency caesarean sections. Eleven women (45.8%) went into spontaneous labour and 5 (20.8%) had an induction of labour. Oxytocin infusion with artificial rupture of the membranes was the induction method most commonly used. Four of the 5 patients who had inductions were administered more than one induction method.

Morbidly adherent placenta was the third most common indi­ cation. Of the 21 cases 7 (33.3%) were emergency hysterectomies as spontaneous labour occurred or a significant antepartum hae­ morrhage occurred before the planned delivery date. Fourteen cases were elective caesarean hysterectomies. Five patients had had one previous caesarean section, 12 had had two previous caesarean sections and one patient had had a previous uterine evacuation of retained products. Three patients had had previous normal vertex deliveries.

There were 14 cases (9.3%) of peripartum hysterectomies related to uncontrollable bleeding from tears. Thirteen were tears at emer­ gency caesarean section that extended into the right or left uterine artery. In one of these cases the tear extended into the cervix and in another the tear extended into the cervix and vagina. There was one case of a vaginal tear during a vaginal delivery that extended to the cervix that required a hysterectomy to control the bleeding.

Uterine rupture occurred in 8 cases (5.3%). Half of these women had previously delivered by caesarean section. One woman needed a hysterectomy after blunt abdominal trauma. She had a scarred uterus from a previous caesarean section and was 30 weeks pregnant. Three women went into spontaneous labour and 4 had an induction of labour, of whom 3 received misoprostol and 1 received Prepidil and a balloon catheter.

Seven women (4.6%) who had a caesarean section for placenta praevia required a peripartum hysterectomy for placental bed bleeding that was not controllable with conservative methods.

Six women (4.0%) had a peripartum hysterectomy due to unspecified haemorrhage. In these cases the cause of haemorrhage was not specified in the medical records.

One hundred and one total abdominal hysterectomies were per­ formed and 50 cases were subtotal hysterectomies. In 3 of the subtotal hysterectomies the cervical stump was removed at a relook laparotomy. The main indication for performing a subtotal hysterectomy was haemodynamic instability and surgical difficulty because of distorted anatomy and/or adhesions.

There were 6 cases of maternal deaths. We were unable to obtain the case records for 2 of these cases but the cause of death on the death certificate was reported to be septic shock. Sepsis was therefore the cause of death in 5 cases and hypovolaemic shock in 1 case. Of the 3 cases whose deaths were related to sepsis, where the clinical records were available, only 1 was HIV-positive. She was a 26-year-old with a CD4 count of 17, not on ART, who was known to have chronic hepatitis B infection with liver failure and a history of perforated peptic ulcer disease. She had a preterm vagi­ nal delivery, complicated by disseminated intravascular coagulation (DIC) and multiple postpartum haemorrhages requiring repeated laparotomies and transfusions. She developed severe puerperal sepsis and had a septic uterus at the time of hysterectomy. She died after she had received 70 units of packed red blood cells, 71 units of fresh frozen plasma, 7 units of platelets and 7 units of cryoprecipitate during her hospital admission. The second death was a 41-year-old multigravida who did not receive any antenatal care.

She had a vaginal delivery at term and presented in septic shock 3 days postpartum. She had a cardiac arrest and was successfully resuscitated but sadly died later that day. Postmortem examination confirmed active pelvic floor sepsis and tonsillar herniation due to brain oedema. The third woman was a 33-year-old who was induced at 38 weeks for hypertension. She had prolonged rupture of membranes and a caesarean section was performed for fetal distress. One week postpartum she presented in septic shock. At the referring hospital she suffered a cardiac arrest and was successfully resuscitated. A laparotomy was performed and a ruptured sigmoid colon was repaired. A repeat laparotomy was required as she was unresponsive to therapy and a total abdominal hysterectomy was performed. She later died as a result of neutropenic septic shock that was unresponsive to treatment. The maternal death related to haemorrhage and hypovolaemic shock involved a 19-yearold primigravida who was induced for a post-date pregnancy and received misoprostol followed by a balloon catheter. The membranes were then artificially ruptured and oxytocin was given.

A caesarean section was performed for poor progress. At caesarean section an atonic uterus was diagnosed. She received ergometrine, prostaglandin F2α and oxytocin. Uterine compression sutures and uterine artery ligation were performed without success. The abdomen and pelvis were then packed with swabs and she was transferred to Tygerberg Hospital. She suffered a cardiac arrest during the surgery and was successfully resuscitated. Unfortunately DIC developed and after two further laparotomies she died.

Peripartum hysterectomies were associated with significant morbi­ dity. A total of 138 women (91.4%) required either intensive care admission or a high-care admission. The majority of women who did not require intensive or high-care admission were elective caesarean hysterectomy cases.

Forty-three women (28.5%) required one or more repeat laparo­ tomies with one requiring nine relook procedures. Most women (137, 90.7%) required blood transfusions. The average (median) amount of packed red blood cells was 6 units with a range of 1 to 70 units. Ninety-six women (63.6%) were given fresh frozen

SAJOG • December 2016, Vol. 22, No. 2

54

RESEARCH plasma. The average amount of fresh frozen plasma required was 4 units, with a maximum of 71 units. Fifty-two women (34.4%) required platelets. On average one pool of platelets was required, with a maximum of 7 units being used. Twenty-six patients (17.2%) required cryo­precipitate.

Sixty-nine women (45.7%) required continuous positive airway pressure support and 53 (35.1%) were ventilated, with 6 (4.0%) needing a tracheostomy as a result of prolonged ventilation.

Twenty-two women (14.6%) required inotropic support, 21 (13.9%) developed renal impairment, 3 (2.0%) developed a deep venous thromboembolism and 3 (2.0%) developed a cardiomyopathy, 2 secondary to sepsis and 1 being diagnosed as a peripartum cardiomyopathy. The mean admission time from peripartum hysterectomy to discharge or death was 10.2 days.

Discussion

Sepsis is an important aetiology for a peripartum hysterectomy, particularly in southern Africa. In this study, sepsis was the most common indication for a peripartum hysterectomy. This is the first study, according to our knowledge, to report this finding. The high rate of sepsis may be due to the high incidence of HIV infection in our community, low socio-economic standards, limited access to healthcare, late diagnosis of complications, sterility issues in the labour ward and theatre, and differences in the definition and inclusion criteria used for a peripartum hysterectomy. If the definition for peripartum hysterectomy had been hysterectomy at delivery or within 24 hours of delivery, or a hysterectomy for uncontrolled haemorrhage, we would have had no cases of sepsis and the incidence in this study would have been significantly lower.

Defining a peripartum hysterectomy as a hysterectomy performed within 6 weeks of delivery, which is in line with the World Health Organization (WHO)’s definition of maternal mortality,[32] would result in more standard reporting enabling one to compare studies with greater accuracy. Some may argue that even hysterectomies due to miscarriages and gestational trophoblast disease should be included in the definition as these are also related to pregnancy.

Improving access to ART is important as all the women who were HIV-positive who needed a hysterectomy for the indication of sepsis had a CD4 count <350 or were not using ART. Implementing the WHO treatment programme Option B+, which entails starting a single daily dose fixed-combination regimen for all HIV-positive women irrespective of the CD4 count and the stage of disease,[33] may decrease the rate of sepsis.

Uterine atony, the morbidly adherent placenta, uterine and cer­ vical lacerations and uterine rupture were also identified as important aetiologies for peripartum hysterectomies. The morbidly adherent placental spectrum of disease and uterine rupture is known to occur more commonly with previous caesarean sections,[34] so avoiding unnecessary caesarean sections is important in decreasing the inci­ dence of peripartum hysterectomies. Delays in self-referral, doctor referral and identification of complications have been described as contributing factors to increased morbidity and mortality.[35] This is especially important in the case of sepsis and haemorrhage, where early identification and treatment may prevent a hysterectomy.

This study did not demonstrate that diabetes is associated with an increased risk for a peripartum hysterectomy, despite our high rate of sepsis and diabetes which is different to published literature.[36] Induction of labour, particularly with prostaglandins, was less of a risk factor than anticipated.

55

This study highlights the high morbidity and mortality associated with a peripartum hysterectomy. Morbidity and mortality after a peripartum hysterectomy are high and the majority of deaths were associated with sepsis. This is particularly challenging as the majority of septic hysterectomies occur in countries where access to high care and intensive care facilities is limited.

Strengths of this study include the size of the sample and the definition of a peripartum hysterectomy used, which includes all hysterectomies performed until 42 days postpartum and does not limit the inclusion to only hysterectomies for uncontrolled bleeding.

Disadvantages of the study include that it was a retrospective review.

Studies assessing the aetiology of peripartum hysterectomies should use a definition that extends to 42 days after the delivery to avoid missing cases of sepsis and delayed haemorrhage, and hysterectomies performed for all obstetric indications should be included.

Further research should be aimed at assessing why the incidence of sepsis is so high in certain areas, particularly South Africa.

Acknowledgement. This article was submitted to the South African Society of Obstetrics and Gynaecology 2016 Congress (1 - 4 May 2016) as an oral presentation.

References 1. Wright JD, Devine P, Shah M, et al. Morbidity and mortality of peripartum hysterectomy. Obstet Gynecol 2010;115(6):1187-1193. http://dx.doi.org/10.1097/aog.0b013e3181df94fb 2. Castaneda S, Karrison T, Cibils LA. Peripartum hysterectomy. J Perinat Med 2000;28(6):472-481. http://dx.doi.org/10.1515/jpm.2000.064 3. Chestnut DH, Eden RD, Gall SA, Parker RT. Peripartum hysterectomy: A review of cesarean and postpartum hysterectomy. Obstet Gynecol 1985;65(3):365-370.

4. Demirci O, Tuğrul AS, Yılmaz E, Tosun Ö, Demirci E, Eren YS. Emergency peripartum hysterectomy in a tertiary obstetric center: Nine years evaluation. J Obstet Gynaecol Res 2011;37(8):1054-1060. http://dx.doi.org/10.1111/j.1447-0756.2010.01484.x 5. Glaze S, Ekwalanga P, Roberts G, et al. Peripartum hysterectomy: 1999 to 2006. Obstet Gynecol 2008;111(3):732-738. http://dx.doi.org/10.1097/aog.0b013e31816569f2 6. Bateman BT, Mhyre JM, Callaghan WM, Kuklina EV. Peripartum hysterectomy in the United States: Nationwide 14 year experience. Obstet Gynecol 2012;206(1):63:61-63,e8. http://dx.doi. org/10.1016/j.ajog.2011.07.030 7. Knight M. Peripartum hysterectomy in the UK: management and outcomes of the associated haemorrhage. BJOG 2007;114(11):1380-1387. http://dx.doi.org/10.1111/j.1471-0528.2007.01507.x 8. Stivanello E, Knight M, Dallolio L, Frammartino B, Rizzo N, Fantini MP. Peripartum hysterectomy and cesarean delivery: A population-based study. Acta Obstet Gynecol Scand 2010;89(3):321-327. http://dx.doi.org/10.3109/00016340903508627 9. Wandabwa J, Businge C, Longo-Mbenza B, Mdaka M, Kiondo P. Peripartum hysterectomy: Two years experience at Nelson Mandela Academic Hospital, Mthatha, Eastern Cape, South Africa.

African Health Sciences 2013;13(2):469-474. http://dx.doi.org/10.4314/ahs.v13i2.38 10. Whiteman MK, Kuklina E, Hillis SD, et al. Incidence and determinants of peripartum hysterectomy.

Obstet Gynecol 2006;108(6):1486-1492. http://dx.doi.org/10.1097/01.aog.0000245445.36116.c6 11. Tadesse W, Farah N, Hogan J, D’arcy T, Kennelly M, Turner M. Peripartum hysterectomy in the first decade of the 21st century. J Obstet Gynaecol 2011;31(4):320-321. http://dx.doi.org/10.3109/ 01443615.2011.560300 12. Sakse A, Weber T, Nickelsen C, Secher NJ. Peripartum hysterectomy in Denmark 1995-2004. Acta Obstet Gynecol Scand 2007;86(12):1472-1475. http://dx.doi.org/10.1080/00016340701692651 13. Bodelon C, Bernabe-Ortiz A, Schiff MA, Reed SD. Factors associated with peripartum hysterectomy.

Obstet Gynecol 2009;114(1):115-123. http://dx.doi.org/10.1097/aog.0b013e3181a81cdd 14. Kwame-Aryee R, Kwakye A, Seff J. Peripartum hysterectomies at the Korle-Bu Teaching Hospital: A review of 182 consecutive cases. Ghana Med J 2010;41(3):133-138. http://dx.doi.org/10.4314/ gmj.v41i3.55281 15. Yucel O, Ozdemir I, Yucel N, Somunkiran A. Emergency peripartum hysterectomy: A 9-year review.

Arch Gynecol Obstet 2006;274(2):84-87. http://dx.doi.org/10.1007/s00404-006-0124-4 16. Wingprawat S, Chittacharoen A, Suthutvoravut S. Risk factors for emergency peripartum cesarean hysterectomy. Int J Gynaecol Obstet 2005;90(2):136-137. http://dx.doi.org/10.1016/j. ijgo.2005.04.013 17. Umezurike CC, Feyi‐Waboso PA, Adisa CA. Peripartum hysterectomy in Aba southeastern Nigeria. Aust N Z J Obstet Gynaecol 2008;48(6):580-582. http://dx.doi.org/10.1111/j.1479828x.2008.00905.x 18. Smith J, Mousa H. Peripartum hysterectomy for primary postpartum haemorrhage: Incidence and maternal morbidity. J Obstet Gynaecol 2007;27(1):44-47. http://dx.doi. org/10.1080/01443610601016925 19. Obiechina N, Eleje G, Ezebialu I, Okeke C, Mbamara S. Emergency peripartum hysterectomy in Nnewi, Nigeria: A 10-year review. Nigerian J Clin Pract 2012;15(2):168-171. http://dx.doi. org/10.4103/1119-3077.97303 20. Kastner ES, Figueroa R, Garry D, Maulik D. Emergency peripartum hysterectomy: Experience at a community teaching hospital. Obstet Gynecol 2002;99(6):971-975. http://dx.doi. org/10.1097/00006250-200206000-00003 21. Karayalçın R, Özcan S, Özyer Ş, Mollamahmutoğlu L, Danışman N. Emergency peripartum hysterectomy.

Arch Gynecol Obstet 2011;283(4):723-727. http://dx.doi.org/10.1007/s00404-010-1451-z

SAJOG • December 2016, Vol. 22, No. 2

RESEARCH 22. Awan N, Bennett MJ, Walters WA. Emergency peripartum hysterectomy: A 10‐year review at the Royal Hospital for Women, Sydney. Aust N Z J Obstet Gynaecol 2011;51(3):210-215. http://dx.doi. org/10.1111/j.1479-828x.2010.01278.x 23. Turner MJ. Peripartum hysterectomy: An evolving picture. Int J Gynecol Obstet 2010;109(1):9-11. http://dx.doi.org/10.1016/j.ijgo.2009.12.010 24. Flood KM, Said S, Geary M, Robson M, Fitzpatrick C, Malone FD. Changing trends in peripartum hysterectomy over the last 4 decades. Obstet Gynecol 2009;200(6):632. e1-632. http://dx.doi. org/10.1016/j.ajog.2009.02.001e6.

25. Kwee A, Bots ML, Visser GH, Bruinse HW. Emergency peripartum hysterectomy: A prospective study in The Netherlands. Eur J Obstet Gynecol Reprod Biol 2006;124(2):187-192. http://dx.doi. org/10.1016/j.ejogrb.2005.06.012 26. Kayabasoglu F, Guzin K, Aydogdu S, Sezginsoy S, Turkgeldi L, Gunduz G. Emergency peripartum hysterectomy in a tertiary Istanbul hospital. Arch Gynecol Obstet 2008;278(3):251-256. http:// dx.doi.org/10.1007/s00404-007-0551-x 27. Yalinkaya A, Güzel AI, Kangal K. Emergency peripartum hysterectomy: 16-year experience of a medical hospital. J Chinese Med Assoc 2010;73(7):360-363. http://dx.doi.org/10.1016/s17264901(10)70078-2 28. Saxena S, Bagga R, Jain V, Gopalan S. Emergency peripartum hysterectomy. Int J Gynecol Obstet 2004;85(2):172-173. http://dx.doi.org/10.1016/j.ijgo.2003.09.011

29. Sebitloane M, Moodley J. Emergency peripartum hysterectomy. East Afr Med J 2001;78(2):70-74. http://dx.doi.org/10.4314/eamj.v78i2.9091 30. Seffah J, Kwame-Aryee R. Emergency peripartum hysterectomy in the nulliparous patient. Int J Gynecol Obstet 2007;97(1):45-46. http://dx.doi.org/10.1016/j.ijgo.2006.11.018 31. Shava J, Masihleho GE, Mazibuko MD. Peripartum hysterectomy at Ga-Rankuwa Hospital: A two and a half year review. Cent Afr J Med 1996;42(1):25-28.

32. Khan KS, Wojdyla D, Say L, Gülmezoglu AM, van Look PF. WHO analysis of causes of maternal death: A systematic review. Lancet 2006;367(9516):1066-1074. http://dx.doi.org/10.1016/s01406736(06)68397-9 33. Ahmed S, Kim MH, Abrams EJ. Risks and benefits of lifelong antiretroviral treatment for pregnant and breastfeeding women: A review of the evidence for the Option B+ approach. Curr Opin HIV AIDS 2013;8(5):474-489. http://dx.doi.org/10.1097/coh.0b013e328363a8f2 34. Knight M, Kurinczuk JJ, Spark P, Brocklehurst P, United Kingdom Obstetric Surveillance System Steering Committee. Cesarean delivery and peripartum hysterectomy. Obstet Gynecol 2008;111(1):97-105. http://dx.doi.org/10.1097/01.aog.0000296658.83240.6d 35. Akar ME, Yilmaz ES, Yuksel B, Yilmaz Z. Emergency peripartum hysterectomy. Eur J Obstet Gynecol Reprod Biol 2004;113(2):178-181 http://dx.doi.org/10.1016/j.ejogrb.2003.10.005 36. Jou H, Hung H, Ling P, Chen S, Wu S. Peripartum hysterectomy in Taiwan. Int J Gynecol Obstet 2008;101(3):269-272. http://dx.doi.org/10.1016/j.ijgo.2007.12.004

SAJOG • December 2016, Vol. 22, No. 2

56

This open-access article is distributed under Creative Commons licence CC-BY-NC 4.0.

RESEARCH

Factors predictive of abnormal semen parameters in male partners of couples attending the infertility clinic of a tertiary hospital in south-western Nigeria O P Aduloju,1 FWACS; P T Adegun,2 FMCS 1 2

Department of Obstetrics and Gynaecology, Faculty of Clinical Sciences, College of Medicine, Ekiti State University, Ado-Ekiti, Nigeria Division of Urology, Department of Surgery, Faculty of Clinical Sciences, College of Medicine, Ekiti State University, Ado-Ekiti, Nigeria

Corresponding author: O P Aduloju (peter.aduloju@yahoo.com)

Background. Infertility is a common gynaecological problem and male factors contribute significantly to its aetiology. Semen analysis has remained useful for investigation of male factor infertility.

Objective. To assess the pattern of semen parameters, and predictive factors associated with abnormal parameters, in male partners of infertile couples attending a Nigerian tertiary hospital.

Methods. A descriptive study of infertile couples presenting at the clinic between January 2012 and December 2015 at Ekiti State University Teaching Hospital, Ado-Ekiti, Nigeria was done. Seminal fluid from the male partners was analysed in the laboratory using the World Health Organization 2010 criteria for human semen characteristics. Data were analysed using SPSS 17 and logistic regression analysis was used to determine the predictive factors associated with abnormal semen parameters.

Results. A total of 443 men participated in the study and 38.2% had abnormal sperm parameters. Oligozoospermia (34.8%) and asthenozoospermia (26.9%) were the leading single-factor abnormalities found, and astheno-oligozoospermia occurred in 14.2% and oligoasthenoteratozoospermia in 3.6% of cases. The prevalence of azoospermia was 3.4%. Smoking habit, past infection with mumps and previous groin surgery significantly predicted abnormal semen parameters (p=0.025, 0.040 and 0.017, respectively). Positive cultures were recorded in 36.2% of cases and Staphylococcus aureus was the most common.

Conclusion. Male factor abnormalities remain significant contributors to infertility and men should be encouraged, through advocacy, to participate in investigations into infertility, to reduce stigmatisation and ostracising of women with infertility, especially in sub-Saharan Africa.

S Afr J Obstet Gynaecol 2016;22(2):57-61. DOI:10.7196/SAJOG.2016.v22i2.1082

Infertility is a common gynaecological problem and is one of the most common reasons for consultations in gynaecological clinics.

Over 80% of laparoscopic investigations are performed for infertility management.[1-3] and infertility remains a sensitive issue in our environment and a source of social stigma.[1,4] The burden of this stigma is felt more by female partners, who are often perceived as responsible for infertility and are faced with the challenges of economic deprivation, social neglect, marital instability, emotional stress and unhappiness.[1,5] Infertility is a global problem with a variation in prevalent rate between regions. Worldwide, infertility is generally quoted as occur­ ring in 8 - 15% of all couples,[3,6] while in sub-Saharan Africa a prevalent rate of 15 - 45% has been variously reported.[1,3,7] However in Nigeria, reports from earlier studies have given an incidence of 20 - 30%.[7,8] Infertility is an underlying pathology, with female factors contributing 30 - 40% of causes, male factors about 30 - 40%, and both factors and unexplained infertility accounting for 20 - 40% of causes.[3,6,7,9,10] The aetiology of male infertility is largely unknown in most cases.[7,11] However, studies have shown upward trends in the prevalence of sexually transmitted and urogenital infections. Semi­

57

nal tract infections play a major contributory role in male infertility, affecting fertility through a number of different mechanisms, including impairment to spermatogenesis and sperm function, and obstruction of the seminal tract.[6,10-12] Other factors that may lead to male infertility include varicocele, endocrine disturbance, immunological conditions, sexual dysfunction and ejaculatory failure.[7,11] Semen analysis has remained a useful investigation in the search for male factor infertility and provides insight into the pro­ cess of sperm production count, and sperm quality-motility and morphology.[3,6,8,9] The semen parameters have been found an important determinant of functional competence of the spermatozoa.[12,13] Therefore, careful evaluation of these parameters may point to the possible causes of abnormal semen parameters and male infertility. This would enable the institution of appropriate treatment targeted at the identified aetiological factors.

Most previous studies on semen pattern have been based on the World Health Organization (WHO) 1999 criteria for human semen characteristics;[1,4-7,9] however, there is a paucity of recent studies using the WHO 2010 criteria.[3,14] This study was conducted to assess the pattern of semen parameters in male partners of infertile couples attending the

SAJOG • December 2016, Vol. 22, No. 2

RESEARCH gynaecological clinic of Ekiti State University Teaching Hospital, Ado-Ekiti, Nigeria using the recent WHO 2010 criteria for semen characteristics to identify the contribution of male factors to the burden of infertility in our environment.

Table 1. Sociodemographic characteristics of male participants involved in the study Characteristic

n (%), N=443

Age group (years)

Materials and methods

≤30

55 (12.4)

31 - 35

216 (48.8)

The study was a descriptive evaluation of seminal fluid of male partners of infertile women presenting at the gynaecological clinic of Ekiti State University Teaching Hospital, Ado-Ekiti between January 2012 and December 2015. The male partners of infertile women who presented at the clinic were invited to the clinic through the women and a total of 443 consecutively consenting male partners of women with infertility were recruited.

36 - 40

90 (20.3)

41 - 45

55 (12.4)

46 - 50

24 (5.4)

≥50

3 (0.7)

Subjects

Family setting Monogamous

362 (81.7)

Polygamous

81 (18.3)

Educational level

Sample collection A semistructured questionnaire with two sections was used to record information from the participants elicited by house officers and nursing staff of the gynaecological clinic. The first section reported the sociodemographic characteristics of the participants in terms of age, educational status, religion, occupation, marital status, family setting, type of infertility, duration of infertility, his­ tory of smoking and alcohol intake, history of childhood mumps infection, past history of chronic medical conditions such as diabetes mellitus, and past history of groin surgery such as herniorrhaphy or hydrocelectomy. In the second section the results of semen analysis in terms of volume, concentration, count, motility, morphology, period of continence and method of collection were recorded. The male partners were adequately counselled and given instructions on how to collect the semen sample. Instructions included abstinence from coitus for 3 - 5 days, washing of their hands before star­ting masturbation, and sample collection by masturbation only, kept close to the body and delivered to the hospital laboratory within 15 20 minutes of semen collection if not collected in the laboratory.

Spilled samples were avoided. Samples were collected into sterile screw-capped plastic universal containers.

The semen samples were collected in a dedicated room with bed and other facilities to promote relaxation within the laboratory, while participants living close to the hospital were allowed to collect at home, bringing the samples to the hospital within 15 - 20 minutes of collection..

Laboratory methods The semen analysis was performed according to the methods and standards outlined by the WHO 2010.[15] The parameters assessed included volume 1.5 mL or more; sperm concentration >15 × 106 cells/mL; motility >40% progressive/forward movement; morphology >4% normal form; and white blood cell count 1 ×106 cells/mL. The sample analysis was done by the same laboratory scientist to avoid inter-laboratory variation, within 1 hour of collection.

The sample was assessed for volume, appearance, liquefaction, concentration, motility, morphology, viability and presence of pus cells. The semen volume was measured using a graduated disposable pipette and pH checked with pH paper. After liquefaction, the semen specimen was thoroughly mixed using a pipette and a thin drop of specimen was spread on a glass slide by placing a cover slip on it. Sperm motility was assessed using an Olympus binocular microscope, magnification ×100, while the sperm concentration was counted in millions per mL using the Meckler counting chamber

Primary

48 (10.8)

Secondary

105 (23.7)

Tertiary

290 (65.5)

Occupation Clergy

37 (8.4)

Teaching

68 (15.3)

Trading

75 (16.9)

Artisan

84 (19.0)

Civil servant

179 (40.4)

Social habits Smoking Yes

41 (9.3)

No

402 (90.7)

Alcohol Yes

18 (4.1)

No

425 (95.9)

Past infection with mumps Yes

19 (4.3)

No

424 (95.7)

Chronic medical condition Yes

34 (7.2)

No

409 (92.3)

Previous groin surgery Yes

45 (10.2)

No

398 (89.8)

Type of infertility Primary

72 (16.3)

Secondary

371 (83.7)

Volume of semen (mL) <2.0

154 (34.8)

≥2.0

289 (65.2) Range

Mean (SD)

Age of male partner (years)

30 - 60

36.36 (5.07)

Duration of infertility (years)

1 - 11

3.13 (2.40)

Volume of semen (mL)

0.5 - 5.0

2.36 (1.22)

Period of abstinence (days)

3-7

4.54 (0.99)

Sperm concentration (10 /mL)

0 - 170

35.41 (31.60)

Total sperm count (106)

0 - 510

90.36 (97.44)

6

SAJOG • December 2016, Vol. 22, No. 2

58

RESEARCH and categorised in accordance with WHO normal and pathological ranges. Bacteriological tests were also carried out on the semen samples by culture on appropriate culture media at 37°C for 24 48 hours to detect bacterial pathogens, and positive samples were subcultured to determine the appropriate antibiotic sensitivity pattern.

Operational definitions The operational definitions used were as follows: • • • • • •

Normospermia: sperm count of ≥15 million/mL Oligospermia: sperm count of ≤15 million/mL Azoospermia: absence of spermatozoa in the ejaculate Asthenospermia: reduced sperm motility <40% Teratozoospermia: reduced sperm morphology <4% Oligoasthenoteratozoospermia (OAT): all variables abnormal

Data analysis Data were analysed using SPSS software version 17 (SPSS Inc, USA) for frequency, mean and χ2 with the level of significance set at p<0.05.

Logistic regression analysis was performed to determine the risk factors significantly associated with abnormal sperm concentration.

Ethical considerations Ethical approval was obtained from the Ethics and Research Com­ mittee of Ekiti State University Teaching Hospital and verbal consent was obtained from each couple participating in the study, following explanation of the study objectives. Questionnaires were made

anonymous and couples were at liberty to withdraw or refrain from the study without any consequence.

Results

A total of 443 men participated in the study. The analysis revealed that 274 (61.8%) had normal and 169 (38.2%) abnormal semen parameters.

The age range of participants was between 30 - 60 years, with a mean (standard deviation (SD)) of 36.36 (5.07) years and the majority (69.1%) aged between 31 - 45 years. The duration of infertility was between 1 - 11 years, with a mean of 3.13 (2.40) years.

A total of 72 (16.3%) participants were investigated for a case of primary infertility while 371 (83.7%) were investigated for secondary infertility. The period of abstinence ranged between 3 - 7 days, with a mean of 4.54 (0.99) days. The sociodemographic characteristics of the male partners are shown in Table 1.

Various risk factors were associated with abnormal semen para­­ meters. These included occupation and smoking habit of participants, past infections with mumps and previous groin surgery (p=0.001, 0.04, 0.022 and 0.004, respectively). Alcohol habits and chronic medical conditions were not significantly associated (p=0.121 and 0.469, respectively) (Table 2).

Multivariate logistic regression showed that smoking habit, past infection with mumps (mumps orchitis) and previous groin surgery in the male participants were significantly associated with abnormal semen parameters (p=0.025, 0.040 and 0.017, respectively) when controlled for multiple risk factors.

Table 2. Risk factors associated with abnormal sperm parameters Sperm concentration, n (%) Low

Normal

Total

p-value

Clergy

3 (8.1)

34 (91.9)

37 (8.4)

0.001*

Teaching

26 (38.2)

42 (61.8)

68 (15.3)

Trading

15 (20.0)

60 (80.0)

5 (16.9)

Artisan

57 (67.9)

27 (37.1)

84 (19.0)

Civil servant

68 (38.0)

111 (62.0)

179 (40.4)

Yes

26 (57.8)

19 (42.2)

45 (10.2)

No

143 (35.9)

255 (64.1)

398 (89.8)

Yes

10 (55.6)

8 (44.4)

18 (4.1)

No

159 (37.4)

266 (62.6)

425 (95.9)

Yes

12 (63.2)

7 (36.8)

19 (4.3)

No

157 (37.0)

267 (63.0)

424 (95.7)

Yes

11 (32.4)

23 (67.6)

34 (7.7)

No

158 (38.6)

251 (61.4)

409 (92.3)

Yes

26 (57.8)

19 (42.2)

45 (10.2)

No

143 (35.9)

255 (64.1)

398 (89.8)

Variable Occupation

Smoking 0.04*

Alcohol 0.121

Past infection with mumps 0.022*

Chronic medical illnesses 0.469

Previous groin surgery

*Indicates statistical significance.

59

SAJOG • December 2016, Vol. 22, No. 2

0.004*

RESEARCH A total of 154 (34.8%) participants produced seminal volume of <2 mL while 289 (66.9%) produced >2 mL (Table 1). The various abnormalities occurring singly or in combinations are shown in Table 4.

Although seminal fluid abnormality was found throughout the age groups, it was higher in those aged 31 - 45 years (Table 5).

A total of 160 (36.2%) samples cultured positive for organisms and Staphylococcus aureus was the most common organism isolated, accounting for 24.4% of organisms cultured (Table 6).

Discussion

The study found that 38.2% of investigated couples had abnormal semen parameters. This result is higher than findings at Ile-Ife and Ibadan in south-western Nigeria.[1,3] but similar to that reported from Abakaliki in south-eastern Nigeria.[12] Semen analysis revealed the various sperm abnormalities contributory to male factor infertility.

Sperm abnormalities due to distortion in the spermatogenesis process may be pretesticular (hormonal), testicular (chromosomal) and post-testicular (disorder in transportation or ejaculation, or caused by infections etc).[3,6,8,12] These abnormal parameters, occuring singly or in combination, impair fertility even with normal sperm concentration. The outcome of treatment of male factor infertility is dependent on the presence of these factors. It has been reported that prognosis is inversely proportional to the number of abnormal patterns, i.e. that having one factor abnormality is better than having two factors, and two factors better than three factors.[1,6,16] The single-factor abnormalities of low sperm count oligozoo­spermia (34.8%) and poor motility-asthenozoospermia (26.9%) were leading factors in sperm parameter abnormality while teratozoo­ spermia contributed 6.9%. This corroborates findings reported by previous studies in this environment.[1,8,12] However, the two-factor abnormality of astheno-oligozoospermia was recorded in 14.2% of cases, which is comparable with findings from Jos and Ile-Ife[3,16] but lower than reported in Ibadan,[1] and the three-factor abnormality OAT occurred in 3.6% of the participants, which is comparable with that reported from Ile-Ife[3] but lower than figures reported from Jos and Ibadan.[1,17] The presence of these factors is associated with poor outcome with the use of conventional methods in the treatment of infertile couples.

However, with newer techniques and advancements in assisted reproduction and conception, which are gradually becoming more available in our environment, pregnancy can be achieved.[1,3,8,17] The prevalence of azoospermia of 3.4% in this study compares well with the rate in the general male population but was lower than findings in previous studies.[3,5,6] About 62% of male partners in this study had normal sperm concentration, while the mean sperm density was 35.41 (31.60) × 106 cells/mL; this implies that not only is absolute sperm count a male factor infertility determinant, but other components such as motility and morphology are equally important. Hence, infertility is not only associated with low sperm count but rather defective sperm parameters or other factors such as female factors.

The majority of the male partners produced normal semen volume although 33.1% had a low semen volume. Mean semen volume was 2.36 (1.22) mL. These values are comparable with earlier reports from studies by Butt et al., Nwafia et al. and Imam et al.[6,18,19] The adequate volumes reported in this study may be related to the period of continence observed by the male partners before presenting for seminal analysis, which was between 3 - 7 days with a mean of 4.54 (0.99) days, and this reflects the importance of abstinence before seminal fluid collection for analysis. Studies have shown that prolonged abstinence is

Table 3. Multivariate logistic regression analysis with sperm concentration as dependent variable AOR (95% CI for AOR)

Variable

p-value

Smoking Yes

1

No

0.479 (0.252 - 0.911)

0.025*

Previous groin surgery No

1

Yes

0.460 (0.243 - 0.871)

0.017*

Past infection with mumps No

1

Yes

0.396 (0.150 - 1.046)

0.040*

Occupation of male partner Professional

1

Artisan

1.486 (0.989 - 2.231)

0.056

AOR = Adjusted odds ratio; CI = confidence interval. *Indicates statistical significance.

Table 4. Pattern of abnormal semen parameters results of couples involved in the study (N=16) Seminal fluid analysis

n (%)

Oligozoospermia

59 (34.9)

Asthenozoospermia

45 (26.6)

Oligoasthenozoospermia

24 (14.2)

Teratozoospermia

11 (6.5)

Asthenoteratozoospermia

10 (5.9)

Oligoteratozoospermia

7 (4.2)

OAT

7 (4.2)

Azoospermia

6 (3.5)

associated with increased sperm concentration, but does not necessarily improve morphology and motility.[3,6,20] There was a positive culture in 36.2% of our cases, with S. aureus accounting for the greatest proportion, comparable with previous findings.[3,7,21,22] This may reflect penile contamination of the semen during collection even though the participants were instructed to observe aseptic technique. It may also be due to male genital infection, an important aetiological factor in male infertility which may lead to distortion of the process of spermatogenesis, impairment of sperm function and obstruction of the seminal tract. This may be contributory to the abnormal semen parameters recorded in this study, among the other factors elucidated.[1,3,6,12,20] Environmental factors such as exposure to heat and chemicals, lifestyle factors such as smoking and alcohol consumption, chronic medical conditions such as diabetes mellitus and thyroid disease, previous groin surgery, including herniorrhaphy or varicocelec­ tomy, and past mumps infection have been demonstrated as having adverse effects on sperm parameters.[12,23] Consistent with previous reports,[24-26] this study demonstrated a significant association

SAJOG • December 2016, Vol. 22, No. 2

60

RESEARCH Table 5. Distribution of abnormal semen parameters by age group Age group of husband (years) Semen fluid analysis

≤30

31 - 35

36 - 40

41 - 45

46 - 50

≥51

Azoospermia

0

3

1

1

1

0

Oligozoospermia

13

40

1

4

0

1

Teratozoospermia

4

4

1

2

3

1

Asthenozoospermia

4

21

6

9

3

2

Oligoteratozoospermia

1

2

1

1

1

1

Oligoasthenozoospermia

4

8

2

3

4

3

Asthenoteratozoospermia

1

3

1

2

2

1

OAT

1

2

1

2

1

0

References

Table 6. Cultured organisms (N=443) Organism

n (%)

Staphylococcus aureus

108 (24.2)

Klebsiella spp.

15 (3.4)

Escherichia coli

10 (2.3)

Streptococcus spp.

8 (1.8)

Candida spp.

6 (1.4)

Multiple coliforms

13 (2.9)

No organism isolated

283 (63.8)

between smoking habit, past infection with mumps (mumps orchitis) and past groin surgery and abnormal sperm parameters, although Okonofua et al.[26] also reported a significant association with alcohol consumption, among other factors.

A total of 371 (83.7%) couples involved in this study presented with secondary infertility, and 72 (16.3%) had primary infertility.

This is higher than figures reported in Ibadan and Ile-Ife[1,3] and reflects a growing pattern in the incidence of secondary infertility in this environment. This may be attributed to the significant contribution of obstruction of the female and male genital tract resulting from a high rate of genital infections in both female (postabortal sepsis, puerperal sepsis) and male partners in our setting.[1,6]

Conclusion

Male factor abnormalities remain significant contributors to infertility, as demonstrated in this study, and the importance of semen analysis cannot be overemphasised in the detection of sperm abnormalities. On the basis of this, society should particularly view infertility as a couple problem rather than ostracising women, and men should be encouraged to take up the challenge and present themselves for appropriate testing and treatment. Government should ensure the establishment of public centres for assisted reproduction, to bring this closer to the less privileged in the society and contribute to solving the challenges posed by male infertility.

Acknowledgements. The authors acknowledged the couples who participated in the study, and the laboratory scientists of the microbiology department of the hospital who assisted in the analysis of the semen samples.

61

1. Adeniji RA, Olayemi O, Okunlola MA, Aimakhu CO. Pattern of semen analysis of male partners of infertile couples at the University College Hospital, Ibadan. W Afr J Med 2003;22(3):243-249. http:// dx.doi.org/10.4314/wajm.v22i3.27959 2. Orhue A, Aziken M. Experience with a comprehensive university-hospital-based infertility program in Nigeria. Int J Gynaecol Obstet 2008;101(1):11-15. http://dx.doi.org/10.1016/j. ijgo.2007.09.034 3. Owolabi AT, Fasubaa OB, Ogunniyi SO. Semen quality of male partners of infertile couple in Ile-Ife, Nigeria. Nig J Clin Pract 2013;16(1):37-40. http://dx.doi.org/10.4103/1119-3077.106729 4. Umeora OJ, Igberase GO, Okogbenin SA, Obu ID. Cultural misconception and emotional burden of infertility in South-East Nigeria. Int J Gynaecol Obstet 2009;10:2. https://dx.doi. org/10.5580/10a 5. Jimoh AAA, Olawuyi TS, Omotoso GO, Oyewepo AO, Dare JK. Semen parameters and hormone profile of men investigated for infertility at Midland Fertility Centre, Ilorin, Nigeria. J Basic Appl Sci 2012;8(1):110-113. http://dx.doi.org/10.6000/1927-5129.2012.08.01.03 6. Butt F, Akran N. Semen analysis parameters: Experiences and insight into male infertility at a tertiary care hospital in Punjab. J Pak Med Assoc 2013;63(5):558-562.

7. Ugboma HAA, Obuna JA, Ugboma EW Pattern of seminal fluid analysis among infertile couples in a secondary health facility in south-eastern Nigeria. Res Obstet Gynaecol 2012;1(2):15-18. http:// dx.doi.org/10.5923/j.rog.20120102.01 8. Ojiyi EC, Dike EI, Anolue BU, Okendo C, Uzoma OI, Uzoma JI. Male factor subfertility at Imo State University Teaching Hospital, Orlu. Int J Gynaecol Obstet 2012;16(1):1-6.

9. Jajoo S, Kalyani KR. Prevalence of abnormal semen analysis in patients of infertility at a rural set up in Central India. Int J Reprod Contracept Gynaecol Obstet 2003;2(2):161-164. http://dx.doi. org/10.5455/2320-1770.ijrcog20130610 10. Shaikh AH, Khalique K, Tanq G, Soomro N. Pattern of semen abnormalities in couples with male factor infertility. Pak J Surg 2011;27(3):204-208.

11. Monavari SH, Vaziri MB, Khalili M, et al. Asymptomatic seminal infection of herpes simplex virus: Impact on male infertility. J Biomed Res 2013;27(1):56-61. http://dx.doi.org/10.7555/ jbr.27.20110139 12. Ugwuja EI, Ugwu NC, Ejikeme BN. Prevalence of low sperm count and abnormal semen parameters in male partners of women consulting at infertility clinic in Abakaliki, Nigeria. Afr J Reprod Health 2008;12(1):67-72.

13. Dohle GR. Inflammatory associated obstruction of the male reproductive tract. Andrologia 2003;35(5):321-324. http://dx.doi.org/10.1111/j.1439-0272.2003.tb00866.x 14. Chukwunyere CF, Awonuga DO, Ogo CN, Nwadike V, Chukwunyere KE. Pattern of seminal fluid analysis in male partners of infertile couples attending gynaecology clinic at Federal Medical Centre, Abeokuta. Niger J Med 2015;24(2):131-136.

15. World Health Organization. WHO laboratory manual for the examination and processing of human semen. 5th ed. Geneva: WHO, 2000:7-113.

16. Jarrow JP, Espeland MA, Lipshultz G. Evaluation of azoospermic patients. J Urol 1989;142(1):62-65.

17. Imade GE, Sagar AS, Pam IOA, Ujah PH. Semen quality in male partners of infertile couples in Jos, Nigeria. Trop J Obstet Gynaecol 2000;17(1):24-26.

18. Nwafia WC, Igweh JC, Udebuani IN. Semen analysis of infertile Igbo males in Enugu, Eastern Nigeria. Niger J Physiol Sci 2006;21(1-2):67-70. http://dx.doi/org/10.4314/njps.v21i1.54254 19. Imam MEI, Siuf A, Mansour MM, et al. Semen analysis of infertile Sudanese males in Gezira State, Central Sudan. Sudanese J Pub Health 2009;4(3):340-344.

20. Ikechebelu JI, Adinma JI, Orie EF, Ikegwuonu SO. High prevalence of male infertility in South Eastern Nigeria. J Obstet Gynaecol 2003;23(6):657-659. http://dx.doi.org/10.1080/014436103100 01604475 21. Ekwere PD, Archibong EI, Bassey EE, Ekabua JE, Ekanem EI, Feyi-Waboso P. Infertility among Nigerian couples as seen in Calabar. Port Med Journ 2007;2(1):35-40. http://dx.doi.org/10.4314/ phmedj.v2i1.38890 22. Ibekwe PC, Mbazor JO. Semen evaluation of infertile couples in Abakaliki, Nigeria. Ebonyi Med Journ 2002;1(1):33-37. http://dx.doi.org.10.4314/ebomed.v1i1.41517 23. Weber RF, Dohle GR, Romijn JC. Clinical laboratory evaluation of male infertility. Adv Clin Chem 2005;40:317-364. http://dx.doi.org/10.1016/s0065-2423(05)40008-6 24. Bayasgalan G, Naranbat D, Radnaabazar J, Lhagvasuren T, Rowe J. Male infertility: Risk factors in Mongolian men. Asian J Androl 2004;6(4):305-311.

25. Chia SE, Lim ST, Tay SK, Lim ST. Factors associated with male infertility: A case control study of 218 infertile and 240 fertile men. Br J Obstet Gynaecol 2000;107(1):55-61. http://dx.doi/ org/10.1111/j.1471-0528.2000.tb11579.x 26. Okonofua FE, Menakaya U, Onemu SO, Omo-Aghoja LO, Bergstrom S. A case-control study of risk factors for male infertility in Nigeria. Asian J Androl 2005;7(4):351-361. http://dx.doi. org/10.1111/j.1745-7262.2005.00046.x

SAJOG • December 2016, Vol. 22, No. 2

This open-access article is distributed under Creative Commons licence CC-BY-NC 4.0.

CASE REPORT

Skene’s gland duct cysts: The utility of vaginal/transperineal imaging in diagnosis and mapping for surgery A case series and review of the literature P F Kruger,1 MB ChB, FRCS(C) O&G; R Kung,2 MD, FRCS(C) O&G; F Hamidinia,3 CRGS, RDMS; R Rahmani,3 MB ChB, FRCP(C), DABR epartment of Obstetrics and Gynaecology, University of Cape Town, South Africa; formerly Department of Obstetrics and Gynecology, D University of Toronto, Canada 2 Department of Obstetrics and Gynecology, University of Toronto, Canada 3 Department of Radiology, University of Toronto, Canada 1

Corresponding author: P F Kruger (pkruger@urogynaecology.co.za)

We report three patients with Skene’s gland cysts diagnosed on transvaginal and transperineal 2D and 3D sonography. We demonstrate that pelvic floor imaging is a useful diagnostic tool and aids in preoperative surgical planning. The real-time nature of this form of imaging and the addition of 3D ultrasonography demonstrate internal architecture and spatial relationships of periurethral pathology, thus aiding presurgical mapping. In this case series patients were followed up and results of surgical procedures and histological findings were recorded. Marsupialisation and resection are accepted surgical options in symptomatic Skene’s gland cysts. Usually simple drainage will not suffice for management of a symptomatic Skene’s gland cyst, as demonstrated in one of our cases. Asymptomatic cysts can be followed conservatively.

S Afr J Obstet Gynaecol 2016;22(2):62-64. DOI:10.7196/SAJOG.2016.v22i2.1063

Periurethral glands, or Skene’s glands, are branched, tubular glands that are adjacent to the distal urethra. Usually Skene’s ducts run parallel to the long axis of the urethra for approximately 1 cm before opening into the distal urethra. Sometimes the ducts open into the area just outside the urethral orifice. Skene’s glands are the largest of the paraurethral glands; however, many smaller glands empty into the urethra. Skene’s glands are homologous to the prostate in the male.[1] Skene’s gland cysts are rare (incidence between 1/2 074 and 1/7 246). They may be either congenital or acquired. Presenting symptoms include a palpable or visible mass at the introitus, pain, dyspareunia, dysuria, a distorted voiding stream and a vaginal discharge. Periurethral cysts may be totally asymptomatic and discovered during routine pelvic examination.[2,3] We discuss the usefulness of imaging the labia, vagina and pel­ vic floor with 2D and 3D vaginal and transperineal pelvic floor sonography.

Ethical approval for this retrospective case series review was gran­ted, and the need for specific patient consent for the study was waived. All patients gave informed verbal consent for all aspects of the sonographic study and signed consent for procedures that were subsequently performed. Examination of the periurethral cysts was done transvaginally and transperineally in all cases.

Case series

Three consecutive patients with periurethral cysts who presented to a women’s ultrasound imaging centre in Toronto between June 2012 and February 2014 were included in this study. Patient information

was extracted from the office database of transvaginal sonographic information, maintained by one of the authors. This author also gathered follow-up information regarding surgical management and pathological results on these cases.

Sonographic technique and interpretation We demonstrate three patients with the sonographic diagnosis of Skene’s gland duct cysts, all proven histologically. Sonography was performed preoperatively with an E8 ultrasound machine (GE Healthcare, UK) employing 2D/3D endovaginal probe (RIC 5-9-D-model, Freq 8-10) to obtain endovaginal and transperineal pelvic floor imaging. All patients underwent colour Doppler sono­ graphy to evaluate the presence of vascularity.

A senior sonologist who specialises in female ultrasound imaging evaluated all three patients and made the diagnosis preoperatively. The ultrasound masses were evaluated for texture, presence or absence of colour flow, and the anatomical location was mapped on 3D sonography and 3D manipulation of the pelvic floor images.

Results

The results are summarised in Table 1. The patients were aged 29, 39 and 55 years. A total of three Skene’s gland duct cysts were seen in 3 patients, and in 2 of the patients additional vaginal cysts were identified. 3D perineal sonography was helpful in establishing the position of the masses in relation to the urethra, aiding the diagnosis.

The cysts varied in size between 2.8 and 3.6 cm (Figs 1 - 3).

SAJOG • December 2016, Vol. 22, No. 2

62

CASE REPORT Table 1. Sonographic findings Case

Age

Sonographic features and size

Colour Doppler sonography

3D sonography

Associated findings

1

29

Lowlevel echoes measuring 3.0 × 3.0 × 2.1 cm (Fig. 1D)

Present in second mass indicating septation or previous infection

Mass lateral to urethra and separate from it

Second mass in region of right labia. Measuring 1.6 × 1.4 × 2.1 cm. More echogenic and has colour flow. Differential diagnosis includes canal of Nuck cyst

2

55

Tubular cystic structure which encircles the urethra. Rightsided going anterior and posterior to urethra towards the left but dominantly on the right. Septated. Measures 3.6 × 1.3 × 2.2 cm (Fig. 2)

Vaginal wall Gartner duct cyst measuring 2.2 × 1.3 cm

3

39

Hypoechoic lesion measuring 2.8 × 2.7 × 1.6 cm with lowlevel echoes (Fig. 3A & B)

None

A

B

C

D

Discussion

Fig. 1. A. Sonogram of mucinous mass in region of right labia. B. 2D transverse view. C. 2D sagittal view. D. Soft-tissue rendered image demonstrates coronal view of the mass and its relation to the distal urethra. The mass is displacing the urethra to the left.

Management and follow-up The referring physicians were informed of the results and management decisions were made after discussion with the patients. Two of the three patients pro­ ceeded with surgical excision as the ini­ tial approach. A complete resection of the peri­ ure­ thral cyst was possible in case 1. Histopathology confirmed a be­nign cyst with transitional epithelium.

The patient had resolution of her symptoms. The second case presented a difficult sur­ gical excision. Owing to the nature of the cyst and proximity

to the urethra, it was not possible to completely excise it. Histopathology confirmed the diagnosis of a benign cyst with transitional epithe­ lium in keeping with a Skene’s gland duct cyst.

This patient was asymptomatic at her 6-week postoperative visit. Case 3 op­ted for drainage at the time of clinical diagnosis. She experienced recur­ rence within a few weeks and then pro­cee­­­­ded with marsupialisation. Histopathology was not obtained. The patient remained asymptomatic at her 6-week follow-up visit.

63

SAJOG • December 2016, Vol. 22, No. 2

The spectrum of female urethral and peri­ urethral disorders includes both benign and malignant entities.[5] Benign aetiologies include urethral caruncles, Skene’s gland abscess/cysts, mucosal prolapse, ectopic ureterocoele, urethral diverticulum, vaginal wall cysts, Gartner’s duct cysts, leiomyo­ mas and hamartomas.[6] Rarely, malig­nant lesions present as periurethral masses, inclu­ ding adenocarcinoma, squa­ mous cell carcinoma, transitional cell carcinoma, histio­cytoma and sarcoma.[5,6] The diagnosis of female periurethral disease is challenging for clinicians because patients present with nonspecific signs and symptoms, including pelvic pain, dyspareunia, dysuria, urinary frequency, urinary urgency, incontinence, urethral bleed­ing and urinary tract infections.[7] Clinical examination can be unrelia­ ble in distinguishing among the many types of urethral or vaginal wall masses.

Although conventional imaging studies such as voiding cysto-urethrography and retrograde double-balloon positive-pressure urethrography may be helpful, they are invasive and cannot help map periurethral tissues preoperatively.[8] MRI has become the imaging modality of choice for diagnosis and preoperative planning in female patients with urethral and periurethral disease because of its superb resolution, increased signal-tonoise ratio, and multiplanar capability. The drawback to MRI is high cost and limited access.

With the advent of improved ultrasound high-frequency probes and 3D pelvic floor imaging, transvaginal and transperineal ultra­

CASE REPORT

A

B

Fig. 2. Transperineal 3D pelvic floor image demonstrating the exact anatomy relationship of Skene’s gland duct cyst wrapping the urethra. sonography is an emerging modality that is being used to evaluate female urethral and periurethral disease in a more cost-effective manner.[5] This form of imaging also has the advantage of being a real-time investigation.

With the addition of 3D scanning, the internal architecture and spatial relation­ ships of periurethral pathology can be clearly displayed. The ability to visualise internal charac­ teristics of vaginal cysts and masses is important because it can help surgeons with management decisions. Evaluating the extent of a vaginal mass or cyst and its anato­ mical landmarks gives important information for precise diagnosis and surgical finding expectations.[1,9] We agree with recent publications that transvaginal ultra­ sonographic scanning with the addition of 3D scanning can make a significant contribution to the diagnosis

Fig. 3. A. 2D transperineal sonogram image of mucinous mass adjacent to the urethra. B. 3D ultrasound demonstrates the accurate relationship of mucinous mass to other structures in the pelvic floor. and management of periurethral disease by supplying preoperative information that can aid surgical planning, obviating the need for more costly MRI imaging.[9] This was well illustrated in this case series.

References 1. Katz VL. Comprehensive gynecology. 5th ed. Philadelphia: Mosby Elsevier, 2007.

2. Battaglia C, Venturoli S. 3-D ultrasonographic appearance of two intermittent paraurethral cysts: A case report. J Sex Med 2010;7(8):2903-2906. http://dx.doi.org/10.1111/j.17436109.2009.01602.x 3. Sharifi-Aghdas F, Daneshpajooh A, Mirzaei M. Parauerthral cyst in adult women: Experience with 85 cases. Urol J 2014; 1(5):1896-1899.

SAJOG • December 2016, Vol. 22, No. 2

64

4. Lucioni A, Rapp DE, Gong EM, Fedunok P, Bales GT. Diagnosis and management of periurethral cysts. Urol Int 2007;78(2):121-1215. http://dx.doi.org/10.1159/000098068 5. Chaudhari VV, Patel MK, Douek M, Raman SS. MR imaging and US of female urethral and periurethral disease. Radiographics 2010;30(7):1857-1874. http://dx.doi. org/10.1148/rg.307105054 6. Dmochowski RR, Ganabathi K, Zimmern PE, Leach GE. Benign female periurethral masses. J Urol 1994;152 (6 Pt 1):1943-1951.

7. Handel LN, Leach GE. Current evaluation and management of female urethral diverticula. Curr Urol Rep 2008;9(5):383-388.

8. Rufford J, Cardozo L. Urethral diverticula: A diagnostic dilemma. BJU Int 2004;94(7):1044-1047. http://dx.doi. org/10.1111/j.1464-410x.2004.05125.x 9. Shobeiri SA, Rostaminia G, White D, Quiroz LH, Nihira MA. Evaluation of vaginal cysts and masses by 3-dimensional endovaginal and endoanal sonography. J Ultrasound Med 2013;32(8):1499-1507. http://dx.doi.org/10.7863/ ultra.32.8.1499

This open-access article is distributed under Creative Commons licence CC-BY-NC 4.0.

CASE REPORT

Three-dimensional colour Doppler of ductus venosus agenesis in the first trimester D Singh, MD, DNB Radiodiagnosis; L Kaur, MD Radiodiagnosis Prime Diagnostic Centre and Heart Institute, Chandigarh, India Corresponding author: D Singh (docdivyas@yahoo.co.in)

Ductus venosus (DV) has a pivotal role in the fetal circulation. It serves as a conduit connecting the fetal umbilical and portal venous system with the inferior vena cava. The absence of DV is an uncommon anomaly. In case of agenesis of DV, the umbilical vein joins the fetal systemic venous circulation via the intrahepatic or extrahepatic route. We report a case of absent DV with associated anomaly diagnosed in the first trimester using three-dimensional (3D) colour Doppler.

S Afr J Obstet Gynaecol 2016;22(2):65-66. DOI:10.7196.SAJOG/2016.v22i2.1085

DV is an important vascular channel in the fetal circulation. It preferentially channels the oxygenated umbilical venous blood to the vital organs like the brain and the heart. It is routinely evalu­a­ted in the first trimester as its abnormality serves as a marker for fetal aneuploidy and cardiac anomalies. Agenesis of DV is a relatively uncommon anomaly. Its prevalence is 1:2 500 between 11 and 14 weeks of gestation.[1] It can be detected in the first trimester using colour Doppler. Our case illustrates the diagnosis of this entity using three-dimensional (3D) colour Doppler.

Case report

A 27-year-old primigravida attended our institution for firsttrimester screening sonography. Transvaginal sonogram revealed a cystic lesion with septations measuring 3.6 × 2.4 cm in the fetal neck, suggestive of a cystic hygroma (Fig. 1A). In addition, there was bilateral pleural effusion with generalised subcutaneous oedema suggestive of hydrops fetalis (Fig. 1B). Colour Doppler sonography revealed the umbilical vein draining into the portal sinus with non-visualisation of the DV, which would connect the portal sinus with the inferior vena cava (IVC) (Figs 1C and 1D). Hence, a diagnosis of agenesis of DV with intrahepatic drainage was made.

There were no other associated anomalies. The couple was offered fetal karyotyping. They opted for termination of the pregnancy.

Karyotype of the products of conception post termination of the pregnancy revealed 45XO karyotype consistent with diagnosis of Turner’s syndrome.

Discussion

DV performs a crucial role of directing 20 - 30% of the oxygenated blood coming from the umbilical vein towards the right atrium, and from there through the foramen ovale into the left atrium, and thereafter the brain and the coronary arteries.[2] In the absence of DV, the umbilical venous blood reaches the heart via other routes: intrahepatic or extrahepatic. In the intrahepatic route, the umbilical venous blood drains directly into the portal sinus. From the portal sinus, it reaches the systemic venous circulation either through an abnormal venous channel to the right atrium or via hepatic

65

sinusoids to the hepatic veins. The extrahepatic route bypasses the portal sinus, such that an abnormal venous channel connects the umbilical vein directly to the systemic circulation. This channel may join the right atrium, IVC, renal vein, iliac veins or, less commonly, the left atrium and coronary sinus.[3,4] Agenesis of DV is diagnosed using colour Doppler. The normal DV is visualised in the sagittal section of the fetal abdomen. It is seen beyond the umbilical vein as an area of aliasing in continuity with the portal sinus and draining into the IVC (Figs 2A and 2B). On

A

B

C

D

Fig. 1. A. B-mode sonogram of the fetal neck axial section showing an anechoic lesion with septation suggestive of a cystic hygroma (arrow). B. Transverse section of the fetal thorax demonstrating bilateral pleural effusion (asterisk). Note the diffuse oedema of the chest wall (arrow). C. Colour Doppler image of the sagittal section of the fetal abdomen showing the umbilical vein (UV) draining into the portal sinus (PS). No channel is seen connecting the PS to the IVC. These features are suggestive of agenesis of DV. Note the hepatic vein (HV) draining into the IVC. D. 3D colour Doppler image showing the venous connections in the fetus with DV agenesis. (UA = umbilical artery, AO = descending aorta.)

SAJOG • December 2016, Vol. 22, No. 2

CASE REPORT

A

B

Fig. 2. A. Sagittal colour Doppler image of a normal first-trimester fetus showing the UV connected to the systemic venous circulation by the DV, which is seen as an area of colour aliasing. B. 3D colour Doppler image showing the normal DV. spectral interrogation, the DV has a triphasic waveform distinct from the monophasic waveform of the umbilical vein. With the recent impetus on first-trimester screening for aneuploidy and availability of highresolution transducers, agenesis of DV can be diagnosed in the first trimester.

The use of 3D colour Doppler gives a panoramic view of the fetal circulation and helps better understanding of the vascular connections. Thubert et al.[5] reported a case of absent DV in the third trimester using 3D power Doppler. Our case illus­ trates the 3D colour Doppler findings of DV agenesis in the first trimester. In case of non-visualisation of the DV, a detailed evaluation of the fetal anatomy is mandatory as it may be associated with several anomalies.[6] A recent study demonstrated that in nearly half the cases with absent DV, the fetal nuchal translucency (NT) is above the 95th percentile. In the group with

increased NT, more than 40% of fetuses had chromosomal abnormalities. The most commonly associated aneuploidy was Turner’s syndrome. It was observed that in most fetuses with NT below the 95th percentile, absent DV was an isolated finding and these pregnancies had a normal outcome.[1] Therefore, in case of absence of DV in the first trimester, the prognosis is determined by the NT measurement. If the NT is normal, the prognosis is favourable.

In cases with isolated absent DV, the site of umbilical venous drainage is important.

Direct drainage of the umbilical venous blood into the right atrium, IVC or its branches can cause volume overload on the fetal heart with the potential to cause cardiomegaly and cardiac failure resulting in hydrops fetalis.[7] Therefore, the prognosis is most favourable for isolated DV agenesis with intrahepatic drainage.

Although our case had an intrahepatic drainage of the umbilical vein, it was

SAJOG • December 2016, Vol. 22, No. 2

66

associated with cystic hygroma and hydrops fetalis. Karyotype of the conceptus revealed Turner’s syndrome. This underscores the sig­ nificance of searching for associated anomalies in agenesis of DV which can determine the prognosis.

In conclusion, agenesis of DV can be diagnosed in the first trimester using colour Doppler. The umbilical venous drainage pathway can be exquisitely depicted using 3D Doppler. The diag­nosis of DV agenesis warrants detailed structural evaluation of the fetus for associated abnormalities. Fetal karyotyping should be offered to rule out chromosomal anomalies. If absent DV is seen in isolation, close fetal surveillance is recommended, especially in cases with extrahepatic drainage because of the risk of developing cardiac failure.

References 1. Staboulidou I, Pereira S, Cruz JdeJ, Syngelaki A, Nicolaides KH. Prevalence and outcome of absence of ductus venosus at 11(+0) to 13(+6) weeks. Fetal Diagn Ther 2011;30(1):35-40. http://dx.doi.org/10.1159/000323593 2. Kiserud T. The ductus venosus. Semin Perinatol 2001;25(1):11-20. http://dx.doi.org/10.1053/sper.2001.22896 3. Berg C, Kamil D, Geipel A, et al. Absence of ductus venosus – importance of umbilical venous drainage site.

Ultrasound Obstet Gynecol 2006;28(3):275-281. http://dx.doi. org/10.1002/uog.2811 4. Jaeggi ET, Fouron JC, Hornberger LK, et al. Agenesis of the ductus venosus that is associated with extrahepatic umbilical vein drainage: Prenatal features and clinical outcome. Am J Obstet Gynecol 2002;187(4):1031-1037. http://dx.doi. org/10.1067/mob.2002.126292 5. Thubert T, Levaillant JM, Stos B, Benachi A, Picone O. Agenesis of the ductus venosus: Three-dimensional power Doppler reconstruction. Ultrasound Obstet Gynecol 2012;39:118-120. http://dx.doi.org/10.1002/uog.10155 6. Volpe P, Marasini M, Caruso G, et al. Prenatal diagnosis of ductus venosus agenesis and its association with cytogenetic/ congenital anomalies. Prenat Diagn 2002;22(11):995-1000. http://dx.doi.org/10.1002/pd.456 7. Acherman RJ, Evans WN, Galindo A, et al. Diagnosis of absent ductus venosus in a population referred for fetal echocardiography: Association with a persistent portosystemic shunt requiring postnatal device occlusion. J Ultrasound Med 2007;26(8):1077-1082.

This open-access article is distributed under Creative Commons licence CC-BY-NC 4.0.

CASE REPORT

Respiratory arrest caused by a large uterine myoma M S Kim,1 MD; G H Lee,1 MD; M C Choi,2 MD; H Park,2 MD, PhD; S G Jung,2 MD, PhD; K A Kim,3 MD, PhD Department of Obstetrics and Gynecology, CHA Bundang Medical Center, CHA University, Seongnam-si, Gyeonggi-do, Korea Comprehensive Gynecologic Cancer Center, Department of Obstetrics and Gynecology, CHA Bundang Medical Center, CHA University, Seongnam-si, Gyeonggi-do, Korea 3 Department of Radiology, CHA Bundang Medical Center, CHA University, Seongnam-si, Gyeonggi-do, Korea 1

2

Corresponding author: M C Choi (oursk79@cha.ac.kr)

Large abdominal masses increase intra-abdominal pressure, thus changing the haemodynamics of the patient by elevating the dia­ phragm and causing partial occlusion of the inferior vena cava (IVC). Large abdominal masses present many challenges, including life-threatening risks due to severe cardiovascular, pulmonary, and circulatory problems, as well as technical difficulties of surgery and postoperative complications. We report a case of a large pelvic-abdominal myoma with perioperative pulmonary compromise. The goal of this report was to familiarise other surgeons with the alterations in the pathophysiology and management of large abdominal masses.

S Afr J Obstet Gynaecol 2016;22(2):67-68. DOI:10.7196/SAJOG.2016.v22i2.1094

It is rare to see patients present with exertional dyspnoea caused by a pelvic mass such as uterine myoma without underlying cardio­ pulmonary disease. The most common symptom of myoma is men­ strual disturbance. However, in cases of large uterine myoma, intraabdominal pressure (IAP) can increase, which interferes with the pulmonary, renal, splanchnic and cardiovascular systems by elevating and splinting the diaphragm and partially occluding the inferior vena cava (IVC).[1] If untreated, IAP rises and multiple organ failure begins, and may progress to abdominal compartment syndrome (ACS), defined as a sustained IAP of 20 mmHg or higher and associated with new organ dysfunction.[2] Removal of the mass is the treatment of choice for ACS; however, gradual decompression is necessary.

We report a case of perioperative respiratory failure precipitated by increased IAP, which was caused by a large abdo­mino­pelvic myoma.

Case report

A 42-year-old virgin woman presented to the emergency room for gradually worsening exertional dyspnoea and remarkable abdominal distension. Gradual abdominal distension had been noted by the patient over the past 3 years, but she became symptomatic only a week before admission. Medical and surgical histories were unremarkable. The patient complained of tachypnoea, and her oxygen saturation (SpO2) on room air was 89%. Abdominal examination revealed generalised abdominal distension with a mass located through the entire abdomen. The patient’s height was 170 cm, and weight was 55 kg. Laboratory tests indicated hypoxaemia in arterial blood gas analysis (pH 7.43, partial pressure of carbon dioxide (pCO2) 45.5 mmHg, partial pressure of oxygen (pO2) 54.7 mmHg) and mildly elevated liver enzymes (aspartate aminotransferase (AST) 68 IU/L, alanine transferase (ALT) 72 IU/L).

Other biochemical parameters, including tumour marker levels (CA 125 15.3 U/mL and CA 19-9 26.82 U/mL) were within normal ranges. Abdominopelvic computed tomography (CT) revealed a large, heterogeneous, multiseptated ovoid pelvic mass measuring 37 × 25 × 12 cm, occupying the pelvis and abdomen up to the diaphragm. It was unclear whether it was from the uterus, adnexa,

67

or another organ (Fig. 1). Chest CT angiogram showed no signs of pulmonary embolism and the echocardiogram revealed normal ejection fraction (EF, 60%) and diastolic function (E/A ratio, 1.33).

To improve her respiratory impairment, 8 L of oxygen was admin­ istered in addition to intravenous diuretics. She was admitted to the Gynecologic Cancer Center for further evaluation of the abdominal mass. On the second hospital day, she complained of aggravated tachyp­noea (respiratory rate, 40 per minute) and oliguria (urine out­put ≤20 mL/h). The clinical signs were thought to be suggestive of organ dysfunction due to progressive ACS. The gynaecological oncology team decided to perform exploratory laparotomy.

At the time of surgery, we noted a 37 × 25 × 12 cm sized sub­ serosal myoma-like mass arising from the uterus with secondary degeneration. Examination of the frozen section indicated a uterine leiomyoma, not a malignant condition. After ascertaining that the other pelvic organs were grossly normal, myomectomy was per­ formed in view of her nulliparity.

After normal respiratory function was confirmed, she was extu­ bated in the operating room and sent to the intensive care unit for monitoring of signs of multi-organ failure. Her vital signs were stable postoperatively; however, 1 hour later, she was tachypnoeic (respiratory rate 35 per minute) and her SpO2 abruptly dropped to 50%. Because of the possibility of inadequate ventilation, re-intubation with positive pressure ventilation to support adequate respiration was performed. The following day, the patient’s urine output and liver enzyme levels returned to normal limits. After respiratory compromise had improved, she was extubated and trans­ ferred to the general ward.

The final pathology indicated a 6.54 kg uterine leiomyoma with hydropic change. She was discharged on postoperative day 6 with no pulmonary symptoms. The postoperative process was uneventful. The patient remains well 31 months postoperatively.

Discussion

Uterine myomas are benign, monoclonal tumours of the smooth muscle of the myometrium. Uterine myomas occur in approximately

SAJOG • December 2016, Vol. 22, No. 2

CASE REPORT

Fig. 1. Abdominopelvic CT imaging of the patient; coronal view image depicting a 47 × 25 × 12 cm heterogeneous, ovoid mass.

15 - 20% of women in their reproductive years.[1] In the present case, the large abdominal mass induced an increase in IAP that resulted in changes in the haemodynamics of the patient. A large intra-abdominal mass may cause immobility, breathlessness, and the inability to lie supine. Increased IAP can cause partial occlusion of the IVC, impaired venous return, and decreased cardiac output.

Chronic vena cava compression produces venous stasis and dilated superficial abdominal veins, so-called caput medusa.[3] Our patient experienced immobility and breathlessness but had normal cardiac output on echocardiography.

Intra-abdominal hypertension (IAH) is defined as a sustained or repeated pathological elevation of IAP to 12 mmHg or higher.[4] IAH is diagnosed with a combination of clinical findings and monitoring of bladder pressure, but the diagnosis is often based only on clinical examination. Chronic IAH involves the insidious development of increased pressure over months (such as that during pregnancy) or years (such as a large intra-abdominal tumour or ascites).[4] In our case, gradual growth of the uterine myoma occurred over 3 years. IAH may have caused preoperative pulmonary compromise in the present case. Elevation and splinting of the diaphragm by the mass reduces functional residual capacity, vital capacity, and total lung capacity, leading to tachypnoea and arterial hypoxaemia.

Pronounced rib flaring and attenuated diaphragm and abdominal muscles have also been observed.[5] Increased IAP may affect the viscoelastic properties of the thoraco-abdominal region, elevating chest wall resistance, decreasing chest wall compliance and inducing pulmonary oedema.

Undiagnosed and untreated chronic increased IAP can pro­gress to an acute condition, ACS. ACS is defined as a sustained IAP of 20 mmHg or higher and is associated with new organ dys­func­­­ tion.[4] In the present case, respiratory distress, hypoxaemia, decreased urine output, and elevated liver enzyme levels were observed, that could be signs of multiple organ dysfunction, suggestive of ACS. ACS can be treated with abdominal decom­ pression.[4] Surgical removal of the mass was attempted in the present case; however, the diaphragm was lax, and the abdominal wall was weak after the removal of the myoma, thus resulting in pulmonary compromise after surgery.[5] Another possible mecha­ nism for postoperative pulmonary compromise is the sudden expansion of chronically collapsed lungs after removal of the large abdominal mass, causing re-expansion pulmonary oedema (RPE).[6] There is no standard method to prevent RPE but re-expanding the lung very slowly with a low tidal volume may be helpful.[6] Any patient admitted for the removal of a large abdominal mass should undergo cardiac and pulmonary assessments. During the ope­ra­tion, placing the patient in the reverse Trendelenburg position with a tilt to the left side and continuous monitoring of blood gases may avert the supine hypotensive syndrome and prevent further respiratory difficulties.[5] Gradual decompression of the mass reduces supine hypotension, and increases vital capacity. If the large mass is cystic, gradual decompression of the cystic fluid by paracentesis preoperatively and intraoperatively may prevent the pheno­menon of splanchnic shock, which occurs when the compressed IVC is suddenly released.[3,6] In case of a large, solid abdominal mass com­ pressing the IVC, rolling the mass gradually off the IVC may be the optimal surgical technique.[3] Delayed extubation is sometimes recommended to prevent pulmonary compromise.[3,5] If extubation had been delayed in the present case to support the lax diaphragm and weak respiratory muscles, postoperative pulmonary compromise could have been prevented.

Uterine myomas are frequently encountered in the gynaecological department. Most clinicians perform surgery without any pre­ paration; however, a chronic large abdominal myoma can alter the patient’s haemodynamics peri-operatively. We make the point that large abdominal masses, regardless of whether they are benign or malignant, can increase the risk of IAH and progression to ACS. Furthermore, surgeons should be familiar with the alterations in physiology and management of large abdominal mass.

References 1. Abdul Ghaffar NA, Ismail MP, Nik Mahmood NM, et al. Huge uterine fibroid in a postmenopausal woman associated with polycythemia – a case report. Maruritas 2008;60:177-179. http://dx.doi. org/10.1016/j.maturitas.2008.03.013 2. Gattinoni L, Pelosi P, Suter PM, et al. Acute respiratory distress syndrome caused by pulmonary and extrapulmonary disease: different syndromes? Am J Respir Crit Care Med 1998;158:3-11. http://dx.doi.org/10.1164/ajrccm.158.1.9708031 3. Kim YT, Kim JW, Choe BH. A case of huge ovarian cyst of 21-year-old young woman. J Obstet Gynaecol Res 1999;4:275-279. http://dx.doi.org/10.1111/j.1447-0756.1999.tb01162.x 4. Peparini N, Di Matteo FM, Silvestri A, et al. Abdominal hypertension in Meigs’ syndrome. Eur J Surg Oncol 2008;34:938-942. http://dx.doi.org/10.1016/j.ejso.2007.08.006 5. Hoile RW. Hazards in the management of large intra-abdominal tumours. Ann R Coll Surg Engl 1976;58:393-397.

6. Ohashi N, Imai H, Tobita T, et al. Anesthetic management in a patient with giant growing teratoma syndrome: A case report. J Med Case Rep 2014;8:32. http://dx.doi.org/10.1186/1752-1947-8-32

SAJOG • December 2016, Vol. 22, No. 2

68

CPD

True (A) or false (B): An Afrikaans pelvic organ prolapse questionnaire 1. The P-QOL questionnaire stands for the prolapse quotient of limi­­tation questionnaire.

2. According to the article, the P-QOL has been previously trans­ lated into four languages, all European.

Birth preparedness in southern Nigeria 3. The article reveals that one-third of the women studied did not have a good knowledge of being pregnant or complications.

Peripartum hysterectomies in a South African setting 4. The article quotes a review from the USA in which the quoted rate of peripartum hysterectomy was 1 per 10 000 live births.

5. Some definitions of peripartum hysterectomy include hysterec­ tomies performed for uncontrollable haemorrhage and exclude those performed for sepsis.

6. The definition of peripartum for the description of a peripartum hysterectomy ranges from within the first 24 hours to within the first 6 weeks of delivery.

Abnormal semen parameters in Nigeria 7. According to the article, 25% of semen samples analysed were abnormal.

8. The most common semen abnormality identified was astheno­ zoo­­spermia.

9. Unexplained infertility accounts for about 20 - 40% of couples investigated.

10. Positive semen cultures were found in 36.2% of samples and the most common organism found was Staphyloccocus aureus.

11.

World Health Organization (WHO) criteria for semen para­ meters were defined in 1999 and in 2010.

12. According to the 2010 WHO criteria, oligozoospermia is defined as a sperm count of <40 million/mL. Skene’s gland duct cysts 13. Usually simple drainage will not suffice for the management of Skene’s gland duct cysts.

14. S kene’s glands are homologous to the prostate gland in the male.

Doppler imaging of ductus venosus (DV) agenesis 15. The DV directs 20 - 30% of the oxygenated blood from the um­bilical vein to the left atrium.

16. The article quotes a recent study which found that in approx­ imately 50% of women with an absent DV the nuchal trans­ lucency (NT) was above the 95th centile.

17. In the quoted series, when an absent DV was associated with an increased NT there was a chromosome abnormality in 40% of cases, the most common aneuploidy being Turner’s syndrome.

18. Agenesis of the DV in the absence of a chromosome abnormality has a poor prognosis.

Respiratory arrest and a large myomatous uterus 19. Abdominal compartment syndrome is defined as a sustained intra-abdominal pressure of 20 mmHg or higher associated with organ dysfunction secondary to it.

20. Following decompression of a pressurised abdomen, re-expan­ sion pulmonary oedema may result.

The CPD programme for SAJOG is administered by Medical Practice Consulting: CPD questionnaires must be completed online at www.mpconsulting.co.za A maximum of 3 CEUs will be awarded per correctly completed test.

Accreditation number: MDB015/178/02/2016(Clinical)

69

SAJOG • December 2016, Vol. 22, No. 2