SAJCC Vol 31, No 1 (2015) by HMPG

SAJCC THE SOUTHERN AFRICAN JOURNAL OF CRITICAL CARE

June 2015 Vol. 31 No. 1

• H1N1 in the paediatric ICU • Obstetric patient outcome in ICU • M icrobial resistance patterns in a secondary centre • Ventilator-associated pneumonia prevalence in KZN • Verifying endotracheal tube placement

THE OFFICIAL JOURNAL OF THE CRITICAL CARE SOCIETY OF SOUTHERN AFRICA

Evidence. Experience. Confidence.

In ICU patients, appropriate antifungal choices are critical in the management of invasive fungal infections. ¹

Help them in the fight of their lives Cancidas®: • ESCMID 2012 Guidelines for the Diagnosis and Management of Candida Diseases Strongly Recommend CANCIDAS® 2

• •

Broad Spectrum Coverage Against Many Species of Candida and Aspergillus. 3 The only Echinochandin registered in children from 3 months and above. 3

Indications 3: • Empirical therapy for presumed fungal infections in febrile, neutropaenic patients. • Treatment of Invasive Candidiasis, including candidaemia. • Treatment of Oesophageal and Oropharyngeal Candidiasis where IV antifungal therapy is appropriate.

•

Treatment of invasive Aspergillosis in patients who are refractory or intolerant of other therapies including amphotericin B, lipid formulations of amphotericin B and itraconazole.

References: 1. Schelenz, S, Management of Candidiasis in the intensive care unit. J Antimicrob Chemoth 2008; 61 ( Suppl 1) i31-i34. 2. Cornely OA, Bassetti M, Calandra T, et al for ESCMID Fungal Infection Study Group(EFISG). ESCMID 2012 Guidelines for the Diagnosis and Management of Candida Diseases: non-neutropaenic adult patients. Clin Microbiol Infect 2012;18(suppl7)19-37. 3. Cancidas Approved Package Insert-14 September 2012. For full prescribing information refer to the package insert approved by the Medicines Regulatory Authority.

S4 CANCIDAS® 50 mg Lyophilised Powder for Solution for Infusion. Reg. No. 37/20.2.2/0544. Each vial contains 50 mg caspofungin anhydrous free base, equivalent to 55,5 mg caspofungin acetate. S4 CANCIDAS® 70 mg Lyophilised Powder for Solution for Infusion. Reg. No. 37/20.2.2/0545. Each vial contains 70 mg caspofungin anhydrous free base, equivalent to 77,7 mg caspofungin acetate. MSD (Pty) Ltd (Reg. No. 1996/003791/07), Private Bag 3, Halfway House 1685. Tel: (011) 655-3000. www.msd.co.za. Copyright © 2013 Merck Sharp & Dohme Corp., a subsidiary of Merck & Co., Inc., Whitehouse Station, NJ, USA. All rights reserved.MSD. AINF-1099838-0000

14945

SAJCC THE SOUTHERN AFRICAN JOURNAL OF CRITICAL CARE

The Official Journal of the Critical Care Society of Southern Africa June 2015 Vol. 31 No. 1

CONTENTS 3

EDITORIAL

H1N1 influenza (‘swine ‘flu’) in the paediatric ICU in South Africa

S Kling

ARTICLES

4 Influenza A(H1N1)pdm09 in critically ill children admitted to a paediatric intensive care unit, South Africa J O Ahrens, B M Morrow, A C Argent

8 Obstetric intensive care admissions at a tertiary hospital in Limpopo Province, South Africa T S Ntuli, G Ogunbanjo, S Nesengani, E Maboya, M Gibango

12 Organisms cultured and resistance patterns seen in a secondary referral centre ICU and burns unit B Greatorex, G Oosthuizen

16 Incidence and outcome of ventilator-associated pneumonia in Inkosi Albert Luthuli and King Edward VIII Hospital surgical intensive care units A Awath Behari, N Kalafatis

20 Endotracheal tube verification in adult mechanically ventilated patients P Jordan, W Ten Ham, D Fataar

ABSTRACTS

24 Abstracts of presentations at the congress of the Critical Care Society of Southern Africa, July 2015

CASE REPORT

30 Postoperative internal iliac artery embolisation as salvage therapy for bleeding in an HIV-positive patient with giant cell tumour of bone T van den Heever, C L Barrett, M J Webb, M G L Spruyt, C J Louw

OBITUARY

32 Max Klein A C Argent

EDITOR Lance Michell DEPUTY EDITOR Brenda Morrow ASSOCIATE EDITORS Andrew Argent (UCT) Dean Gopalin (UKZN) Lauren Hill (Private Practice) Ivan Joubert (UCT) David Linton (Hadassa University, Jerusalem) Rudo Mathiva (Wits) Mervyn Mer (Wits) Sam Mokgokong (UP) Fathima Paruk (Wits) Helen Perrie (Wits) Guy Richards (Wits) Juan Scribante (Wits) PUBLISHED BY Health and Medical Publishing Group (HMPG), a subsidiary of the South African Medical Association Suites 9 & 10, Lonsdale Building, Gardner Way, Pinelands, 7405 Tel: 082 635 9825 Email: publishing@samedical.org HMPG CEO AND PUBLISHER Hannah Kikaya HMPG EDITOR-IN-CHIEF Janet Seggie CONSULTING EDITOR J P de V van Niekerk EXECUTIVE EDITOR Bridget Farham SCIENTIFIC EDITOR Simon Nye TECHNICAL EDITORS Emma Buchanan Paula van der Bijl PRODUCTION MANAGER (CMC) Emma Jane Couzens DTP & DESIGN (CMC) Carl Sampson DISTRIBUTION MANAGER Edward Macdonald HEAD OF SALES AND MARKETING Diane Smith tel. (012) 481-2069 ISSN 1562-8264

Correction In the article ‘Dysmagnesaemia and outcome in a trauma ICU’ by Ilicki et al., which appeared on pp. 45-50 of Volume 30(2) of SAJCC, T C Hardcastle and D J J Muckart are both affiliated with the University of KwaZulu-Natal. Articles listed in EXCERPTA MEDICA (EM BASE), BIOLOGICAL ABSTRACTS (BIOSIS), SCIENCE CITATION INDEX (SCISEARCH), CURRENT CONTENTS/CLINICAL MEDICINE, SCIENTIFIC ELECTRONIC LIBRARY ONLINE (SCIELO) This Journal is accredited by the South African Department of Higher Education and Training. This Journal is also published online at www.sajcc.org.za. Critical Care Society of Southern Africa Society contact details: Lorraine Palm, Administrative Secretary, P O Box 521, Melville, 2109 cell: 083 464 1304, fax: 0866 340 259 e-mail: critcare@tiscali.co.za website: www.criticalcare.org.za The views and opinions expressed in the SA Journal of Critical Care are those of the authors and do not necessarily reflect the views of the Editors of the Journal or the Critical Care Society of Southern Africa. The appearance of advertising in the Journal does not denote a guarantee or an endorsement by the Society of the products or the claims made for the products by the manufacturers. Copyright 2000 by the SA Medical Association. This work is copyright under the Berne Convention. It is also copyright in terms of the Copyright Act 98 of 1978. No part of this publication may be reproduced, stored in a retrieval system, or transmitted in any form or by any means, electronic, mechanical, photocopying, recording, or otherwise, without permission of the copyright holder.

Plagiarism is defined as the use of another’s work, words or ideas without attribution or permission, and representation of them as one’s own original work. Manuscripts containing plagiarism will not be considered for publication in the SAMJ. For more information on our plagiarism policy, please visit http://www.samj.org.za/index.php/ samj/about/editorialPolicies Use of editorial material is subject to the Creative Commons Attribution – Noncommercial Works License. http://creativecommons.org/licenses/by-nc/3.0

PRINTED BY TANDYM PRINT

EDITORIAL

H1N1 influenza (‘swine ‘flu’) in the paediatric ICU in South Africa The 2015 influenza season officially started during the second week of May, according to the National Institute of Communicable Diseases (NICD). [1] The NICD website explains that the influenza strains in circulation change every year and that this year the ‘swine ‘flu’ strain (influenza A(H1N1)pdm09) is behaving similarly to any of the other influenza strains. However, in 2009 this strain caused an influenza pandemic.[1] Influenza viruses are endemic in many species, including humans, birds and pigs, and they are known to result in annual seasonal outbreaks of disease, which cause both significant morbidity and mortality.[2] Occasionally, however, influenza viruses cause pandemics, characterised by ‘sustained community spread in multiple regions of the world.’[2] The epidemiological definition of a pandemic is ‘an epidemic occurring worldwide or over a very wide area, crossing international boundaries and usually affecting a large number of people.’[3] The definition does not define the severity of the outbreak. In South Africa, the 2009 outbreak coincided with the winter months and thus the usual season for respiratory virus infections. What was unusual about the H1N1 outbreak was its predilection for older children, young adults and pregnant women. The median age of patients all over the world in this pandemic was 10 - 20 years.[4] In their article published in this journal, Ahrens and co-authors present their experience of critically ill children at Red Cross War Memorial Children’s Hospital (RCWMCH) who were admitted with H1N1 infection during the outbreak from 1 August to 30 September 2009 and compare these patients with children affected by other respiratory viruses.[5] During this period, 19 children with H1N1 were admitted to the Paediatric Intensive Care Unit (PICU) out of 20 admissions. The data from this study reveal a number of interesting characteristics. Most of the H1N1-infected children in the study were younger than 3 years of age, with only three patients older than this.[5] This is in contrast to our own experience of H1N1 affecting predominantly older children, as well as the description from the literature.[4,6] In a study of all the paediatric deaths associated with the 2009 pandemic in the USA, the median age at death was 9.4 years, and 72% of the children were >5 years of age at the time of death.[6] Comorbidities were prevalent in both of the

RCWMCH groups and are in accordance with data from the USA, where 68% of the children for whom the information was available had an associated high-risk medical condition.[6] These conditions included neurodevelopmental and seizure disorders, asthma and other lung diseases, and cardiac disease. Four of the five deaths in the RCWMCH study were in children with significant underlying comorbidities.[5] Patients with H1N1 infection had greater morbidity and longer PICU stays than children with other respiratory virus infections.[5] This inevitably has a knock-on effect in limiting turnover of beds and the availability of these beds to other children, particularly those requiring elective surgery. What is of concern is the high prevalence of presumed hospitalacquired H1N1 infection in the RCWMCH study, namely 36.8%. Six out of the seven children with nosocomially acquired H1N1 infection had underlying chronic conditions; the seventh child was referred from another hospital.[5] As the authors point out, the high bed occupancy rate in a very busy tertiary hospital serving the public health sector does increase the risk for hospital-acquired infections, especially during the respiratory virus season. It is incumbent upon us to emphasise the importance of prevention of transmission of infection between patients in our wards.

S Kling General Paediatrics, Department of Paediatrics and Child Health, Faculty of Medicine and Health Sciences, Stellenbosch University, South Africa References 1. http://www.nicd.ac.za/?page=alerts&id=5&rid=553 (accessed 31 May 2015). 2. Fineberg HV. Pandemic preparedness and response – Lessons from the H1N1 Influenza of 2009. N Engl J Med 2014;370;14:1335-1342. [http://dx.doi.org/10.1056/NEJMra120882] 3. Porta M, ed. A Dictionary of Epidemiology, 5th ed. City: Oxford University Press, 2008. http://www.oxfordreference.com/view/10.1093/acref/9780195314496.001.0001/acref9780195314496-e-1373?rskey=oFXArL&result=1372. (Online version 2014) (accessed 1 June 2015). 4. Schoub B. Swine flu – implications for South Africa. Communicable Diseases Surveillance Bulletin 2009;7(3):5-7. 5. Ahrens JO, Morrow BM, Argent AC. Influenza A(H1N1)pdm09 in critically ill children admitted to a paediatric intensive care unit, South Africa. S Afr J Crit Care 2015;31(1):4-7. 6. Cox CM, Blanton L, Dhara R, et al. 2009 Pandemic Influenza A (H1N1) deaths among children – United States, 2009 - 2010. CID 2011;52(Suppl 1):S69-S74. [http://dx.doi.org/10.1093/cid/ ciq011]

S Afr J Crit Care 2015;31(1):3. DOI:10.7196/SAJCC.238

The Critical Care Society of Southern Africa works for the benefit of critically ill patients. Membership is open to all health care professionals involved in the management of the critically ill. Visit the Society’s web page at: www.criticalcare.org.za

SAJCC June 2015, Vol. 31, No. 1

ARTICLE

Influenza A(H1N1)pdm09 in critically ill children admitted to a paediatric intensive care unit, South Africa J O Ahrens,1,2 FCPaeds (SA); B M Morrow,2 PhD; A C Argent,1,2 MD, FCPaeds (SA) 1 2

Paediatric Intensive Care Unit, Red Cross War Memorial Children’s Hospital, Cape Town, South Africa Department of Paediatrics and Child Health, Faculty of Health Sciences, University of Cape Town, South Africa

Corresponding author: B M Morrow (brenda.morrow@uct.ac.za) Objective. To describe the clinical course of critically ill children with confirmed pandemic influenza A(H1N1)pdm09 (H1N1) infection in a southern African paediatric intensive care unit (PICU), and to compare them with a similar group with respiratory virus infections other than H1N1 admitted to the same PICU during the same period. Methods. A retrospective descriptive study of all patients admitted to a PICU in Cape Town, South Africa, who tested positive for H1N1 and other respiratory viruses from 1 August to 30 September 2009. Results. A total of 19 children in 20 PICU admissions tested positive for H1N1 (Group 1). Of these, 14 (70%) had major comorbidities and 4 tested positive for another respiratory virus. Five (26.3%) children in this group died and seven (36.8%) had nosocomial infection. Eight patients in nine PICU admissions who tested H1N1-negative (Group 2), tested positive for other respiratory viruses. Of these, five (55.6%) had major comorbidities. None in this group died. Children in Group 1 had significantly longer ICU stays, ventilator days and worse indices of organ dysfunction than those in Group 2. Conclusions. Children admitted to the PICU with confirmed H1N1 tended to have longer ICU stays, prolonged ventilation, more severe organ dysfunction and higher mortality than those with other respiratory viruses. Hospitalisation was identified as a major risk factor for chronically ill children to acquire H1N1 infection requiring intensive care in our setting. S Afr J Crit Care 2015;31(1):4-7. DOI:10.7196/SAJCC.202

SAJCC June 2015, Vol. 31, No. 1

southern hemisphere[10-13] countries showed that younger children and those with preexisting chronic medical conditions were at greater risk of more severe disease, including shock, multi-organ failure and higher mortality. The initial wave of the 2009 flu pandemic in the northern hemisphere countries took place during their spring and summer months, which coincided with the end of their seasonal influenza season, whereas in the southern hemisphere the epidemic started in the winter months and coincided with the seasonal outbreaks of other respiratory viral infections, including

seasonal influenza. Few data exist as to what extent the clinical picture of H1N1 differed from that of the seasonal respiratory virus infections in children, and what the effect of the 2009 flu pandemic was on paediatric critical care resources in southern Africa. Such data would help in planning optimal use of paediatric critical care resources in a country with limited resources and limited ability to test for respiratory virus infections, leaving clinicians to make decisions about clinical care and resource allocation purely on clinical grounds rather than in conjunction with laboratory testing.

No. of Isolates

30 25

pH1N1 (RCWMCH)

pH1N1 (PICU)

Seasonal influenza A

10 5 0

W ee W k1 ee W k2 e W ek 3 ee W k4 ee W k5 ee W k6 e W ek 7 ee W k8 e W ek ee 9 W k1 e 0 W ek 1 ee 1 W k1 ee 2 W k1 ee 3 W k1 ee 4 W k1 ee 5 W k1 ee 6 W k1 ee 7 W k1 ee 8 W k1 ee 9 W k2 ee 0 W k2 ee 1 W k2 ee 2 W k2 e 3 W ek 2 e 4 W ek 2 ee 5 W k 26 ee W k2 ee 7 k2 8

The first case of ‘swine flu’ (pandemic influenza A(H1N1) pdm09 (H1N1)) in South Africa (SA) was diagnosed on 17 June 2009.[1] Red Cross War Memorial Children’s Hospital (RCWMCH) in Cape Town, SA, was part of a sentinel epidemiological testing site to monitor the pandemic in SA. The first case of H1N1 at RCWMCH was confirmed on 3 August 2009, following which a further 92 children at RCWMCH tested positive for the virus until the end of September 2009. Surveillance data of the 2009 H1N1 epidemic in SA supplied by the SA National Institute of Communicable Diseases[2] show that there were 12 636 laboratory-confirmed cases during the period May - December 2009, with the peak incidence in early August and a rapid decline towards the end of September 2009. The incidence rate of laboratoryconfirmed cases was 25.6 and 39.5 per 100 000 population in SA and the Western Cape Province, respectively. There were 93 recorded deaths associated with the outbreak nation ally (about 0.7%). The weekly incidence of laboratory-confirmed cases at RCWMCH was similar to the national trend (Fig. 1). Early studies of critically ill children with H1N1 infection from northern[3-9] and

Week (April - September)

Fig. 1. The prevalence of seasonal and pandemic H1N1 (pH1N1) influenza A at RCWMCH and the paediatric intensive care unit (PICU) from April to September 2009.

This was the first study of critically ill children with H1N1 infection in southern Africa.

Objectives The objectives of this study were to describe the clinical course of children managed in the paediatric intensive care unit (PICU) at RCWMCH with laboratory-confirmed H1N1 infection, and to compare the clinical course of these children with a cohort of other respiratory viral isolates admitted to the PICU during the same period.

Methods This was a retrospective descriptive study of all children admitted to the PICU at RCWMCH from 1 August to 30 September 2009, who tested positive for the following respiratory viruses: H1N1, seasonal influenza A, influenza B, parainfluenza 1/2/3, rhinovirus, respiratory syncytial virus (RSV), human metapneumovirus (HMPV) and adenovirus. Patients were tested for the above respiratory viruses if they met the clinical case definition for a moderate or severe acute respiratory infection (SARI),[14] had household contact with suspected H1N1 infection, or at the discretion of the PICU consultant. The children were divided into two groups, Group 1 being those who tested positive for the H1N1 virus, and Group 2 those who tested positive for respiratory viruses other than H1N1. Non-directed bronchial alveolar lavage specimens, tracheal aspirates or nasopharyngeal aspirates obtained as part of standard practice were submitted to the virology laboratory of the National Health Laboratory Services at Groote Schuur Hospital, Cape Town, for detection of respiratory viruses. Initially samples were tested using the Multiplex 7 virus polymerase chain reaction (PCR) (Seeplex RV detection kit, Seegene, Seoul, Korea), with influenza A-positive samples typed with supplementary primers included in the kit. In mid-August 2009, testing was changed to H1N1-specific screening using an Advanced Realtime PCR kit (Luminex Molecular Diagnostics, Inc., Toronto, Canada) for Swine H1N1 influenza. Negatives were tested by respiratory viral PCR (RV-PCR) the following day. In mid-September 2009, the real-time screening was stopped and testing was resumed with RV-PCR. Every patient admitted to the PICU with a SARI or proven H1N1 infection during the study period was given a 5-day course of the neuraminidase inhibitor, oseltamivir, which was extended in some patients who had prolonged illness and who continued to test positive for the H1N1 virus. The following data were collected and recorded on an Excel 2007 (Microsoft Corporation, USA) spreadsheet: patient demographics (age, gender, address), clinical observations (temperature, weight, nutritional status, daily fluid balance, primary diagnoses and comorbidities), treatment (including antimicrobial treatment, oseltamivir, corticosteroids, diuretics, mode of respiratory support and use of inotropes), special investigations (arterial blood gases, blood electrolytes, full blood count and liver function tests, tests for respiratory viruses and bacterial co-infections and inflammatory markers), measures of oxygenation (oxygenation index, PaO2:FiO2 ratio on admission), Pediatric Index of Mortality Score (PIM2) on admission, daily Pediatric Logistic Organ Dysfunction (PELOD) score, duration of PICU stay and mechanical ventilation, and PICU mortality. Extracorporeal membrane oxygenation was not available. Full approval for the study was obtained from the Human Research Ethics Committee of the Faculty of Health Sciences, University of Cape Town. The study adhered to the provisions laid down in the Declaration of Helsinki (2013).[15]

Analysis Data were tested for normality using Kolmogorov Smirnov and Lilliefors tests. Data were not normally distributed and were there fore presented as median (interquartile range (IQR)) or proportions for categorical variables. Chi-square tests (for categorical data), with Yates correction for cells values <10, and the Mann-Whitney U-test (for continuous data) were used for comparisons between Groups 1 and 2, and between survivors and non-survivors in Group 1. The Kruskal-Wallis analysis of variance (ANOVA) by ranks was used to assess differences among multiple variables. Statistica version 10 (2011) (StatSoft Inc., USA) was used for data analysis. A significance level of p<0.05 was chosen.

Results Nineteen children in 20 PICU admissions tested positive for H1N1 during the study period (Group 1), comprising 20.2% of the total H1N1-positive children presenting to RCWMCH (both in- and outpatients). Three patients in Group 1 were co-infected with rhinovirus, and one was co-infected with adenovirus. Eight children in nine PICU admissions tested positive for a respiratory virus other than H1N1 (Group 2), namely: rhinovirus (n=4, 50%); rhinovirus plus influenza B (n=1), rhinovirus plus seasonal influenza A (n=1, 12.5%), RSV plus seasonal influenza A (n=1, 12.5%) and adenovirus alone (n=1, 12.5%). Table 1 summarises the clinical features of patients in Group 1 and Group 2, and also compares survivors and non-survivors in Group 1. The median age of H1N1-infected children was 12 months, with only three patients older than three years. Both sexes were equally prone to severe H1N1 infection. Despite seven of the eight patients in Group 2 being male, the difference in gender distribution between groups did not reach statistical significance (Table 1). In Group 1, the underlying reasons precipitating PICU admission were respiratory failure from pneumonia (n=14), elective cardiac surgery (n=3), laryngotracheobronchitis (n=1), severe burns (n=1) and trauma from a motor vehicle accident (n=1). The reasons for PICU admission of the children in Group 2 were pneumonia (n=3), laryngotracheobronchitis (n=3), Guillain-Barre syndrome (n=1), septic shock in a child who had a previous liver transplant (n=1) and seizures and obstructive hydrocephalus from tuberculous meningitis (n=1). Both groups of patients had a high incidence of comorbidities. The rate of HIV and bacterial co-infections was not significantly different between the two groups (Table 1). Seven children (36.8%) in Group 1 had a presumed nosocomially acquired respiratory virus infection, defined as having first evidence (clinically or by laboratory confirmation) of a respiratory virus infection at least 72 hours after hospital admission, whereas none of the patients in Group 2 acquired infection nosocomially (p<0.05) (Table 1). Of the seven who likely acquired the H1N1 infection nosocomially, one was transferred from a regional hospital for PICU care. Three others had been admitted to the cardiac ward awaiting cardiac surgery, where pre- and postoperative cardiac surgery patients co-habited. They had no preoperative symptoms or signs of a respiratory virus infection, had not received prophylactic oseltamivir, underwent elective cardiac surgery and had a postoperative course in the PICU complicated by pneumonia, with prolonged periods of ventilation and PICU stay. All three had bacterial co-infections. Three other children with chronic medical conditions (end-stage renal failure, chronic lung disease, epilepsy) had been inpatients in general paediatric wards for long-term care where no isolation facilities were available to protect patients at risk from nosocomial infections.

SAJCC June 2015, Vol. 31, No. 1

Table 1. Patient characteristics* H1N1-positive (Group 1)

H1N1-negative (Group 2)

Total (n=19; 20 admissions)

ICU survivors (n=14; 15 admissions)

ICU nonsurvivors (n=5)

p-value†

Total (n=8; 9 admissions)

p-value‡

Age (months)

12 (1 - 100)

12 (1 - 90)

18 (3 - 100)

0.5

12 (7 - 25)

0.6

Gender (n), M:F

9:10

6:8

3:2

0.9

7:1

0.1

Hospital-acquired infections

7 (35)

5 (33)

2 (40)

0.4

0.047

Pre-ICU oseltamivir

4 (20)

4 (28.5)

0 (0)

0.5

0.7

Mortality

5 (25)

0.3

0.26

PIM2

0.059 (0.0005 - 0.714)

0.065 (0.002 - 0.593)

0.7

0.055 (0.0004 - 0.59)

0.8

Admission PELOD score

11 (0 - 22)

11 (1 - 22)

0.6

1 (0 - 11)

0.02

PICU stay (days)

8 (1 - 27)

7 (2 - 16)

17 (1 - 27)

0.1

6 (0 - 11)

0.4

Invasive ventilation (days)

17 (85)

12 (80)

5 (100)

0.7

8 (89)

0.8

Ventilator days

5 (0 - 27)

4 (0 - 11)

17 (1 - 27)

0.05

1 (0 - 11)

0.9

Day 3 PaO2:FiO2

170.16 (67.4- 352.3)

175 (69.3 - 352.3)

101.64 (67.4 - 178.7)

0.07

306.7 (170.6 - 493.3)

0.002

Day 3 OI

6.07 (0.44 - 16.97)

5.07 (0.44 - 16.97)

16.30 (1.01 - 22.41)

0.07

1.98 (0.30 - 7.08)

0.008

Significant comorbidities

16 (80)

12 (85)

4 (80)

0.7

6 (75)

0.3

HIV infection

3 (5.2)

2 (14.3)

1 (20)

0.7

1 (12.5)

0.8

Inotropes

7 (35)

5 (33.3)

2 (40)

0.8

1 (11.1)

0.4

Lowest white cell count during ICU stay (×10 /L)

5.3 (1.1 - 9.7)

6.4 (1.2 - 9.7)

4.9 (1.1 - 9.4)

0.3

6.2 (2.0 - 8.3)

0.8

Lactate on admission (mmol/L)

1.1 (0.6 - 16.0)

1.0 (0.7 - 9.1)*

1.2 (0.6 - 16.0)

0.8

0.8 (0.5 - 2.2)

0.02

C-reactive protein <48 hours (mg/L)

64.5 (6.7 -379.0)

49.0 (6.7 - 148.0)

120.0 (9.4 - 379.0)

0.2

15.0 (6.8 - 68.0)

0.1

Bacterial co-infection

9 (45)

6 (40)

3 (60)

0.8

5 (55)

0.9

* Continuous data are presented as median (range); categorical data as n (%). †

Comparing H1N1-infected survivors v. non-survivors.

‡

Comparing H1N1-negative patients with all H1N1-positive patients.

Table 2. Summary of characteristics of the patients with H1N1 influenza A who died (n=5) Patient

Age (months)

Nutritional status (weight-for-age)

Comorbidities

Bacterial sepsis

Type of ventilation

PICU stay (days)

Immediate cause of death

Normal

Nil

IPPV

Tension pneumothorax

UWFA

Obstructive sleep apnoea, cor pulmonale

HFOV

ARDS, pulmonary hypertension

HFOV

Bullous lung disease, pneumothorax

Nil

Normal

Disseminated tuberculosis, Previous RSV / rhinovirus A infection Ebstein-Barr Virus infection

UWFA

HIV, preterm

Enterococcus, Acinetobacter baumannii

HFOV

Nosocomial bacterial sepsis with ARDS

100

Normal

70% total body surface area burn

Haemophilus influenzae, Acinetobacter baumannii, Pseudomonas, Ralstonia

IPPV

Nosocomial bacterial sepsis with shock

Nil

UWFA = underweight for age; IPPV = intermittent positive pressure ventilation; HFOV = high-frequency oscillation ventilation; ARDS = acute respiratory distress syndrome.

Indices of illness severity and multi-organ involvement were higher in Group 1 than in Group 2 on admission, with median admission PELOD score and serum lactate 11 v. 1 (p=0.02) and 1.1 v. 0.8 (p=0.02), respectively (Table 1). A similar percentage of patients in both groups required invasive ventilation and, although patients in Group 1 had longer durations of ventilation than patients in Group 2 (5 v. 1 median

SAJCC June 2015, Vol. 31, No. 1

ventilator days), this was not statistically significant (Table 1). Of clinical importance was that patients in Group 1 had worse indices of oxygenation on day 3 of ventilation than those in Group 2, as suggested by higher median oxygenation index (OI) (6.07 v. 1.98; p=0.008) and lower median PaO2:FIO2 ratio (170.16 v. 306.67; p=0.002) (Table 1). The predicted PICU mortality according to the median PIM2

scores was similar for both groups (Table 1). Five (26.3%) children infected with H1N1 died: three as a result of respiratory failure, and two from overwhelming sepsis (Table 2). Four of the patients who died had major comorbidities (Table 2). No deaths occurred in Group 2 (p=0.3). There were no significant differences between survivors and non-survivors in Group 1 (Table 1).

Discussion This is the first study from SA to describe children severely affected by the 2009 H1N1 pandemic and who required PICU admission. The RCWMCH PICU admission rate (22%) for H1N1-positive children was comparable with the 9.3 - 26.0% reported from Canada[7] and the USA[16] during the outbreak. Although small patient numbers precluded identification of risk factors for H1N1 infection or mortality, it is notable that 80% of critically ill children with H1N1 had major comorbid conditions, including HIV infection (Table 1). Other centres around the world have also reported that children with comorbidities such as asthma and neurological or developmental conditions were at particular risk of severe H1N1 disease, and at greater risk of dying from H1N1 infection.[16,17] Bacterial co-infection has been reported as a major risk factor for mortality in studies from nearly every major influenza pandemic, including the 2009 pandemic.[5,11,16] A high proportion of our patients, particularly in the H1N1-positive group, had documented or suspected bacterial co-infection (Table 1). The H1N1 outbreak had an immediate effect on the availability of PICU beds for elective surgery during August 2009. In accordance with the hospital’s escalation plan, all elective surgery cases requiring postoperative PICU admission were cancelled for a 2-week period during the peak of the pandemic in the Western Cape, during which time period the PICU ran at 100% occupancy. Fortunately, the number of H1N1-positive patients declined rapidly during September 2009 and the elective surgery list could return to normal. Some other centres have also commented that the pandemic placed strain on critical care resources.[18,19] Patients with H1N1 infection admitted to the PICU had more severe organ dysfunction and disease severity compared with the H1N1negative group, as evidenced by higher admission PELOD scores and poorer oxygenation on day 3. Although clinical outcome measures were not found to be statistically different, children in the H1N1 group had clinically significantly longer durations of PICU stay and mechanical ventilation, as well as higher mortality. Observations from other countries have been varied, with some countries reporting higher mortality rates and more severe illness among critically ill children with p(H1N1) disease,[5,18-20] whereas some countries have reported lower disease acuity and mortality rates than expected, particularly if compared with the previous seasonal influenza seasons.[21,22] The mortality rate of children with H1N1 was 26.3%, whereas mortality rates in other PICUs have been between 0% (Netherlands)[23] and 39% (Argentina).[18] Of the children who died, four had major underlying comorbidities and none tested positive for any other respiratory virus. The high rate of presumed nosocomially acquired H1N1 infection is concerning, but not unexpected in the context of lack of isolation and cohorting facilities, and the fact that an effective vaccine was not yet available in our hospital at the outbreak of the epidemic. The neuraminidase inhibitor oseltamivir also only became available for prophylaxis for exposed H1N1 contacts half-way through the epidemic. Once available, four of the children admitted to PICU had received oseltamivir prior to PICU admission, and all four survived to discharge.

Conclusion Children admitted to an SA PICU with H1N1 infection had evidence of more severe inflammatory disease, greater organ dysfunction and worse respiratory indices, and had poorer outcome in terms of PICU stay and mortality rate compared with children admitted to the PICU with other respiratory viruses. Comorbities and co-infection with bacterial pathogens were common. These findings are similar to those of other studies of critically ill children in other centres in the world, particularly from the southern hemisphere, relating to the 2009 swine flu pandemic. This study also highlights the problem of hospital-acquired pandemic influenza infections in a setting of a tertiary hospital with a high bed occupancy rate, emphasising the potential for acquisition of these viruses by hospitalised children with significant comorbidities. Attention must be paid to implementing appropriate infection control procedures in order to prevent crossinfection in case of future pandemics. References 1. Schoub B. Swine flu - implications for South Africa. Commun Dis Surveill Bull 2009;7(3):5-7. 2. National Institute for Communicable Diseases of the National Health Laboratory Services. Situation Update Pandemic Influenza A(H1N1) 2009, South Africa. Report no. SWIN110809 SITREP. Johannesburg: NICD, 2009. http://www.nicd.ac.za (accessed 15 August 2015). 3. Hackett S, Hill L, Patel J, et al. Clinical characteristics of paediatric H1N1 admissions in Birmingham, UK. Lancet 2009;374(9690):61511-61517. [http://dx.doi.org/10.1016/S01406736(09)61511-7] 4. Lister P, Reynolds F, Parslow R, et al. Swine-origin influenza virus H1N1, seasonal influenza virus, and critical illness in children. Lancet 2009;374(9690):605-607. [http://dx.doi.org/10.1016/ S0140-6736(09)61512-9] 5. Lockman JL, Fischer WA, Perl TM, Valsamakis A, Nichols DG. The critically ill child with novel H1N1 influenza A: A case series. Pediatr Crit Care Med 2010;11(2):173-178. [http://dx.doi. org/10.1097/PCC.0b013e3181ccedae] 6. Koliou M, Soteriades ES, Toumasi MM, Demosthenous A, Hadjidemetriou A. Epidemiological and clinical characteristics of influenza A(H1N1)v infection in children: The first 45 cases in Cyprus, June - August 2009. Euro Surveill 2009;14(33):19312. 7. Jouvet P, Hutchison J, Pinto R, et al. Critical illness in children with influenza A/pH1N1 2009 infection in Canada. Pediatr Crit Care Med 2010;11(5):603-609. [http://dx.doi.org/10.1097/ PCC.0b013e3181d9c80b] 8. Jain S, Kamimoto L, Bramley AM, et al. Hospitalized patients with 2009 H1N1 influenza in the United States, April - June 2009. N Engl J Med 2009;361(20):1935-1944. [http://dx.doi. org/10.1056/NEJMoa0906695] 9. Gilsdorf A, Poggensee G, Working Group Pandemic Influenza A(H1N1)v Influenza A(H1N1)v in Germany: The first 10 000 cases. Euro Surveill 2009;14(34):19318. 10. Gomez J, Munayco C, Arrasco J, et al. Pandemic influenza in a southern hemisphere setting: The experience in Peru from May to September, 2009. Euro Surveill 2009;14(42):19371. 11. ANZIC Influenza Investigators, Webb SA, Pettila V, et al. Critical care services and 2009 H1N1 influenza in Australia and New Zealand. N Engl J Med 2009;361(20):1925-1934. [http://dx.doi. org/10.1056/NEJMoa0908481] 12. Oliveira W, Carmo E, Penna G, et al. Pandemic H1N1 influenza in Brazil: Analysis of the first 34 506 notified cases of influenza-like illness with severe acute respiratory infection (SARI). Euro Surveill 2009;14(42):19362. 13. Archer BN, Timothy GA, Cohen C, et al. Introduction of 2009 pandemic influenza A virus subtype H1N1 into South Africa: Clinical presentation, epidemiology, and transmissibility of the first 100 cases. J Infect Dis 2012;206 (Suppl 1):S148-S153. [http://dx.doi.org/10.1093/ infdis/jis583] 14. The National Institute for Communicable Diseases (NICD) of the National Health Laboratory Service (NHLS). Revised Health Workers Handbook on Pandemic Influenza A (H1N1) ‘swine flu’, 2009. http://www.kznhealth.gov.za/h1n1handbook.pdf (accessed 15 August 2015). 15. World Medical Association (WMA). WMA Declaration of Helsinki - Ethical Principles for Medical Research Involving Human Subjects, 2013. http://www.wma.net/en/30publications/10policies/ b3/17c.pdf (accessed 15 August 2015). 16. Randolph AG, Vaughn F, Sullivan R, et al. Critically ill children during the 2009 - 2010 influenza pandemic in the United States. Pediatrics 2011;128(6):e1450-e1458. [http://dx.doi. org/10.1542/peds.2011-0774] 17. Libster R, Bugna J, Coviello S, et al. Pediatric hospitalizations associated with 2009 pandemic influenza A (H1N1) in Argentina. N Engl J Med 2010;362(1):45-55. [http://dx.doi.org/10.1056/ NEJMoa0907673] 18. Farias JA, Fernandez A, Monteverde E, et al. Critically ill infants and children with influenza A (H1N1) in pediatric intensive care units in Argentina. Intensive Care Med 2010;36(6):10151022. [http://dx.doi.org/10.1007/s00134-010-1853-1] 19. Torres SF, Iolster T, Schnitzler EJ, et al. High mortality in patients with influenza A pH1N1 2009 admitted to a pediatric intensive care unit: A predictive model of mortality. Pediatr Crit Care Med 2012;13(2):e78-83. [http://dx.doi.org/10.1097/PCC.0b013e318219266b] 20. Kendirli T, Demirkol D, Yildizdas D, et al. Critically ill children with pandemic influenza (H1N1) in pediatric intensive care units in Turkey. Pediatr Crit Care Med 2012;13(1):e11-e17. [http:// dx.doi.org/10.1097/PCC.0b013e31820aba37] 21. Morgan CI, Hobson MJ, Seger B, Rice MA, Staat MA, Wheeler DS. 2009 pandemic influenza A (H1N1) in critically ill children in Cincinnati, Ohio. Pediatr Crit Care Med 2012;13(3):e140-e144. [http://dx.doi.org/10.1097/PCC.0b013e318228845f ] 22. Baird JS, Buet A, Hymes SR, et al. Comparing the clinical severity of the first versus second wave of 2009 influenza A (H1N1) in a New York city pediatric healthcare facility. Pediatr Crit Care Med 2012;13(4):375-380. [http://dx.doi.org/10.1097/PCC.0b013e31823893df ] 23. Augustyn B. Ventilator-associated pneumonia: Risk factors and prevention. Crit Care Nurse 2007;27(4):38-39.

SAJCC June 2015, Vol. 31, No. 1

ARTICLE

Obstetric intensive care admissions at a tertiary hospital in Limpopo Province, South Africa T S Ntuli,1 BSc, Bsc (Hon), MSc; G Ogunbanjo,2 MB ChB, FCCS (SA), FACRRM, FAFP (SA); S Nesengani,3 MB ChB, MMed (O&G); E Maboya,4 MB ChB, MMed (Anaes); M Gibango,4 MB ChB, MMed (Anaes) Research Development and Administration, University of Limpopo (Turfloop Campus), Sovenga, South Africa Department of Family Medicine, Sefako Makgatho Health Sciences, University of Pretoria, South Africa 3 Department of Obstetrics and Gynaecology, University of Limpopo (Turfloop Campus), Polokwane, South Africa 4 Department of Anaesthesiology, University of Limpopo (Turfloop Campus), Polokwane, South Africa 1 2

Corresponding author: T S Ntuli (tsntuli@hotmail.com)

Objective. To determine the characteristics of obstetric patients admitted to the intensive care unit (ICU) at a tertiary hospital in the Limpopo Province, South Africa. Methods. Hospital files of all obstetric patients admitted to the Pietersburg provincial referral hospital ICU from 1 January 2008 to 31 December 2012 were retrospectively reviewed. Age, parity, admission diagnosis, length of stay, information on the referring hospitals, and maternal outcomes were analysed. Results. There were 138 obstetric ICU admissions during the study period (6.7% of all ICU admissions and 0.95% of all deliveries). The most common reasons for obstetric ICU admissions were pre-eclampsia or eclampsia (52.9%, n=73/138) and obstetric haemorrhage (18.1%, n=25/138). The mean age of the patients was 28 years, and mean duration of ICU stay was 8 days (range 0 - 163 days). Forty-eight maternal deaths occurred (34.8%), and of these, 27 were referrals from other hospitals (district and regional hospitals). Pre-eclampsia or eclampsia accounted for 25 (52%) of all deaths. Conclusion. Obstetric patients formed a small proportion of ICU admissions, but mortality among these patients was high. It is recommended that obstetric registrars rotate through a multidisciplinary ICU, and the need for a critical care specialist should be considered. S Afr J Crit Care 2015;31(1):8-10. DOI:10.7196/SAJCC.164

Admission of critically ill obstetric patients to the intensive care unit (ICU) is a common phenomenon in both developed and developing countries. A number of studies have reported that <1% of ICU admissions are obstetric patients, [1-4] while other studies indicate that ICU admission rates of 1 - 10% are obstetric patients. [5-7] However, there are few studies in which obstetric patients comprise >10% of ICU admissions.[8,9] The most common reasons for obstetric admission are pre-eclampsia or eclampsia, and obstetric haemorrhage.[1,3,8,10-12] A few studies report that single and multiple organ failure may also be as a possible reason for obstetric admission.[8,13] In the last 10 years, maternal mortality in the ICU is increasingly rare in developed countries, and some studies have reported no maternal deaths at all.[2,11,14] Other studies have reported maternal mortality of between 1 and 5%,[2,12] while others have reported a maternal mortality of >10%.[8,10,15-17] A retrospective study of obstetric patients admitted to the former Johannesburg General Hospital ICU reported a maternal death rate of 38%.[18] The ICU of the Obstetrics and Gynaecology Department in King Edward VIII Hospital, Durban, has reported a maternal mortality of 21%.[5] The studies reviewed illustrate that there is a wide variation in the mortality rates of obstetric patients admitted to the ICU. Tertiary referral centres usually admit more high-risk obstetric patients.[5,18] In order to assist local healthcare teams in knowing which health conditions to focus on, there is a need to identify the indication for obstetric ICU admission. Therefore, a retrospective descriptive study was undertaken to determine the characteristics of obstetric patients admitted to a tertiary hospital in the Limpopo Province, South Africa (SA).

SAJCC June 2015, Vol. 31, No. 1

Methods A retrospective review of obstetric admissions to the ICU of Pietersburg Hospital, Limpopo, SA, was carried out over a period of 5 years (1 January 2008 to 31 December 2012). The 12-bed multi disciplinary ICU admitted 350 - 450 patients on average per annum. ICU admission in this institution is based on the clinical judgment of the admitting discipline in consultation with an anaesthetist. The ICU records and patient files were extracted and reviewed. Ethics approval to conduct the study was obtained from the University of Limpopo ethics committee, and anonymity and confidentiality of patient personal information were protected. The data for the study were collected by a trained nurse assistant. The data collected included patient age, parity, admission diagnosis, length of stay in the ICU and maternal outcome. Categorical data were displayed as percentages; continuous data were reported as mean (standard deviation (SD)). Statistical software (STATA 9.0, StataCorp, USA) was used for data analysis.

Results Over the 5-year period, 138 obstetric patients were admitted (6.7% of 2 073 ICU admissions and 0.95% of 14 478 deliveries at the hospital). Six patients were admitted twice, and two of these patients died. The characteristics of the admitted obstetric patients are presented in Table 1. The average age of the patients was 28 (8.1) years (range 16 - 45 years). More than half (n=79/138) of the patients had a parity of 2 or more. Sixtytwo patients (45%) were referred, and the majority of the referred patients were from district hospitals (n=53/62 (86%)). Table 2 shows the yearly obstetric deliveries and maternal ICU admissions from 2008 to 2012. The indications for admission and outcome of obstetric ICU admissions are shown in Table 3. Pregnancy-induced hypertension (52.9%, n=73/138) and severe obstetric haemorrhage (18.1%,

n=25/138) were the most frequent causes of admission. Eleven of the patients with pre-eclampsia or eclampsia presented

Table 1. Characteristics of obstetric patients admitted to the ICU n (%) Age (years) <20

22 (16)

20 - 24

36 (26)

25 - 29

33 (24)

30 - 34

23 (17)

35 - 39

16 (11)

â&#x2030;Ľ40

9 (6)

with HELLP (haemolysis, elevated liver enzymes and low platelet count) syndrome. Anaesthetic complications were seen in 5.8% (n=8/138) of the patients. The average duration of obstetric ICU admission was 8 days (range 0 - 163 days). There were 48 obstetric deaths documented during the study period (mortality rate 34.8%). This included 34% (n=25/73) of the women admitted with pre-eclampsia or eclampsia and 36% (n=9/25) of patients with severe obstetric haemorrhage. Half (n=24/48) of the deaths were referrals from district hospitals.

Discussion

Parity (n) 0

39 (29)

20 (14)

53 (38)

â&#x2030;Ľ3

26 (19)

Table 2. Frequency of obstetric ICU admissions Year

Obstetric deliveries, n

Obstetric ICU admissions, n (%)

2008

3 007

24 (0.80)

2009

3 037

27 (0.89)

2010

2 770

18 (0.65)

2011

2 758

45 (1.63)

2012

2 906

24 (0.83)

In our study, 6.7% of the ICU admissions were owing to obstetric problems. This finding is similar to those in studies with results that ranged between 1 and 10%.[5-7] However, in some other studies obstetric ICU admission rates were reported to be <1%.[1-4] The latter were exclusively conducted in developed countries, where obstetric care at peripheral referring hospitals is much improved compared with that in developing countries. Previous studies have reported that pre-eclampsia or eclampsia and obstetric haemorrhage were the most common indications for ICU admission.[1,3,8,10-12] The findings of our study confirmed that these indications are the most frequent, although anaesthetic complications also

Table 3. Indications and outcomes of obstetric ICU admission (N=138) Diagnosis

Total, n (%)

Survivors, n (%)

Non-survivors, n (%)

Pre-eclampsia or eclampsia

73 (52.9)

48 (66)

25 (34)

Obstetric haemorrhage

25 (18.1

16 (64)

9 (36)

Anaesthetic complication (spinal)

8 (5.8)

5 (62)

3 (38)

Ruptured ectopic pregnancy

4 (2.9)

3 (75)

1 (25)

Pulmonary oedema

4 (2.9)

3 (75)

1 (25)

Renal failure

4 (2.9)

2 (50)

Ruptured uterus

4 (2.9)

3 (75)

1 (25)

Placenta abruptio

3 (2.2)

2 (67)

1 (33)

Peripartum cardiomyopathy

2 (1.4)

1 (50)

Respiratory distress

2 (1.4)

2 (100)

Postabortional sepsis

2 (1.4)

2 (100)

Abdominal pregnancy

1 (0.7)

1 (100)

Bowel obstruction

1 (0.7)

1 (100)

Hepatic encephalopathy

1 (0.7)

1 (100)

Hypokalaemia

1 (0.7)

1 (100)

Parasuicide

1 (0.7)

1 (100)

Tuberculosis

1 (0.7)

1 (100)

Septicaemia during labour

1 (0.7)

1 (100)

featured commonly. Other studies have shown that acute pulmonary oedema is also a frequent cause of admission to an ICU,[2] and a leading cause of death in women with pre-eclampsia.[19,20] In our study, four patients were admitted owing to pulmonary oedema, and one succumbed to pre-eclampsia and multiple organ failure. Most of these conditions are preventable causes of maternal mortality and may indicate relative inexperience with regard to the early identification and management of obstetric emergencies. Interestingly, sepsis is one of the most common non-obstetric causes of admission into ICUs, accounting for 8 - 30%;[1,12,21,22,23] however, in our study, only 4 (2.8%) of the obstetric ICU admissions were owing to infectious causes. It is likely that most of these patients died without admission to ICU.[24] In the present study, the average length of ICU stay was longer than that reported in previous studies.[1,3,10-12] The reasons for this remain to be established; however, it is possible this was owing to admission of more severe obstetric cases from the referring hospitals. The management of obstetric patients in ICU is complex owing to the physiological changes due to pregnancy and poorly understood pathophysiology of pregnancy-related diseases such as preeclampsia;[25,26] this requires collaboration between critical care specialists and obstetricians. [21,27,28] In our study, 34.8% of patients admitted to the obstetric ICU died. This finding is similar to the Johannesburg Hospital study, which reported 38%,[18] but higher than the mortality rate reported in the Durban studies, i.e. 21%.[5,22] Possible reasons for the high maternal mortality rate in our study are a lack of proper antenatal care, late referrals, poor transport facilities, limited specialist obstetrician and critical care specialist support, long distances to the referral hospital and inadequate emergency obstetric care at referral centres close to patient residences. To reduce this high maternal mortality rate, it would be necessary for the provincial health department to implement audit processes to identify areas for improvement in obstetric care at the district and regional referring hospitals within the province.

Study limitations The duration of this study was only 5 years. A longer duration could have resulted in either a higher or lower maternal mortality rate from the obstetric ICU admissions at this hospital. As with retrospective studies, any missing data from patient files affect the reliability of the data, but this was minimised by reviewing all files from the records department and the ICU over the SAJCC June 2015, Vol. 31, No. 1

study period. Finally, in this 12-bed, multidisciplinary ICU, the severity of illness is assessed using the SOFA (Sequential Organ Failure Assessment) score; however, the scores were incomplete for obstetric ICU admissions.

Conclusion In this 12-bed, multidisciplinary ICU, we found that obstetric patients form a small proportion of ICU admissions but the mortality is high. It is recommended that obstetric registrars rotate through a multidisciplinary ICU, and the need for a critical care specialist should be considered. Acknowledgements. We thank the staff of the records department and ICU of the Pietersburg Hospital, especially Dr M E Gonzalez for her co-operation during this study. References 1. Richa F, Karim N, Yazbeck P. Obstetric admissions to the intensive care unit: An eight year review. J Med Liban 2008;56(4):215-219. 2. Sriram S, Robertson MS. Critically ill obstetric patients in Australia: A retrospective audit of 8 years experience in a tertiary intensive care unit. Crit Care Resusc 2008;10(2):124. 3. Keizer JL, Zwart JJ, Meerman RH, Harinck BL, Feuth HD, Van Roosmalen J. Obstetric intensive care admissions: A 12-year review in a tertiary care centre. Eur J Obstet Gynecol Reprod Biol 2006;128(1-2):152-156. [http://dx.doi.org/10.1016/j.ejogrb.2005.12.013] 4. Ramachandra Bhat PB, Navada MH, Rao SV, Nagarathna G. Evaluation of obstetric admissions to intensive care unit of a tertiary referral center in coastal India. Indian J Crit Care Med 2013;17(1):34-37. [http://dx.doi.org/10.4103/0972-5229.112156] 5. Platteau P, Engelhadt T, Moodley J, Muckart DJ. Obstetric and gynaecological patients in an intensive care unit: A 1 year review. Trop Doctor 1997;27(4):2002-2006. 6. Cohen J, Singer P, Kogan A, Hod M, Bar J. Course and outcome of obstetric patients in a general intensive care unit. Acta Obstet Gynecol Scand 2000;79(10):846-850. 7. Okafor UV, Aniebue U. Admission pattern and outcome in critical care obstetric patients. Int J Obstet Anesth 2004;13(3):164-166. [http://dx.doi.org/10.1016/j.ijoa.2004.04.002] 8. Vasquez DN, Estenssoro E, Canales HS, et al. Clinical characteristics and outcomes of obstetric patients requiring ICU admission. Chest 2007;131(3):718-724. [http://dx.doi.org/10.1378/chest.06-2388] 9. Mjahed K, Hamoudi D, Salmi S, Barrou L. Obstetric patients in a surgical intensive care unit: Prognostic factors and outcome. J Obstet Gynaecol 2006;26(5):418-423. [http://dx.doi. org/10.1080/01443610600720188] 10. Baloch R, Jakhrani N, Zeb E, Hafeez F, Abassi M, Naz Abbasi F. Pattern and outcome of obstetric admissions to the surgical intensive care unit – A ten years study. J Surg Pakistan 2010;15(4)171-176.

11. Crozier TM, Wallace EM. Obstetric admissions to an integrated general intensive care unit in a quaternary maternity facility. Aust N Z J Obstet Gynaecol 2011; 51(3):233-238. [http://dx.doi. org/10.1111/j.1479-828X.2011.01303.x] 12. Leung NY, Lau AC, Chan KK, Yan WW. Clinical characteristics and outcomes of obstetric patients admitted to the intensive care unit: A 10-year retrospective review. Hong Kong Med J 2010;16(1):18-25. 13. Muench MV, Baschat AA, Malinow AM, Mighty HE. Analysis of disease in the obstetric intensive care unit at a university referral center: A 24-months review of prospective data. J Reprod Med 2008;53(12):914-920. 14. Madan I, Jain NJ, Grotegut C, Nelson D, Dandolu V. Characteristics of obstetric intensive care unit admissions in New Jersey. J Matern Fetal Neonatal Med 2009;22(9):785-790. [http://dx.doi. org/ 10.3109/14767050902874097] 15. Al Suleiman SA, Outub HO, Rahman J, Rahman MS. Obstetric admissions to the intensive care unit: A 12 year review. Arch Gynecol Obstet 2006;274(1):4-8. [http://dx.doi.org/10.1007/s00404-004-0721-z] 16. Ghike S, Asegaonkar P. Why obstetric patients are admitted to intensive care unit? A retrospective study. J South Asian Feder Obstr Gynae 2012;4(2):90-92. 17. Bhadabe R, De’ Souza R, More A, Harde M. Maternal outcomes in critically ill obstetric patients: A unique challenge. Indian J Crit Care Med 2012;16(1):8-16. [http://dx.doi.org/10.4103/0972-5229.94416] 18. Taylor R, Richards GA. Critically ill obstetric and gynaecological patients in the intensive care unit. S Afr Med J 2000;90(11):1140-1144. 19. Duley L, Williams J, Henderson-Smart DJ. Plasma volume expansion for treatment of women with pre-eclampsia. Cochrane Database Syst Rev 2000;(2):CD001805. [http://dx.doi. org/10.1002/14651858.CD001805] 20. Ganzevoort W, Rep A, Bonsel GJ, et al. A randomised controlled trial comparing two temporising management strategies, one with and one without plasma volume expansion, for severe and early onset pre-eclampsia. BJOG 2005;112(10):1358-1368. [http://dx.doi. org/10.1111/j.1471-0528.2005.00687.x] 21. Zwart JJ, Dupuis JR, Richters A, Ory F, van Roosmalen J. Obstetric intensive care unit admission: A 2-year nationwide population-based cohort study. Intensive Care Med 2010;36(2):256-263. [http://dx.doi.org/10.1007/s00134-009-1707-x] 22. Ngene NC, Moodley J, Songca P. Maternal and fetal outcomes of HIV-infected and noninfected pregnant women admitted to two intensive care units in Pietermaritzburg, South Africa. S Afr Med J 2013;103(8):543-548. [http://dx.doi.org/10.7196/samj.6590] 23. Chawla S, Nakra M, Mohan S, Nambiar BC, Agarwal R, Marwaha A. Why do obstetric patients go to the ICU? A 3-year-study. Med J Armed Forces India 2013;69(2):134-137. [http://dx.doi. org/10.1016/j.mjafi.2012.08.033] 24. NCCEMD. Saving Mothers 2008-2010: Fifth Report on the Confidential Enquiries into Maternal Deaths in South Africa – Short Report. http://www.doh.gov.za (accessed 1 October 2014). 25. Neligan PJ, Laffey JG. Clinical review: Special populations – critical illness and pregnancy. Crit Care 2011;15(4):227. [http://dx.doi.org/10.1186/cc10256] 26. Munnur U, Bandi V, Guntupalli KK. Management principles of the critically ill obstetric patient. Clin Chest Med 2011;32(1):53-60,viii. [http://dx.doi.org/10.1016/j.ccm.2010.10.003] 27. Cartin-Ceba R, Gajic O, Iyer VN, Vlahakis NE. Fetal outcomes of critically ill pregnant women admitted to the intensive care unit for nonobstetric causes. Crit Care Med 2008;36(10):2746-2751. 28. South African Society of Anaesthesiologists. SASA Practice Guidelines 2013 - 2012, Revision. S Afr J Anaest Analg 2013;19(1):S1-S42.

CCRC CRITICAL CARE REFRESHER COURSE

PORT ELIZABETH 2015

CRITICAL CARE REFRESHER COURSE --------------20 - 22 November 2015 Boardwalk Hotel and Convention Centre Port Elizabeth www.ccrc2015.co.za

Multiple modes of action sustain activity1,2 1,5,7,8

FEIBA is indicated for therapy and prophylaxis of haemorrhage and to cover surgical interventions in:3 • Haemophilia A patients with FVIII inhibitor • Haemophilia B patients with FIX inhibitor

FEIBA® is a multicomponent therapeutic agent 1 • Other proteins besides FVlla are involved in its 2

• FXa and prothrombin (Fll) in FEIBA® play a critical role by inducing thrombin generation1 • FIX, which can be activated by FXIa and FVlla, and FX, which can be activated by FIXa and FVlla contribute to the potency of FEIBA® by increasing the respective substrate concentrations1

FEIBA: Sites of Action

Intrinsic Pathway

FEIBA FIX, FIXa

FXIIa

FXII

FXI

FVIIIF

FEIBA FX, FXa

FXIa

FVIIa

FIX FIXa Ca++-PL VIIIa

FEIBA FVII, FVIIa TF

FVII Extrinsic Pathway

FXa FEIBA FII

FII

FEIBA FXa/FII

Blood coagulation Feedback mechanisms PL = Phospholipids TF = Tissue Factor

FVaF Common Pathway Leading to Clot

FEIBA FIIa

FXa

V Fibrinogen

FXIII FXIIIa Fibrin

Fibrin Polymer CLOT Adapted from Turecek PL, et al 20041

FEIBA® acts by targeting steps in both the intrinsic and extrinsic coagulation pathways8

FEIBA therapy gives him extra hours to enjoy his day Up to 12 hours between doses

Help your patients by making FEIBA therapy an integral part of your inhibitor management strategy.

ARTICLE

Organisms cultured and resistance patterns seen in a secondary referral centre ICU and burns unit B Greatorex, BMBS, FRCA; G Oosthuizen, MB ChB, FCS (SA) Departments of Intensive Care and Surgery, Edendale Hospital, Pietermaritzburg, KwaZulu-Natal, South Africa Corresponding author: B Greatorex (benjamingreatorex@hotmail.com)

Background. Infections are common in intensive care units (ICUs) and burns units. Empiric antibiotic therapy is often required, and as such it is important to have a good knowledge of the resident organisms in these departments. Antibiotic resistance is becoming an increasing problem both internationally and in South Africa (SA) and it is important to monitor organism sensitivity. Objectives. To establish the spectrum and sensitivity of nosocomial pathogens in an SA government referral hospital ICU and burns unit. Methods. We report the findings from a retrospective audit of all cultures sent from the ICU and burns unit of an SA urban hospital for a 6-month period between January and June 2008. Results. The results showed a prevalence of Gram-negative organisms in the ICU department, in particular Klebsiella pneumoniae and Escherichia coli. There was a prevalence of Gram-positive organisms in the burns unit. Overall resistance to co-amoxiclav and erythromycin was found to be high (49% and 53%, respectively), resistance to ciprofloxacin and gentamicin was moderate (30% and 35%, respectively) and resistance to piperacillin-tazobactam and the carbapenems remained low (21% for piperacillin-tazobactam, 2% for ertapenem and 19% for meropenem). When looking at individual species, it was noted that K. pneumoniae had high resistance to ampicillin (97%), moderate resistance to co-amoxiclav and ciprofloxacin (35% and 43%, respectively) and low resistance to piperacillin-tazobactam, ertapenem, meropenem and colistin (12%, 0%, 5% and 0%, respectively). E. coli was seen to have high resistance to ampicillin (79%), but low resistance to co-amoxiclav (4%), ciprofloxacin (9%), piperacillin-tazobactam (0%), ertapenem (0%), meropenem (4%) and colistin (0%). Conclusion. This study demonstrates the prevalence of Gram-negative organisms in an SA government hospital ICU. It also demonstrates the presence of resistance mechanisms in the organisms cultured for almost all available classes of antibiotics, albeit some at low levels. The development of multi- and pan-resistant pathogenic organisms is both an SA and worldwide problem. In particular, the threat posed by resistant Gram-negative bacteria is likely to manifest itself in ICUs where septic patients unresponsive to standard antimicrobial regimens will inevitably end up. Frequent assessment of resistance patterns and appropriately designed empirical treatment protocols must remain a priority for all critical care departments. S Afr J Crit Care 2015;31(3):12-15. DOI:10.7196/SAJCC.185

Invasive infections are common in intensive care units (ICUs) and burns units, and mortality from severe sepsis in the critical care setting ranges from 28 to 50%. [1] Antibiotics are commonly prescribed to critically ill patients. The use of antibiotics that are ineffective against the causative organism is associated with increased morbidity and mortality, and is an unnecessary additional strain on what are often stretched healthcare resources.[2,3] A study from 2005 looking at prescribing practices in ICUs across South Africa (SA) identified inappropriate empirical antibiotic choice in 54.9% of patients.[3] Inappropriate use can also result in the development of antibiotic resistance among bacteraemic pathogens.[4] As such, it is important to ensure that empirical antibiotic choice is based on good knowledge of the resident organisms in critical care departments. It is also well known that resistance among pathogens is increasing rapidly both internationally and in SA.[5,6] New forms of antibiotic resistance can spread between countries with ease. Of particular concern are the so-called ‘ESKAPE’ pathogens: Enterococcus faecium, Staphylococcus aureus, Klebsiella pneumoniae, Acinetobacter baumannii, Pseudomonas aeruginosa and Enterobacter spp. [7] Of the aforementioned pathogens, while there is still concern around methicillin-resistant Staphylococcus aureus (MRSA), it is infections secondary to the Gram-negative bacteria that are causing greater concern due to the existence of pan-resistant strains and the

SAJCC June 2015, Vol. 31, No. 1

paucity of novel antibiotics. [8] A recently published audit from an ICU in Durban, South Africa, revealed that from samples sent from ventilated patients, 62% of the positive results were Gram-negative bacteria. Extended spectrum β-lactamase production was found in 19% of Escherichia coli cultured and 25% of Klebsiella spp. isolates.[9] Over time, as antibiotics are used within a department, the resident organisms will adapt and develop resistance mechanisms such as the production of β-lactamases. Therefore, it is important to monitor organisms and their spectrum of resistance on a regular basis.

Objectives To establish the spectrum and sensitivity of nosocomial pathogens in an SA government referral hospital ICU and burns unit.

Methods We report the findings from a retrospective audit of all cultures sent from the ICU and burns unit of Edendale Hospital, Pietermaritzburg, SA, for a 6-month period between January and June 2008. All culture and sensitivity testing was carried out at an on-site laboratory. The data that were assessed were kept in hard-copy form in the microbiology department of the hospital. Edendale is a government hospital with 900 beds, which included at the time of the audit a 6-bed adult ICU and an 8-bed acute burns unit. There were no isolation rooms. There were no ventilators in

the burns unit; patients requiring invasive respiratory support were admitted to the ICU. It was not policy in either the ICU or burns unit to send surveillance cultures at the time the audit was performed. As such, all specimens were sent in response to suspected infection. Specifically, neither tracheal aspirates nor wound swabs were sent routinely. Bronchoscopy and bronchoalveolar lavage were not possible for sputum sample collection in the ICU. All data were recorded in an Excel (Microsoft, USA) spreadsheet, including those samples with no growth. The data that were recorded included the type of sample sent, any organisms grown and their susceptibility to any antibiotics tested. Any results which were from the same patient, during the same visit and which grew the same organism were only counted once. Results for which the genus was given, but not the specific name, were categorised separately under the title spp. (for example, Klebsiella spp.).

Results Table 1 shows the organisms cultured from patients in the ICU. The total number of positive cultures was 127. Gram-negative organisms predominated, in particular K. pneumoniae, which represented 27% of cultured isolates, E. coli (22%) and A. baumannii (17%). Table 2 shows the organisms cultured from patients in the burns unit. The total number of positive cultures was 20. There was prevalence for Gram-positive organisms, in particular S. aureus and S. epidermidis. Table 3 shows the number and type of cultures sent from the ICU during the 6-month period, and the percentage of cultures that came back positive. Blood was the most common sample type sent (158Â samples), of which 60 were growth positive. Despite fewer samples of sputum and tracheal aspirate being sent (76 samples), the number of samples that were growth positive was higher than for blood, at 68. The high percentage of positive growth seen in sputum and tracheal aspirates may reflect the techniques used to collect the samples. The number of growth-positive wound swabs was also high, at 62, a proportion of which almost certainly represented skin commensals. Table 4 shows the number and type of cultures sent from the burns unit. As might be expected, the majority of samples sent were wound swabs (38 samples). Similarly

Table 1. Organisms cultured from patients in the ICU

Table 3. Number and type of cultures sent from the ICU

Gram-negative bacteria cultured

Number of positive cultures (N=127), n (%)

Type of culture

Total number

% growth positive

K. pneumoniae

34 (26.8)

Blood

158

E. coli

28 (22.0)

Sputum/ tracheal aspirate

A. baumannii

22 (17.3)

Urine

Klebsiella spp.

18 (14.2)

Wound swab

S. aureus

13 (10.2)

P. aeruginosa

12 (9.4)

Central venous line tip

Table 2. Organisms cultured from patients in the burns unit

Table 4. Number and type of cultures sent from the burns unit

Gram-positive bacteria cultured

Number of positive cultures (N=20)

Type of culture

Total number

% growth positive

S. aureus

Wound swab

S. epidermidis

Urine

P. mirabilis

Blood

Central venous line tip

to what was seen in the ICU, a high proportion (84%) of wound swab cultures came back positive. Table 5 demonstrates the resistance rates of each genus to all of the antibiotics routinely tested. Low levels of resistance were seen to linezolid (0%), amikacin (8%), vancomycin (10%) and colistin (12%). Moderate levels of resistance were seen to piperacillin-tazobactam (21%), ciprofloxacin (30%), gentamicin (35%) co-amoxiclav (49%) and the 3rd-generation cephalosporins cefotaxime and ceftazidime (49% and 44%, respectively). The level of resistance to the carbapenems varied from 2% for ertapenem up to 23% for imipenem. Table 6 demonstrates resistance rates among individual species. A. baumannii displayed the highest levels of resistance with 80% and 78% resistance to piperacillintazobactam and meropenem, respectively. By comparison, 12% of K. pneumoniae was resistant to piperacillin-tazobactam, while 5% was resistant to meropenem.

Discussion Nosocomial infections cause significant morbidity and mortality among patients. Of these infections, those affecting the lower respiratory tract, urinary tract, bloodstream and postsurgical account for the majority. [10] The increasing resistance rate seen in nosocomial pathogens combined with a paucity of new antimicrobials presents a serious challenge for physicians. [11] There

are concerns that we are entering a postantibiotic era where the treatment of previously curable infections becomes impossible. It is likely that this phenomenon will be seen most acutely in ICUs, as they are often the final destination of patients with septic shock who are not responding to initial, empirical antimicrobial therapy. The organisms that were grown from the samples sent during the 6 months of this audit represent a combination of community-acquired and nosocomial pathogens. Distinguishing which are hospital acquired is not possible from our data. However, knowing the patterns of resistance across all pathogenic species helps us to establish more effective empirical antibiotic protocols.

Organisms grown In this study, Gram-negative bacteria made up the majority of the growthpositive cultures (114 cultured organisms of a total of 127) in the ICU, the three most prevalent of which were K. pneumoniae (26.8% of cultured isolates), E. coli (22.0%) and A. baumannii (17.3%). S. aureus was the only Gram-positive organism cultured in the ICU in significant numbers (10.2%). The preponderance for Gram-negative organisms is consistent with the findings of Brink et al. [4] for private institutions in

SAJCC June 2015, Vol. 31, No. 1

Table 5. Resistance rates to all antibiotics routinely tested across each genus Antibiotic

% resistance

Antibiotic

% resistance

Amikacin

Colistin

Ampicillin

Ertapenem

Co-amoxiclav

Erythromycin

Benzyl penicillin

100

Gentamicin

Cefotaxime

Imipenem

Cefoxitin

Linezolid

Ceftazidime

Meropenem

Cephamdole

Piperacillin-tazobactam

Ciprofloxacin

Trimethoprim

Cloxacillin

Vancomycin

Table 6. Resistance rates among individual species (%) Antibiotic

K. pneumoniae

E. coli

A. baumannii

S. aureus

P. aeruginosa

Amikacin

Ampicillin

100

Cloxacillin

Co-amoxiclav

Cefotaxime

Ciprofloxacin

Colistin

Ertapenem

Erythromycin

Gentamicin

100

Meropenem

Oxacillin

Piperacillin

Piperacillintazobactam

Vancomycin

SA in 2008. They found that K. pneumoniae made up 24.9% of cultured isolates, while E. coli and A. baumannii made up 18.5% and 7.4%, respectively. S. aureus made up 24.6%. Notably, unlike Brink et al.,[4] we saw very few Enterobacter spp. (n=7 positive cultures, all E. cloacae). We also saw very few Enterococci spp. (n=8), none of which were resistant to vancomycin. In 2013, when looking at an SA trauma ICU, Ramsamy et al.[9] found that Gram-negative organisms made up the majority of the cultured organisms. However, in their study, Gram-positive organisms made up a much bigger proportion of the total (119 out of 323). This emphasises the importance of knowing the common pathogenic organisms in a unit when designing empirical antimicrobial protocols. In contrast to the ICU, Gram-positive organisms were the most common organism

SAJCC June 2015, Vol. 31, No. 1

cultured in the burns unit, in particular S. aureus. This isnâ&#x20AC;&#x2122;t surprising perhaps, given the fact that most of the cultures were wound swabs and therefore possibly represented skin flora rather than invasive pathogens. However, immediate colonisation of burns by patientsâ&#x20AC;&#x2122; normal skin flora is well documented and, depending on the individual organismâ&#x20AC;&#x2122;s virulence, can often lead to an invasive burn wound infection.[12]

Specific organisms of concern K. pneumoniae (n=34) K. pneumoniae showed similarly high rates of resistance to ampicillin, co-amoxiclav, ciprofloxacin and the 3rd-generation cephalosporins to those seen by Brink et al. [4] in private institutions. In contrast to

their findings, however, we found low levels of resistance to piperacillin-tazobactam (12% compared with 40%). Resistance to the carbapenems was also comparably low. Despite the increasing prevalence of carbapenemase-producing Klebsiella isolates in Europe (8.3% in 2013)[13] and the US (11% for all Klebsiella spp. in 2013),[14] documented rates of carbapenem resistance in Africa remain low, at between 0 and 4% (although there is a paucity of data).[15] E. coli (n=28) Although resistance to ampicillin among E.coli isolates was also high (79%), they were surprisingly sensitive to all other classes of antimicrobials (4%, 9% and 8% for co-amoxiclav, ciprofloxacin and cefotaxime, respec tively). This is in contrast to the findings of both Brink et al., [4] who documented resistances of 37%, 20% and 10%, respectively, for the same three antimicrobials, and the World Health Organization ( WHO) [15] data for 2014, which showed resistance in Africa to fluoroquinolones of between 34 and 53% in invasive isolates and resistance to 3rd-generation cephalosporins of between 28 and 36%. There was no documented resistance to either piperacillin-tazobactam or ertapenem, and only 4% to meropenem. A. baumannii (n=22) Of all the organisms cultured from patients in the Edendale ICU, A. baumannii exhibited the most extensive range of antimicrobial resistance. This is in contrast to Brink et al.,[4] who saw significantly lower levels of resistance to ciprofloxacin (90% compared with 36%), 3rd-generation cephalosporins (91% compared with 43%) and piperacillintazobactam (80% compared with 42%). Even more concerning, however, was the level of resistance to meropenem (78%). The Centers for Disease Control reported in 2013 that 63% of all nosocomial A. baumannii infections in the USA were multidrug resistant. The only effective antimicrobial tested at Edendale was the polymyxin, colistin (0% resistance). The use of colistin as monotherapy, part of combination therapy and via novel routes such as nebulisation has seen a renewed interest in recent years. While there have been doubts in the past about the virulence of A. baumannii, given the ability of micro-organisms to transfer resistance genes between genus, the presence of bacteria with such an extensive spectrum of resistance is a large concern.

S. aureus (n=20) The overall resistance to cloxacillin or oxacillin, thus conferring methicillin-resistant status, was 46%. This is within the range reported by the WHO [15] for the African continent in 2014 (12 80%). However, it is slightly greater than that seen by Brink et al.[4] (36%) and significantly higher than the EU average for 2013 of 18% (although the EU intercountry variation is 1 - 60%). It should be noted that MRSA rates in both the US and Europe have been falling over the past 6 years, possibly secondary to focused national infection control campaigns.[13,14] In line with most other studies, resistance to vancomycin was low at 7%. P. aeruginosa (n=12) Resistance to ciprofloxacin, gentamicin and piperacillin-tazobactam was low at 7%, 0% and 7%, respectively. This is in marked contrast to Brink et al.,[4] who saw resistance in the private sector of 46% for ciprofloxacin, and 48% for both amikacin and piperacillintazobactam. Resistance to both colistin and meropenem was around 15%, again lower than Brink et al. The CDC have reported that ~8% of all healthcare-associated infections were caused by P. aeruginosa and that 13% of these were multi-drug resistant.

Study limitations These data are historical, and while useful in guiding empirical antimicrobial protocols at the time they were collected, it is likely that the spectrum of antimicrobial resistance has changed significantly since. To guide current protocols, further data should be collected. Data regarding length of stay before cultures were taken was not collected, so no distinction can be made between community- and hospital-acquired infections. Distinguishing the hypermetabolic response to burns from sepsis is challenging, and as such, the wound swabs sent from the burns unit may often reflect colonising bacteria rather than invasive pathogens.

Conclusions This study demonstrates the prevalence of Gram-negative organisms in an SA government hospital ICU. It also demonstrates that while at

low levels for certain classes of antibiotics, specifically carbapenems, resistance mechanisms exist in the organisms cultured for almost all available classes of antibiotics. The development of multi- and pan-resistant pathogenic organisms is both an SA and worldwide problem. In particular, the threat posed by resistant Gramnegative bacteria is likely to manifest in ICUs where septic patients unresponsive to standard antimicrobial regimens will inevitably end up. Frequent assessment of resistance patterns and appropriately designed empirical treatment protocols must remain a priority for all critical care departments. References 1. Sablotzki A, MĂźhling J, Czeslick E. Sepsis and multiple organ failure: Update of current therapeutic concepts. Anaesthesiol Intensivmed Notfallmed Shmerzther. 2005;40(9):511-520. [http://dx.doi.org/10.1055/s-2005-870104] 2. Colardyn F. Appropriate and timely empirical antimicrobial treatment of ICU infections: A role for carbapenems. Acta Clin Belg 2005;60(2):51-62. [http://dx.doi.org/10.1179/acb.2005.011] 3. Paruk F, Richards G, Scribante J, Bhagwanjee S, Mer M, Perrie H. Antibiotic prescription practices and their relationship to outcome in South African intensive care units: Findings of the Prevalence of Infection in South African Intensive Care Units (PISA) Study. S Afr Med J 2012;102(7):613-616. 4. Brink A, Moolman J, Da Silva MC, Botha M, National Antibiotic Surveillance Forum. Antimicrobial susceptibility profile of selected bacteraemic pathogens from private institutions in South Africa. S Afr Med J 2007;97(4):273-279. 5. Brink A, Feldman C, Richards G, Moolman J, Senekal M. Emergence of extensive drug resistance (XDR) among gram negative bacilli in South Africa looms nearer. S Afr Med J 2008;98(8):586-592. 6. Jones RN. Resistance patterns among nosocomial pathogens: Trends over the past few years. Chest 2001;119(2 Suppl):S397-S404. 7. Rice LB. Federal funding for the study of antimicrobial resistance in nosocomial pathogens: No ESKAPE. J Infect Dis 2008;197:1079-1081. [http://dx.doi.org/10.1086/533452] 8. Pop-Vicas A, Opal SM. The impact of multi-drug resistant gram-negative bacilli in the management of septic shock. Virulence 2014;5(1):206-212. [http://dx.doi.org/10.4161/viru.26210] 9. Ramsamy Y, Muckart DJ, Han KS. Microbiological surveillance and antimicrobial stewardship minimise the need for ultrabroad-spectrum combination therapy for treatment of nosocomial infections in a trauma intensive care unit: An audit of an evidence-based empiric antimicrobial policy. S Afr Med J 2013;103(6):371-376. [http://dx.doi.org/10.7196/samj.6459] 10. DusĂŠ AG. Infection control in developing countries with particular emphasis on South Africa. South Afr J Epidemiol Infect 2005;20(2):37-41. 11. Pertrosillo N, Capone A, Di Bella S, Taglietti F. Management of antibiotic resistance in the intensive care unit setting. Expert Rev Anti Infect Ther 2010;8(3):289-302. [http://dx.doi. org/10.1586/eri.10.7] 12. Church D, Elsayed S, Reid O, Winston B, Lindsay R. Burn wound infections. Clin Microbiol Rev 2006;19(2):403-434. [http://dx.doi.org/10.1128/CMR.19.2.403-434.2006] 13. European Centre for Disease Prevention and Control. Summary of the latest data on antibiotic resistance in the European Union. 2014. http://ecdc.europa.eu/en/eaad/Pages/antibioticsdata-reports.aspx (accessed 1 February 2015). 14. US Department of Health and Human Services Centre for Disease Control and Prevention. Antibiotic Resistance Threats in the United States, 2013. http://www.cdc.gov/drugresistance/ pdf/ar-threats-2013-508.pdf (accessed 1 February 2015). 15. World Health Organization. Antimicrobial Resistance Global Report on Surveillance. 2014. http:// www.who.int/drugresistance/documents/surveillancereport/en/ (accessed 1 February 2015).

SAJCC June 2015, Vol. 31, No. 1

ARTICLE

Incidence and outcome of ventilator-associated pneumonia in Inkosi Albert Luthuli and King Edward VIII Hospital surgical intensive care units A Awath Behari, MB ChB, DA (SA), FCA (SA); N Kalafatis, MB BCh, DA (SA), FCA (SA), Cert Critical Care (SA) Department of Anaesthesia and Critical Care, University of KwaZulu-Natal, Durban, South Africa Corresponding author: A Awath Behari (a.behari1@gmail.com) Background. Ventilator-associated pneumonia (VAP) is one of the most common causes of hospital morbidity and mortality, but has been poorly studied in the South African context. Objective. To evaluate the incidence and outcome of VAP in the intensive care units (ICUs) of two major centres in the Durban metropolitan area. Methods. The study was conducted over a period of 6 months with all intubated and mechanically ventilated patients who were screened on admission to ICU. A questionnaire was prepared to note patients’ age, gender, date and time of intubation or reintubation. Patients were monitored from date of admission to the date of discharge from ICU or death. A diagnosis of VAP was made on a clinical pulmonary infection score (CPIS) of ≥6. Results. Of 32 patients evaluated, eight patients (25%) were diagnosed with VAP. Median duration of ventilation in the VAP group was 249 hours v. 65.5 hours in the non-VAP group (p=0.0002). We found no statistically significant association between age or gender with the development of VAP (p=0.28 and p=0.59, respectively). The most common organism isolated was Acinetobacter baumannii, followed by Pseudomonas aeruginosa. Three of the eight (37.5%) patients diagnosed with VAP died in the ICU. Conclusion. VAP is common in critically ill patients, possibly associated with poor outcome. These results highlight the need for strict adherence to evidence-based preventive measures. S Afr J Crit Care 2015;31(1):16-18. DOI:10.7196/SAJCC.227

Ventilator-associated pneumonia (VAP) is one of the most common causes of hospital morbidity and mortality. [1] VAP refers to pneumonia developing in patients who have been receiving mechanical ventilation for at least 48 hours, and may be further categorised into early-onset VAP (<96 hours) and late-onset VAP (≥96 hours).[2] Prevalence ranges from 10 to 25% in tertiary care hospitals, and can reach 76% in some settings.[3] VAP is associated with substantial morbidity and excess cost, and patients with VAP have been found to be twice as likely to die than those without VAP.[4] Gram-negative organisms are the most commonly associated microbial flora.[5] The fundamental problem with the diagnosis of VAP is the lack of an internationally accepted gold standard. According to the Centers for Disease Control and Prevention,[6] VAP may be diagnosed by: (i) a new or progressive pulmonary infiltrate; (ii) fever, leukopenia or leukocytosis; (iii) purulent tracheobronchial secretions; (iv) and worsening gas exchange. However, these criteria are nonspecific and of little utility in the diagnosis of VAP. An autopsy investigation showed that only 52% of patients with pneumonia had a localised infiltrate on their chest radiograph.[7] Furthermore, fever and leukocytosis may be caused by other foci of infection in the intensive care unit (ICU) setting. In an attempt to increase the likelihood of diagnosing VAP, Pugin et al.[3] created the Clinical Pulmonary Infection Score (CPIS, Table 1) based on sputum smear microscopy and tracheal aspirate culture, as well as on the clinical findings at the time of diagnostic suspicion. In that study, the authors concluded that there was a good

SAJCC June 2015, Vol. 31, No. 1

correlation between clinical score and quantitative bacteriology, with a sensitivity of 72% and specificity of 80%. A CPIS threshold of 6 was found to be a fairly accurate measure of the presence or absence of pulmonary infection, as signified by bacterial culture. VAP has been poorly studied in the South African (SA) context. A literature search showed no reported studies pertaining to VAP incidence and aetiology in adults in SA. The primary objective of this study was to evaluate the incidence and outcome of VAP in the ICUs of two major centres in the Durban metropolitan area.

Methods A literature search was done on PubMed and Google using the search terms ‘ventilator-associated pneumonia, incidence, outcome, mortality.’ The study was conducted over a period of 6 months in the King Edward VIII Hospital ICU and Inkosi Albert Luthuli Central Hospital surgical ICU. Ethical approval was obtained, and consent was waived on the basis of the study being a prospective chart review with no patient contact. All intubated and ventilated patients were screened on admission to the ICUs. A questionnaire was prepared to note each patient’s age, gender, date and time of initial intubation and initiation of mechanical ventilation, and date and time of reintubation and ventilation. The parameters included in the CPIS were tabulated to enable a daily score to be calculated by the attending doctor. Inclusion criteria were as follows: all patients ≥18 years of age admitted to the ICU who were intubated and ventilated for a minimum of 48 hours. Exclusion criteria were suspected

or confirmed community-acquired or nosocomial pneumonia on admission, patient age <18 years, and patients who were managed on non-invasive ventilation. Patients were monitored from the date of admission into the ICU to the date of discharge or date of death. A diagnosis of VAP was made on a CPIS of ≥6. Due to the cost and ethical issues surrounding daily radiographic and microbiological testing, a score of 0 was allocated to these parameters (chest X-ray infiltrates and microbiology) if none was available on the day of evaluation. Semiquantitative and qualitative tracheal aspirations and blood cultures were obtained. At the time of study, patients were managed by the attending clinician and all interventions were decided upon based on the clinician’s individual assessment of the patient, including the initiation of antimicrobial therapy according to the unit’s antimicrobial stewardship programmes. The association between the diagnosis of VAP and duration of ventilation was statistically analysed using the MannWhitney test, the association between age and development of VAP was studied using the two-sample t-test and the association between gender and VAP was studied using the one-sided Fisher’s exact test. A p-value of <0.05 was considered significant.

Results The study cohort comprised 32 patients (male n=19, 59.4%) admitted for various surgical and medical pathologies. Eight of the 32 patients (male n=5, 25.0%) were diagnosed with VAP, corresponding to a rate of 9.9 per 1 000 ventilator days. There were no statistically significant associations between age or gender and the develop ment of VAP (Table 2). Patients with VAP had a significantly longer duration of mechanical ventilation than those without VAP (Table 2). Of the patients diagnosed with VAP, three out of eight (37.5%) died in the ICU. Mortality in the non-VAP group was not recorded. All of the VAP cases except Case 1 fell under the definition of late-onset VAP. The most common organisms associated with VAP in were Acinetobacter baumannii (37.5%), Pseudomonas aeruginosa (25.0%), followed by Haemophilus influenzae, Streptococcus pneumoniae, and Escherichia coli (12.5% each) (Table 3), with no organsim isolated in one of the patients diagnosed with VAP by CPIS.

Table 1. CPIS parameters and scoring[3] CPIS points

Tracheal secretions

Rare

Abundant

Abundant and purulent

Chest X-ray infiltrate

No infiltrate

Diffuse

Localised

Temperature (˚C)

≥36.5 and ≥38.4

≥38.5 and ≤38.9

≤36.5 or ≥39.0

Leukocytes (mm3)

>4 000 and <11 000

<4 000 and >11 000

<4 000 or >11 000 and band forms

PaO2/FiO2 (mmHg)

>240 or ARDS

≤240 and no ARDS

Microbiology

Negative

Positive

ARDS = acute respiratory distress syndrome; PaO2/FiO2 = ratio of partial pressure of arterial oxygen to fraction of inspired oxygen.

Table 2. Selected patient characteristics and outcomes in those with and without development of VAP VAP group (n=8)

Non-VAP group (n=24)

p-value

Age (years), median (IQR)

53 (27 - 63)

50 (27 - 59)

0.28

Male gender, mean (%)

5.0 (62.5)

14.0 (58.3)

0.59

Duration of mechanical ventilation (hours), median (IQR)

249.0 (130.5 - 333.5)

65.5 (56.0 - 87.0)

0.0002

Discussion The primary objective of this study was to assess the incidence, aetiology and outcome of VAP in a heterogenous patient population admitted to our ICUs with both medical and surgical pathologies. The incidence of VAP in our setting was 25%, which is in keeping with rates of 15.5 - 27.5% quoted in other studies using similar methodology. [8] However, a recent surveillance study by Kollef et al. [9] on the epidemiology of VAP due to P. aeruginosa found the global incidence of VAP to be lower, at 15.6%, with a regional incidence of 13.5% in the USA, 19.4% in Europe, 13.8% in Latin America and 16.0% in Asia Pacific. Despite the clinical popularity of the CPIS, debate continues regarding its diagnostic validity. Its apparent straightforward calculation is beneficial; however, the inter-observer variability in CPIS calculation remains substantial, jeopardising its routine use in clinical trials.[10] The lack of an international gold standard for the diagnosis of VAP makes comparison of different studies difficult and inaccurate. Unfortunately, we as clinicians are nowhere near achieving this goal due to the overlapping of clinical presentations with other causes of sepsis or ARDS, resulting in a high sensitivity but low specificity using clinical diagnostic

Table 3. VAP group organisms and outcomes Case number

Microorganisms isolated

ICU outcome

No result

Discharged

E. coli and P. aeruginosa

Discharged

A. baumannii

Died

S. pneumoniae

Discharged

A. baumannii

Discharged

H. influenzae

Discharged

A. baumannii

Demised

P. aeruginosa

Demised

Mortality rate

37.5%

criteria. It is hoped that new developments in the isolation of specific biomarkers would provide a solution to this diagnostic conundrum. Age and gender were not found to be statistically significant contributors to the development of VAP, but duration of mechanical ventilation was found to be highly significantly associated with the development of VAP. However, we are unable to determine cause and effect on the basis of this study. The association between duration of mechanical ventilation

SAJCC June 2015, Vol. 31, No. 1

and the development of VAP has been reported previously, for example Gadani et al.,[8] in a study of 100 patients, concluded that the incidence of VAP was directly proportional to the duration of mechanical ventilation. This highlights the need to avoid intubation and mechanical ventilation if at all possible. The rate of VAP in noninvasive positive pressure-ventilated patients is lower and should be the ventilation modality utilised if proven to be equal if not superior to invasive means of ventilation for the disease process.[11] Both units in our study had VAP bundles in place, including head-up position, hand-washing protocols and early weaning protocols incorporating sedation holds. We found that Gram-negative organisms were the most common associated pathogens, with A. baumannii and P. aeruginosa being the most common organisms isolated in the patients with VAP. Although these organisms are commonly associated with VAP in different settings, patient profile and nature of the ICU can contribute to a higher prevalence of other organisms.[5] The majority of the VAP cases were of the late-onset subtype in which multidrug-resistant bacteria are known to be the most prevalent.[11] Aetiological data collection and interpretation provides vital information on most likely causative organisms and resistance patterns. This assists clinicians in directing their choice of empirical treatment if VAP is suspected. Therefore, it is suggested that surveillance studies be adopted in all ICU settings. The question of the effect of VAP on mortality of critically ill patients is certainly a pertinent one to the clinician. Of the patients who developed VAP, 37.5% died in the ICU. A recent meta-analysis on the attributable mortality of VAP conducted by Melsen et al.[12] showed an overall attributable mortality of VAP to be 13%. Due to the heterogenous nature of our study population, with differing comorbidities and admission diagnoses as well as limited outcome data collection in those who did not develop VAP, we are unable to comment on VAP being an independent risk factor for mortality. However, the mortality rate in our study does warrant further investigation.

Study limitations The major limitation is the small sample and the lack of patient stratification prior to investigation, preventing multivariate analysis. Our sample was based on other similarly designed studies, with populations of between 51 and 100 patients.[3,8] Furthermore, we did not look at medical and surgical admissions separately but instead

SAJCC June 2015, Vol. 31, No. 1

looked at the patients in both ICUs as a single population. Patientsâ&#x20AC;&#x2122; baseline function, comorbidities, injury severity score and management confounders such as dialysis and patient transfer complications need to be documented and evaluated as possible contributing factors for the development of VAP. Possible areas of study for this unit include adherence to VAP bundle elements and the utilisation of other prevention studies such as feeding protocols, monitoring of endotracheal tube cuff pressures, minimising transfers out of ICU, etc., with subsequent similar surveillance studies to ascertain if these interventions affect the development of VAP. Larger, multicentred studies are recommended to address and help minimise the effect of this disease process on a national level.

Conclusion VAP is a common pathology in critically ill patients, possibly associated with poor outcome. We found a significant association between duration of ventilation and development of VAP, which highlights the essential need for implementation of VAP preventive bundles, weaning protocols and strict adherence to infection control policies. References 1. Rotstein C, Evans G, Born A, et al. Clinical practical guidelines for hospital-acquired pneumonia and ventilator-associated pneumonia in adults. Can J Infect Dis Med Microbiol 2008;19(1):1953. 2. Vanhems P, BĂŠnet T, Voirin N, et al. Early-onset ventilator-associated pneumonia incidence in intensive care units: A surveillance-based study. BMC Infect Dis 2011;11:236. [http://dx.doi. org/10.1186/1471-2334-11-236] 3. Rakshit P, Nagar VS, Deshpande VAK. Incidence, clinical outcome, and risk stratification of ventilator-associated pneumonia: A prospective cohort study. Indian J Crit Care Med 2005;9(4):211-216. [http://dx.doi.org/10.4103/0972-5229.19761] 4. Nseir S, Di Pompeo C, Jozefowicz E, et al. Relationship between tracheotomy and ventilatorassociated pneumonia: A case-study. Euro Resp Journal 2007;30(2):314-320. [http://dx.doi.org /10.1183/09031936.06.00024906] 5. Rello J, Quintana E, Austina V, et al. Incidence, etiology, and outcome of nosocomial pneumonia in mechanically ventilated patients. Chest 1991;100(2):439-444. [http://dx.doi. org/10.1378/chest.100.1.439] 6. Centres for Disease Control. Ventilator associated pneumonia guideline. http://www.cdc.gov/ nhsn/pdfs/pscmanual/6pscvapcurrent (accessed 9 September 2014). 7. Prescott HC, Oâ&#x20AC;&#x2122;Brien JM. Prevention of ventilator-associated pneumonia in adults. Med Rep 2010;2(15):15. [http://dx.doi.org/10.3410/M2-15] 8. Gadani H, Vyas A, Kar A. A study of ventilator associated pneumonia: Incidence, outcome, risk factors and measures to be taken for prevention. Indian J Anaesth 2010;54(6):535-540. [http:// dx.doi.org/10.4103/0019-5049.72643] 9. Kollef MH, Chastre J, Fagon JY, et al. Global prospective epidemiologic and surveillance study of ventilator-associated pneumonia due to Pseudomonas aeruginosa. Crit Care Med 2014;42(10):2178-2187. [http://dx.doi.org/ 10.1097/CCM.0000000000000510] 10. Kalanuria AA, Zai W, Mirsk M. Ventilator-associated pneumonia in the ICU. Crit Care 2014;18(2):208 [http://dx.doi.org/10.1186/cc13775] 11. Hess DR. Non-invasive positive-pressure ventilation and ventilation-associated pneumonia. Respir Care 2005;50(7):924-931. 12. Melsen WG, Rovers MM, Groenwald RH, et al. Attributable mortality of ventilator-associated pneumonia: A meta-analysis of individual patient date from randomised prevention studies. Lancet Infect Dis 2013;13(8):665-671. [http://dx.doi.org/10.1016/S1473-3099(13)70081-1]

ARTICLE

Endotracheal tube verification in adult mechanically ventilated patients P Jordan, PhD; W Ten Ham, PhD; D Fataar, RN, CCN School of Clinical Care Sciences, Nelson Mandela Metropolitan University, Port Elizabeth, South Africa Corresponding author: P Jordan (portia.jordan@nmmu.ac.za) Objective. To explore the methods that can be used to verify endotracheal tube (ETT) placement in adult mechanically ventilated patients. Methods. An integrative literature search was conducted in 2012 - 2013 of research citations published in English on the topic of discussion. Electronic databases searched were: the Cumulative Index of Nursing and Allied Health (CINAHL), MEDLINE, PubMed, the Joanna Briggs Institute (JBI) systematic review library, the Cochrane Library and the National Guidelines Clearinghouse. In addition, reference lists of articles, conference summaries and hand searching was performed. Citations were selected based on the inclusion and exclusion criteria as decided upon by the researchers. The process of critical appraisal was done by the researchers as well as an independent reviewer, all skilled in the research methodology and subject matter related to the topic of discussion. A total of 45 articles were included for critical appraisal. On completion of the critical appraisal, which was done by two independent reviewers, 34 articles were excluded and 11 articles were included in the integrative review analyses. Data were extracted following the critical appraisal process. Owing to the heterogeneity of studies, a metasynthesis could not be done. Results. Based on the reviewed studies, various methods have been identified to verify ETT placement in adult mechanically ventilated patients, namely ultrasonography, the use of centimetre scale printed on the ETT, manual cuff palpation, bilateral auscultation of chest and palpation of symmetrical chest movements, oesophageal detector devices, visualisation of the ETT, use of chest X-ray, pulse oximetry and capnography. Both ultrasonography and capnography had excellent sensitivity and specificity in verifying ETT placement. Conclusion. Although there are various methods reported for ETT verification, the review results recommended ultrasonography and capnography as the most accurate and reliable verification methods. S Afr J Crit Care 2015;31(1):20-23. DOI:10.7196/SAJCC.199

Mechanical ventilation is a lifesaving treatment modality, used most commonly in critical care units. A retrospective cohort over a defined period from six states in the USA reported that of 6 469 674 hospitalisations, 180 326 (2.8%) received invasive mechanical ventilation. The estimated national cost related to mechanical ventilation was USD27 billion, representing 12% of all hospital costs. [1] A paucity of literature exists with regard to South African statistics related to the incidence and cost of mechanical ventilation in critical care units. In order to mechanically ventilate critically ill patients, an endotracheal tube (ETT), the most commonly used artificial airway,[2] has to be inserted into the patient’s trachea. ETT intubation has been classified as a highly technical and clinical skill that is accompanied by the danger of complications. These complications can occur during the intubation procedure, while the ETT is in place or after the ETT has been removed. Incorrect placement of the ETT may lead to inadequate ventilation, displacement, aspiration, ineffective oxygenation, hypoxia, hypotension and oesophageal intubation. Verification of the ETT in mechanically ventilated patients is therefore important.[3-8] Different methods can be used to verify the placement of the ETT.

Objective To search for evidence on the methods to verify ETT placement in adult mechanically ventilated patients.

SAJCC June 2015, Vol. 31, No. 1

Methods

Literature search The search strategy was designed to access both published and unpublished literature, and was conducted by D Fataar under supervision of an experienced researcher (P Jordan).

Electronic search An initial search was done to identify relevant keywords contained in the titles of research studies found. Thereafter, a specific search was done using the following electronic databases: Cumulative Index of Nursing and Allied Health (CINAHL), EBSCOhost, MEDLINE (via PubMed), the Joanna Briggs Institute (JBI) systematic review library, the Cochrane Library, the National Guidelines Clearing house and internet searching engines (Google and Google Scholar). A combination of search terms was used to search the literature: ‘verification of endotracheal OR tracheal tube placement; ‘complications AND incorrect endotracheal tube placement’; ‘critical care AND endotracheal tube placement’; ‘full text journal articles related to endotracheal tube placement AND verification’; ‘endotracheal tube placement AND adults NOT neonates’; ‘ventilation AND verification of endotracheal tubes’.

Hand searching relevant journals Hand searching through the contents pages of journals pertaining to critical care nursing and medicine helped to identify any relevant evidence.

Grey literature The search for grey literature was to ensure that unpublished research on the topic was consulted. University databases were searched for unpublished treatises and theses. Reference lists and bibliographies of potential eligible articles and summaries of conference proceedings were searched, and authors known in the field were contacted to enquire if there were any data findings that were pending for publication.

Inclusion criteria The types of studies included in this review were evaluated according to the study design and classified from Level I to VI as described by LoBiondo-Wood and Haber [9] ( Table 1). Studies with human, adult patients older than 18 years, and who were intubated and attached to a mechanical ventilator in a critical care unit were included. Interventions related to the verification of the ETT placement in the patient were considered.

Table 1. Evidence hierarchy for rating levels of evidence associated with study design[9] Level Description I

Systematic review or meta-analysis of RCTs Evidence-based clinical practice guidelines based on systematic reviews

RCT

III

Quasi-experimental study

Non-experimental study (survey, case-control, case study, cohort, observational, prospective, correlation)

Systematic reviews of descriptive and qualitative studies

Single descriptive or qualitative study

VII

Opinion of authorities and/or reports of expert committees

RTC = randomised control trial

Initial search databases n=350 Grey literature n=5

Duplicates n=2 Included for critical appraisal n=45

Exclusion criteria All studies that focused on paediatric patients and neonates, as well as animal studies were excluded, as there are anatomical and physiological differences between these population groups.

Excluded after critical appraisal n=34 Included in study n=11

Critical appraisal The methodological rigour of the included studies was assessed using the critical appraisal tools available in the JBI SUMARI software packages, version 4.0 (Joanna Briggs Institute, Australia). The MAStARI (JBI Meta-analysis of Statistics Assessment and Review Instrument) was used for crit ical appraisal process. Two reviewers, who are experts in critical care, independently appraised the studies found using the selected appraisal tools. Once the reviews were done independently, the JBI system was used to extract and consolidate the results.

Data extraction I nformation on the author, journal, publication date, setting, sample size, inter vention, outcomes, allocation concealment, loss to follow-up, appropriate statistics and adequate follow-up were extracted by D Fataar, using the appropriate data extraction tools as per the JBI-MAStARI version 4.0 software package. The extraction process was verified by the current author (P Jordan).

Excluded (not meeting criteria) n=33

Included n=80

Fig. 1. PRISMA

Ethical considerations Although the ethical principles of beneficence, respect for human dignity and justice were not directly applicable to an integrative literature review, ethical approval to conduct the study was nevertheless obtained by the ethics committee of the university where the study was conducted (ethics number: H11-HEA-NUR-007).

Results The initial search delivered 350 articles, of which five were obtained from grey literature; the majority of the other articles were obtained from Google Scholar and PubMed. After exclusion due to not being relevant to the topic of discussion, 80 articles were included. After the second screening was done, another 35 articles were excluded as there were two duplicates, and the other 33 articles did not answer the research question. A total of 45 articles were included for critical appraisal. Based on the critical appraisal results of the researcher and independent reviewer,

34 of these articles were excluded, with only 11 articles included in the integrative review analyses. The selection process of the studies is outlined in Fig. 1. The characteristics of the 11 included articles (Level II n=3, Level IV n=8) are presented in Table 2.

Discussion From the data analysed, different methods were identified to verify ETT placement in adult mechanically ventilated patients. The accuracy of any technique to identify correct ETT placement is based on its sensitivity (ability to detect whenever tracheal intubation does occur) and specificity (ability to detect whenever tracheal intub ation does not occur).

Ultrasonography Two of the 11 studies, which were both randomised controlled trials, reported that ultrasonography is a reliable verification method; Muslu et al. [10] reported 100%

SAJCC June 2015, Vol. 31, No. 1

Table 2. Characteristics of 11 studies reviewed Study

Setting and participants

Study design

Intervention

Verification methods identified

Muslu et al., 2011[10]

150 adult patients who were scheduled for elective surgery and postop ICU admittance

RCT (Level II)

Two treatment groups to investigate the use of sonography for ETT confirmation

Ultrasonography was reported as an effective method to confirm ETT placement (100% sensitive and 100% specific).

Werner et al., 2007[11]

33 adult ICU patients who underwent elective surgery

RCT (Level II)

Two treatment groups to evaluate the accuracy of ultrasonography in confirming ETT placement

Ultrasonography was 100% sensitive and 97% specific in accurately verifying ETT placement.

Sitzwohl et al., 2010[12]

160 adult ICU patients who underwent elective surgery

RCT (Level II)

Eight study groups to compare three different bedside methods of verification of correct ETT placement

Bilateral auscultation of the chest, observation and palpation of symmetrical chest movements, use of cm scale printed on the tube and a combination of the three methods were evaluated. The highest sensitivity and specificity were obtained when combining all three methods when verifying ETT placement.

Varshney et al., 2011[13]

200 adult ICU patients who underwent elective surgery

Nonexperimental study: prospective (Level IV)

Using the centimetre scale printed on the ETT as a verification method for placement

The centimetre scale printed on the ETT can be used. However, optimal depth of the ETT placement can be estimated by a formula (height in centimetres/7 -2.5); the method is height dependent and should not be used as a sole verification method.

Simpson et al., 2012[14]

794 ICU adult patients

NEO (Level IV)

Assessing practices related to ETT placements

Capnography was recommended to confirm ETT placements.

Delorio, 2005[15]

550 emergency physicians involved in intubation of ICU patients

Nonexperimental: survey (Level IV)

Availability of capnography as a Although capnography was the recommended method to confirm ETT placement method for ETT verification, it is not widely available or consistently used.

Dittrich, 2002[16]

23 ICU patients

Nonexperimental: case study (Level IV)

Evaluating clinical methods (auscultation and palpation of chest), cuff palpation, CXR, pulse oximetry, capnography and EDD as verification methods

Grmec, 2002[17]

345 adult intubated ICU patients

NEO (Level IV)

Comparing three methods: Of three methods tested, capnography was the auscultation, capnography and cap most reliable method to confirm ETT placement in nometry to confirm ETT placement emergency conditions.

Ledrick et al., 2008[18]

163 ICU intubated patients

NEO (Level IV)

Evaluating manual cuff palpation as Manual cuff palpation can be used as a verification a method to confirm ETT placement method; however, the use of this method is limited and not recommended as a sole method.

Takeda, et al., 2003[19]

137 ICU intubated patients

NEP (Level IV)

Assessing three methods, namely Of three methods assessed, capnography was the the use of clinical signs, capnography most reliable method of verifying ETT placement and EDD to verify ETT placement in non-cardiac arrest patients.

Hussain et al., 2006[20]

400 intubated adult patients

NEP (Level IV)

Efficacy of EDDs as a verification method

Among the verification methods explored, capnography was the most reliable, sensitive and specific method to confirm ETT placement (sensitivity of 93% and specificity of 97%).

The use of EDD as a verification method shown to be effective.

NEO = non-experimental observational, NEP = non-experimental prospective.

sensitivity and specificity and Werner et al. [11] reported 100% sensitivity and 97% specificity. The use of ultrasound may be limited by availability of both expertise and equipment.

Using centimetre scale printed on the ETT Sitzwohl et al.[12] explored three methods, namely the use of bilateral auscultation of the chest, observation and palpation of symmetrical chest movements, and the use of centimetre scale printed on the ETT. Their study showed that among single tests, the best way of excluding endobronchial intubations with the highest sensitivity was by observing the centimetre scale on the ETT. However, when all three bedside methods were combined, sensitivity was higher than observing the centimetre scale alone. Varshney et al.[13] reported in their study that the centimetre scale printed on the ETT could be used as a verification method. They suggested that optimal depth of the ETT placement could be estimated by means of a formula: height in cm/7 – 2.5. The study further recommended the 21/23 cm

SAJCC June 2015, Vol. 31, No. 1

rule, i.e. a correct depth of 21 cm for women and 23 cm in men when checking the ETT length. According to their study, the method is height dependent and should not be used as the sole verification method. As only two studies were found that highlighted this verification method, more research is needed to confirm this method for verification of ETT placement.

Manual cuff palpation Two of the 11 studies highlighted the use of ballottement, or manual cuff palpation, to determine if the tube is at the correct depth. This involves palpating the pilot balloon by applying pressure in the suprasternal notch. Manual cuff palpation is a simple technique, but is limited in identifying intubations of appropriate depth. Ballottement was also found to lead to complications due to pain and unwanted rise in both blood and intracranial pressures, and is therefore not recommended for use in the critical care setting.[13,18] This method is thus not recommended for ETT placement verification.

Bilateral auscultation of chest and palpation of symmetrical chest movements Bilateral auscultation of the chest can be done to identify and prevent possible endobronchial intubation. Although auscultation of the lungs can be used to verify the position of the ETT, it may be deceptive in patients with decreased lung compliance or in patients who experience severe bronchospasm.[12] In false negative results by auscultation, examiners did not clearly hear breath sounds and did not see good chest wall excursions because of obesity, or breath sounds were mistakenly identified as stomach gurgling in some clinical conditions such as pulmonary oedema, excessive secretions or aspiration.[17,19] Auscultation is a common method to ensure correct placement of the ETT; however, it is inaccurate when used alone and by inexperienced examiners. Furthermore, auscultation does not reveal how well the lungs are functioning and whether or not blood is being oxygenated effectively for gaseous exchange.[17] Auscultation and palpation of symmetrical chest movements are most reliable when used with other methods, such as capnography.[13,17]

Capnography Five of the 11 studies (all observational) recommended capnography as the most reliable method for confirming ETT placement in all settings, including the operating theatre. [14-17,19] According to Simpson et al., [14] capnography should be used to confirm ETT position in all intubations, including those performed outside the operating theatre.[13] Delorio[15] confirmed that capnography is the recommended method for ETT verification, but also recognised that the use of capnography is limited and is often either not available or inconsistently applied in ETT verification. Dittrich[16] confirmed that capnography is the most reliable method for verification of the ETT. This study showed that of three methods explored, capnography had the highest sensitivity (93%) and specificity (97%). An advantage of using capnography is that the method may be more easily applied without the need for specific expertise, compared with ultrasound.

EDDs It has been reported in three of the 11 studies that EDDs, consisting of either a self-inflating bulb or a 60 mL syringe, have become one of the simplest methods to confirm ETT placement. In a prospective study by Hussain et al.,[20] it was shown that EDDs had a sensitivity, specificity and positive predictive value of discriminating oesophageal from endotracheal intubation of nearly 100% in healthy adults who were intubated. The effectiveness of this method can be affected by the rigidity and structural differences of the trachea, as well as secretions, vomit, blood or any other fluids in the airway.[16,19,20] During cardiac arrest, negative results caused by the use of this method are not uncommon and clinical methods should then be applied as an adjunct method of verification. EDDs appear to be highly reliable in controlled settings such as the operating theatre, but should be used with greater caution in other settings.

Other methods The use of chest radiographs as a method to verify ETT placement is not recommended due to a delay between taking the film and having the film developed, as well as possible incorrect interpretation.[16] Pulse oximetry could be used in verifying ETT position if there is a perfusing rhythm. However, desaturation is usually a late indicator of deterioration in the patientâ&#x20AC;&#x2122;s condition and its use as a rapid indicator of oesophageal intubation is inadequate. Pulse oximetry requires

adequate peripheral perfusion, and is of limited utility in shocked, hypovolaemic and vasoconstricted patients.[16,19] These two methods are the least recommended ETT position verifiers. Visualisation of the ETT is implicitly mentioned in two studies as a method of verification of the ETT. However, this method might not be possible due to trauma, bleeding, vomitus, secretions or oedema and is not without potential hazards.[17,19] Direct visualisation should be the first confirmatory method, as the practitioner can assure that the tube is in the correct place, as it has incomparable speed and achievement rates when compared with some of the alternative methods of placing ETTs.[19]

Study limitations The integrative literature review was conducted using a standard method for literature review, but statistical synthesis of quantitative data was not possible. Due to the heterogeneity of study interventions and levels of evidence used, meta-analysis of data was not possible.

Conclusion Various methods have been identified to verify ETT placement in adult mechanically ventilated patients: ultrasonography, the use of centimetre scale printed on the ETT, manual cuff palpation, bilateral auscultation of chest and palpation of symmetrical chest movements, EDDs, visualisation of the ETT, use of CXR, pulse oximetry and capnography. Both ultrasonography and capnography were found to be highly sensitive and specific for verifying ETT placement, and are recommended for clinical practice. References 1. Wunsch H, Linde-Zwirble WT, Angus DC, Hartman ME, Milbrandt EB, Khan JM. The epidemiology of mechanical ventilation in the United States. Crit Care Med 2010;38(10):19471953. [http://dx.doi.org/10.1097/CCM.0b013e318ef4460] 2. Estaban A, Anzueto A, Alia I, et al. How is mechanical ventilation employed in the critical care unit? An international utilization review. Am J Respir Crit Care Med 2000;161(5):1450-1458. [http://dx.doi.org/10.1164/ajrccm.161.5.9902018] 3. Griesdale DE, Bosma TL, Kurth T, Isac G, Chittock DR. Complications of endotracheal intubation in the critically ill. Intensive Care Med 2008;34(1):1835-1842. [http://dx.doi.org/10.1007/s00134-008-1205-6] 4. Jaber S, Amraoui J, Lefrant JYl. Clinical practice and risk factors for immediate complications of endotracheal intubation in the intensive care unit: A prospective, multiple-center study. Crit Care Med 2006;34(9):2355-2361. [http://dx.doi.org/10.1097/01.CCM.0000233879.58720.87] 5. Bowles TM, Freshwater-Turner DA, Janssen DJ, Peden CJ; the RTIC Severn Group. Out-oftheatre tracheal intubation: Prospective multicenter study of clinical practice and adverse events. Br J Anesth 2011;107(5):687-692. [http://dx.doi.org/10.1093/bja/aer251] 6. Sitzwohl C, Langheinrich A, Schober A, et al. Endobronchial intubation detected by insertion depth of endotracheal tube, bilateral auscultation, or observation of chest movements: Randomised trial. BMJ 2010;41:c5943 [http://dx.doi.org/10.1136/bmj.c5943] 7. Salem MR. Verification of endotracheal tube position. Anesthesiol Clin North America 2001;19(4):813-839. 8. Angelotti T, Weiss EL, Lemmens HJM, Brock-Utne J. Verification of endotracheal tube placement by professional providers: Is a portable fiberoptic bronchoscope of value? Air Med J 2006;25(2):74-80. [http://dx.doi.org/10.1016/j.amj.2005.12.001] 9. LoBiondo-Wood G, Haber J. Nursing Research: Methods and Critical Appraisal for Evidencebased Practice. 7th ed. St Louis, USA: Mosby, 2010. 10. Muslu B, Sert H, Kaya A, et al. Use of sonography for rapid identification of oesophageal and tracheal intubations in adult patients. J Ultrasound Med 2011;30(5):671-676. 11. Werner SL, Smith CE, Goldstein JR, Jones RA, Cydulka RK. Pilot study to evaluate the accuracy of ultrasound in confirming endotracheal tube placement. Ann Emerg Med 2007;49(1):75-80. [http://dx.doi.org/10.1016/j.annemergmed.2006.07.004] 12. Sitzwohl C, Kettner S, Langheinrich A, Schoenberg C, Weinstabl C. Performance of three different bedside methods to detect inadvertent endobronchial intubation. Anesthesiology 2010;105:A532. 13. Varshney M, Sharma K, Kumar R, Varshney PG. Appropriate depth of placement of oral endotracheal tube and its possible determinants in Indian adult patients. Indian J Anaesth 2011;55(5):488-493. [http://dx.doi.org/10.4103/0019-5049.89880] 14. Simpson GD, Ross MJ, McKeown DW, Ray DC. Tracheal intubation in the critically ill: A multicentre national study of practice and complications. Br JAnaesth 2012;108(5):792-799. [http:// dx.doi.org/10.1093/bja/aer504] 15. Delorio NM. Continuous end-tidal carbon dioxide monitoring for confirmation of endotracheal tube placement is neither widely available nor consistently applied by emergency physicians. Emerg Med J 2005;22(7):490-493. [http://dx.doi.org/10.1136/emj.2004.015818] 16. Dittrich KC. Delayed recognition of oesophageal intubation. Can J Emerg Med 2002;4(1):41-44. 17. Grmec S. Comparison of three different methods to confirm tracheal tube placement in emergency intubation. Intensive Care Med 2002;28(6):701-704. [http://dx.doi.org/10.1007/s00134-002-1290-x] 18. Ledrick D, Plewa M, Casey K, Taylor J, Buderer N. Evaluation of manual cuff palpation to confirm proper endotracheal tube depth. Prehosp Disaster Med 2008;23(3):270-274. 19. Takeda T, Tanigawa K, Tanaka H, Hayashi Y, Goto E, Tanaka, K. The assessment of three methods to verify tracheal tube placement in the emergency setting. Resuscitation 2003;56(2):153-157. [http://dx.doi.org/10.1016/S0300-9572(02)00345-3] 20. Hussain N, Jaffri A, Siddiqi R. Efficacy of oesophageal detector device in verification of endotracheal tube placement. Pak Armed Forces Med J 2006;1(3):1-4.

SAJCC June 2015, Vol. 31, No. 1

ABSTRACTS

Abstracts of presentations at the congress of the Critical Care Society of Southern Africa, July 2015 S Afr J Crit Care 2014;31(3):24-28. DOI:10.7196/SAJCC.237

ORAL PRESENTATIONS Intensivists’ practices and perceptions of HIV testing in South African intensive care units D Singh University of the Witwatersrand, University of KwaZulu-Natal singhd6@ukzn.ac.za Background. Testing for HIV infection requires informed consent. Little guidance is given in the case of incapacitated patients who cannot consent but who may benefit from earlier diagnosis and directed therapy. Also, little is known regarding intensivists’ perceptions and practices concerning ethical and legal aspects of HIV testing. Objective. To ascertain intensivists’ practices and ethical per ceptions of diagnostic HIV testing in South African intensive care units (ICUs). Method. A semi-structured online questionnaire was emailed to 47 intensivists at university hospital ICUs probing HIV testing practices and policies, ethical and legal perceptions, and views regarding available legal and ethical guidelines, surrogate consent, and the disclosure of the test results. Results. The response rate was 51%. The majority of ICUs did not have a policy or protocol in place. The majority of respondents considered unconsented HIV testing ethical and beneficial in certain critically ill patients, as early diagnosis may help guide management. The majority also felt that current testing guidelines were inadequate, that surrogate consent for testing was not reliable, and that test results not be disclosed except in specific circumstances. Conclusion. The majority of South African ICUs lack a policy for HIV testing. However, physicians perceive HIV testing as beneficial in specific circumstances. Guidelines regarding testing based on patient autonomy are perceived as not in the patient’s best interests. Surrogate consent for testing is viewed as unreliable and it may be preferable to perform unconsented testing at the discretion of the intensivist.

A review of transplantation activity in South Africa (1991 - 2011) DA Thomson,* E Muller, F McCurdie, D Kahn University of Cape Town, Groote Schuur Hospital *david.thomson@uct.ac.za Background. South African organ transplantation has evolved with limited government oversight across two health systems. Objective. To analyse trends in organ transplant practices across the hospitals, provinces and healthcare systems in South Africa from 1991 to 2011. Method. Statistics submitted to the Organ Donor Foundation were analysed as to organs transplanted, living or deceased donation, paediatric or adult recipients, location and whether publicly or privately funded.

SAJCC June 2015, Vol. 31, No. 1

Results. A total of 7 280 transplants were performed across 10 public and 13 private hospitals in 5 provinces with the majority (84%) being kidney transplants. Deceased donors decreased from 4.2 per million population (pmp) in 1991 to 1.7 pmp in 2011. An increase in living related donation kept transplant numbers relatively constant. Initially, the majority of kidney (91%) and all heart and liver transplants were in the public sector. By 2011, the majority of kidney (56%), heart (96%) and liver (80%) transplants were in the private sector. Organ trafficking from 1999 to 2003 increased numbers temporarily. KwaZulu-Natal, where this practice came to light, showed a decrease in transplants from 110 in 2003 to 5 in 2010. Paediatric transplants comprised 2.6% of all transplants. Conclusion. Deceased donation has decreased markedly in South Africa. There has been a marked shift towards the private sector and increased government support is required to ensure equitable access to organ transplantation.

The effect of body position on regional distribution of ventilation and muscle activity in infants and children A Lupton-Smith,1* A Argent,1 B Morrow,1 P Riemensberger2 University of Cape Town 2 Geneva University Hospital *aluptonsmith@gmail.com

Background. Recent studies have questioned the pattern of ventilation distribution ( VD) in the spontaneously breathing paediatric population. There are no recent studies examining the effect of body position in mechanically ventilated infants and children. There are also few studies reporting respiratory muscle activity in relation to body position in this population. Objective. To determine the effect of body positions on regional VD and diaphragmatic muscle activity in mechanically ventilated children. Method. Thoracic electrical impedance tomography (EIT ) measurements and surface electromyography (sEMG) measure ments of hemidiaphragm activity (V) were taken in left and right side lying positions in mechanically ventilated infants and children. Functional EIT images were produced offline and total regional relative tidal impedance in the left and right lungs was calculated for each patient in each position. Results. Preliminary data on the first 17 patients (aged 6 months to 6 years) are presented. Eleven (65%) children demonstrated varied patterns of VD between left and right side lying. No significant differences were found between left and right lungs in left (p=0.99) and right (p=0.56) side lying. There was no significant difference in hemidiaphragmatic activity between positions. Conclusion. The paediatric pattern of ventilation during mechanical ventilation is not predictable. Diaphragm activity may not be affected by side lying positions.

Parent perception of quality of care in a South African paediatric intensive care unit (PICU) B Morrow,1* C Mol,2 M van Dijk,2 A Argent3 1 Department of Paediatrics and Child Health, University of Cape Town 2 Departments of Paediatric Surgery and Paediatrics, Erasmus MC-Sophia Children’s Hospital, The Netherlands 3 Department of Paediatrics, University of Cape Town and Red Cross War Memorial Children’s Hospital *brenda.morrow@uct.ac.za Background. The effect of physical, psychosocial and environmental factors on perceived quality of PICU care has been poorly studied in South Africa. Objective. To explore how parents of children admitted to PICU perceive the current environment and quality of care. Method. Parents/guardians of stable children admitted to PICU completed the locally adapted EMPATHIC-30 questionnaire developed by Latour et al. (2013). This questionnaire assesses parental satisfaction with provided care in the PICU context. Institutional Research Ethics Committee approval was obtained. Results. 100 questionnaires were completed, 77% by mothers. Thirty-five percent of children’s admissions were unplanned and 88% were mechanically ventilated. Teamwork: 95% reported that doctors and nurses worked well and efficiently together. Information: 91% responded that information provided about their child’s condition and management was clear. Care: 86% and 94% felt nurses and doctors were concerned about their child’s comfort; 91% felt staff were respectful. Doctors’ and nurses’ performance was rated as 9.6 (1.5) and 8.9 (1.4) (mean (standard deviation)) on a 10-point scale. Environment: 97% reported the PICU was clean; 85% felt there was sufficient space around the bedside; 78% reported acceptable noise levels. Conclusion. The experience of PICU care by parents of critically ill South African children was generally positive.

Screening critically ill patients with an adapted early mobility readiness protocol ensures safety of a therapeutic early mobility position E Conradie,1* C Fourie,2 S Hanekom1 1 Department of Physiotherapy, Stellenbosch University 2 Surgical Unit, Tygerberg Hospital *vanheerden.elmarie@gmail.com Background. The effect of the 45 semi-recumbent position on the haemodynamic stability of critically ill patients is questionable. This routine nursing position could minimise the negative effects bedrest has on the cardiac and pulmonary systems. Objective. To evaluate the feasibility of an adapted early mobility readiness protocol and the effect of a therapeutic early mobility position on two haemodynamic parameters: the mean arterial pressure (MAP) and central venous oxygen saturation (ScvO2%). Methods. Twice weekly, all patients nursed in the surgical and respiratory units were screened with the protocol. Patients who passed the protocol and inclusion criteria were tested in the baseline nursing position followed by the testing position. Haemodynamic parameters were measured at 0, 3 and 10 minutes. Results. We screened 138 patients. Eleven patients passed, male/ female (9/2) with a median age of 47 (20 - 67) years. The mean MAP

(95% confidence interval (CI)) increased from 95.87 (91.72 - 100.03) mmHg to 97.86 (94.16 - 101.57) mmHg and the mean ScvO2% (95% CI) values increased from 81.43 (79.43 - 83.42)% to 82.06 (79.43 - 84.20)% when the baseline nursing position was followed by the testing position. Conclusion. Patients who passed the protocol were not physiologically challenged when placed in the testing position. The protocol can be used by clinicians to identify patients suitable for the testing position. Work is needed to investigate the outcome of patients nursed in this position.

Plasma glutamine levels in adult intensive care unit patients A Nienaber,1* RC Dolman,1 AE Van Graan,1 R Blaauw2 1 Centre of Excellence for Nutrition, North-West University, Potchefstroom Campus 2 Division of Human Nutrition, Stellenbosch University *aristahefer@yahoo.com Background. Glutamine (Gln) deficiency is an independent predictor of mortality in intensive care unit (ICU) patients and its supplementation is recommended for proven outcome benefits. Recent data suggest that early Gln supplementation increases mortality in certain patient groups. Objective. To investigate plasma Gln levels of adult ICU patients and to determine relationships between Gln levels, gender, diagnosis and selected inflammatory markers. Methods. This cross-sectional study included 60 mixed adult ICU patients, paired with 60 healthy controls (CG). Gln levels of patients were compared to that of CG and the relationship between Gln levels and Interleukin-6 (IL-6) or C-reactive protein (CRP) was examined. A non-parametric ROC curve was computed to determine the CRP concentration cut-off above which Gln becomes deficient. Results. ICU patients had significantly lower Gln levels than CG (496 mol/L v. 718 mol/L, p<0.0001). Of the patients, 38.3% (n=23) had deficient (<420 mol/L) and 6.7% (n=4) had supra-normal Gln levels (>930 mol/L). Gln was inversely associated with CRP (r=–0.44, p<0.05) and IL-6 (r=–0.23, p=0.08) levels. A CRP cut-off value of 95.5 mg/L was determined above which Gln levels became deficient. Conclusion. ICU patients had lower Gln levels compared with healthy controls; not all patients were Gln deficient and some presented with supra-normal levels. This highlights the importance of selecting patients for Gln supplementation. Since Gln levels were inversely associated with CRP, the latter might be useful as a proxy marker for Gln status.

Hypoxaemia on arrival in a multidisciplinary intensive care unit K de Vasconcellos,* D Singh, DL Skinner King Edward VIII Hospital and University of KwaZulu-Natal *kimdevasconcellos@gmail.com Background. Transport of the critically ill patient poses the risk of numerous complications. Hypoxaemia is one such serious adverse event and is associated with potential morbidity and mortality. It is however potentially preventable. Objective. To determine the incidence of hypoxaemia on arrival in a tertiary multidisciplinary intensive care unit (ICU) and to identify risk factors for this complication. Method. Retrospective observational study using data collected during a prospective transport audit at King Edward VIII Hospital from May 2013 to February 2014.

SAJCC June 2015, Vol. 31, No. 1

Results. Hypoxaemia occurred in 23/148 (15.5%) admissions. Statistically significant risk factors for hypoxaemia on univariate analysis (p<0.05) included lack of SpO 2 monitoring, transfer by an intern as opposed to other medical/paramedical staff, and transfer from internal medicine. Use of neuromuscular blockers and transfer from theatre were protective. Binary logistic regression analysis revealed lack of SpO2 monitoring to be the only significant independent predictor of hypoxaemia (OR=5.59 (95% confidence interval 1.33 - 23.57), p=0.019). Conclusion. Hypoxaemia is common on admission to ICU and could potentially be prevented by simple interventions such as appropriate transport monitoring.

The physiotherapy management of thoracotomy patients: A survey of current practice in Gauteng Schwellnus L,* Roos R, Naidoo V University of the Witwatersrand *grassman.liezel@gmail.com Background. Physiotherapy is an essential clinical component in the management of patients after thoracic surgery to prevent respiratory complications and improve mobility. To date, published literature is scarce regarding physiotherapy practice in this area. Objective. To establish the physiotherapy modalities used in the management of thoracotomy patients during all phases of recovery. Method. A cross-sectional study of 1 389 physiotherapists registered with the South African Society of Physiotherapy in Gauteng was conducted. A self-administered questionnaire was distributed electronically. The data collection period was 2 months and data were analysed with descriptive statistics. Results. Three hundred and twenty-three physiotherapists (23.3%) responded. Preoperative physiotherapy management was deter mined by the patients’ risk profiles and consisted of information and respiratory techniques. Prophylactic postoperative management was high. The modalities used most commonly were respiratory techniques, e.g. deep breathing exercises (97.6%; n=83) and exercise interventions, e.g. early mobilisation (95.3%; n=81), trunk (85.9; n=73) and upper limb mobility exercises (91.8%; n=78). Pain-reducing modality use was less common, e.g. transcutaneous electrical ner ve stimulation (12.9%; n=11). Post-hospital physiotherapy was uncommon (32.6%; n=46). Conclusion. Physiotherapists in Gauteng use techniques to prevent and manage postoperative pulmonary complications in patients who undergo thoracic surgery. Education is needed regarding the importance of effective pain management by physiotherapists for these patients both during and after hospital stay.

POSTER PRESENTATIONS Use of proton pump inhibitors in the ICUs of three academic hospitals in Johannesburg N Biyase,* H Perrie, J Scribante, S Chetty Department of Anaesthesia, University of the Witwatersrand *nana.biyase@yahoo.com Background. Stress ulcer profylaxia (SUP) is an important part of management of critically ill patients in intensive care

SAJCC June 2015, Vol. 31, No. 1

units (ICUs). However, inappropriate use of these drugs has important clinical implications such as ventilator-associated pneumonia and gastrointestinal tract infections. The overuse of proton pump inhibitors (PPIs) SUP is a rapidly growing problem internationally. Objectives. To describe the use of SUP in ICUs and to compare the appropriate use of PPIs versus the inappropriate use according to the risk factors of the patient as per American Society of Hospital Pharmacists guidelines. Method. A retrospective, descriptive, contextual study design was used. A 3-month audit of ICU charts of adult patients admitted to ICUs at Chris Hani Baragwanath Academic Hospital, Charlotte Maxeke Johannesburg Academic Hospital and Helen Joseph Hospital that fulfilled the inclusion criteria. Results. A total of 174 patients were included in the study. Of these patients, 156 were started on SUP, 60.9% (95) of them were appropriately started on SUP and 39.1% (61) were inappropriately on SUP. There was overuse of SUP of 39.1% (61). The number of patients who qualified for SUP came to 113. In that group only 28.3% (32) were on PPIs and the remainder of the patients 71.7% (81) were either on other agents or were not started on SUP, reflecting an underuse of PPIs of 71.7% (81). Conclusion. Our study found inappropriate overuse of SUP but underuse of PPIs where they should have been used.

Outcome and severity of surgical patients admitted to a non-tertiary multidisciplinary critical care unit O Swart,* R Duvenage Worcester Provincial Hospital *oostewalt.swart@westerncape.gov.za Background. Demand for critical care services exceeds avail ability. Limited published public health sector critical care unit (CCU) data, outside of academic institutions, are available. Worcester Provincial Hospital’s ( WPH) 5-bed CCU is classified as a high-care unit (HCU) and services a rural population of approximately 600 000 in the Cape Winelands East and Overberg district of the Western Cape. Objective. To describe the severity and in-hospital all-cause mortality of surgical patients admitted to the WPH CCU. Method. A descriptive study of surgical patients admitted to WPH CCU collected prospectively from January to December 2014. Results. The WPH CCU had 114 (8 re -admissions) surgical admissions out of a total of 610 during 2014. The mean (standard deviation (SD)) age and median length of stay were 46.0 (SD 18.7) years and 2 (range 0.5 - 13.5) days, respectively. The mortality rate was 17.0% (18/106) with a mean Apache II score of 11.1 (SD 6.6). Ventilatory support was required in 68.4% (78/114) of admissions with 156 ventilation and 67 continuous positive airway pressure days. Ventilation and inotropic support (total of 70 inotrope days) was required in 28.9% (33/114) of admissions, indicating multiorgan failure. Only 18.9% (20/106) of patients were referred to tertiary CCUs. Conclusion. Disease severity and mortality were similar to the only non-tertiary, although non-surgical, CCU data published. Almost a third of the CCU admissions had multi-organ failure that is deemed to be beyond the scope of an HCU.

A comparison of excess fluid to be removed in haemodialysis patients, as estimated by haemodialysis staff versus multiple frequency bioelectrical impedance analysis J Downs,* F Veldman, S Kassier Dietetics Department, King Edward VIII Hospital and Discipline of Dietetics & Human Nutrition, School of Agriculture, Earth & Environmental Sciences, University of KwaZulu-Natal *jane.downs3@gmail.com Background. Currently, in most haemodialysis (HD) units in South Africa, the excess fluid to be removed in HD is estimated by the staff who compare the previous post-HD versus the subsequent pre-HD body weight (wt). Bioelectrical impedance analysis (BIA) has been shown to have clinical value and may potentially protect the HD patients from risks associated with under- and over-hydration. Objective. To compare excess fluid to be removed in HD as estimated by the staff versus the volume measured per multiple frequency BIA. Methods. A prospective, non-randomised observational study was conducted. Repeated measures of 24 BIA pre- and post-HD measurements were taken over 3 months on 20 chronic HD subjects (50% male; ages 21 - 63 years) at the King Edward VIII Hospital HD unit. Results. There was a significant difference between the excess fluid measured per BIA v. the staff estimate. Subjects with a body fat (BF) >35% and a BMI >28 kg/m2, had a negative 3rd water space v. those with lower BF and BMIs. Conclusion. Excess fluid measured per BIA v. staff estimates differed. Previous studies have shown HD patients with a higher BF and BMI have a lower mortality risk; we postulate that the negative 3rd water space in overweight subjects is due to greater fluid and sodium lost in perspiration.

Outcome of children admitted to a combined paediatric/neonatal ICU in a low to middle income country D Ballot,* V Davies, P Cooper University of the Witwatersrand *daynia.ballot@wits.ac.za Background. In South Africa, limited resources result in a lack of sufficient ICU staff and facilities for patient demand. Doctors use the ethical principle of distributive justice to ration scarce intensive care unit (ICU) facilities. The mortality rate in paediatric ICU (PICU) is variable and reports range between 5% and 14.6%. Only 19.6% of all ICU beds in both the private and public sectors in South Africa were dedicated to paediatric and neonatal intensive care. Charlotte Maxeke Johannesburg Academic hospital (CMJAH) has a combined paediatric and neonatal ICU with 14 ventilator beds. Objective. To establish the short-term outcome of children admitted to CMJAH paediatric and neonatal ICU between 1 January 2013 and 31 December 2014. Methods. This study is a retrospective descriptive review of children admitted to the CMJAH PICU. Information was obtained from the CMJAH PICU and neonatal databases at CMJAH. Results. A total of 1 073 patients were admitted to PICU at CMJAH during the study period: 472 paediatric patients, 245 very low birth weight (VLBW) neonates and 356 other neonates. Over one-third of

the patients were surgical (414/1 073; 38.5%). The overall mortality rate was 24.3% (261/1 073): 100 (40.8%) for VLBW, 91 (25.5%) for other neonates and 70 (14.8%) for paediatric patients Conclusions. There is a high demand for PICU beds. Separate paediatric and neonatal ICUs are justified. Mortality of PICU patients is within the reported range.

Predicting mortality rates: Hospital Standardised Mortality Ratio v. APACHE IV R Toua,* J De Kock University of Cape Town and Mediclinic *rene.toua@mediclinic.co.za Background. Mortality is widely used as a qualitative measure. However, actual mortality rates can’t be directly compared and to accurately and simply predict the risk of death is a big challenge. Objective. To correlate predicted mortality as calculated with an administrative model (Hospital Standardised Mortality Ratio) to a physiological model (APACHE IV): combined cohort and samples stratified by physiological prediction level (<10% predicted mortality, 10 - 50% predicted mortality or >50% predicted mortality). Method. Cross-sectional: 40 private healthcare group hospitals (68 critical care units) . One-way analysis was done. A total of 47 982 critical and high-dependency patients were scored from 1 June 2013 to 31 July 2014. 1 921 records (0.4%) were excluded due to missing values, duplicate records and values not within parameters (n=46 061). Results. Correlation was moderate for the combined cohort (Pearson’s correlation index 0.62 (95% confidence interval (CI) 0.62 - 0.63), R-squared 0.38); very good for the <10% stratum (Pearson’s correlation index 0.88, R-squared 0.78 (95% CI 0.878 - 0.882)); good for the 10 - 50% predicted mortality rates (Pearson’s correlation index 0.78, R-squared 0.61 (95% CI 0.77 - 0.79)) and no correlation for the >50% predicted mortality stratum (Pearson’s correlation index 0.09, R-squared 0.01 (95% CI 0.03 - 0.15)). Conclusion. The administrative predictive model is not suitable for predicting mortality in the highest stratum.

Profile of ICU bed requests at Helen Joseph Hospital H Hurri,* S Chetty, J Scribante, H Perrie University of the Witwatersrand *hurrih@gmail.com Background. Intensive care unit (ICU) beds are a scarce resource at Helen Joseph Hospital (HJH). A limited number of beds serve a population with a large burden of disease. Medical practitioners request ICU beds for patients they deem in need of ICU; however, the decision to admit into ICU remains with the ICU consultant on call, consultants being from different disciplines. No formal triage such as APACHE or SAPS scoring or admission guidelines is currently in place. Objective. To compile a profile of the ICU admission requests at HJH. Method. A contextual, prospective, descriptive research design was followed in this study. Data were collected during one winter and one summer month in 2012. Results. A total of 139 patients were included. The median age was 44 years. The majority of patients (79%) were under the age of 60 years. The overall admission rate was 35.25% and the most common reason for admission was mechanical ventilation. Reasons for refusal were 41% assessed as too ill, 30% assessed as too well and 29% were

SAJCC June 2015, Vol. 31, No. 1

refused due to a lack of resources. Patients admitted to the ICU had a 77.55% survival rate. The relationship between ICU admission and 30-day outcome was statistically significant. Conclusion. The lack of resources is problematic at HJH ICU. Survival rates correlate with international trends and triage methods appear to be effective.

Are weight estimation methods applicable for rural African children? A comparative study Z Nazo* Walter Sisulu University *zandisile.nazo@gmail.com Background. Accurate paediatric weight estimation is essential in the management of critically ill children as it is impractical to weigh these patients. The Advanced Paedaitric Life Support (APLS) formula has been used for weight estimation globally. Recent studies done in developed countries report that this formula significantly under-estimates children’s weights, which may result in medical errors. Objective. To investigate the accuracy of APLS and Luscombe and Owen formulae in estimating weight of rural African children. To compare the performance of these formulae in this population. Method. A prospective, observational and cross-sectional study involving a convenient population of children aged 1 - 10 years was conducted from September 2012 to March 2013. These children were attending a private paediatric practice and a public hospital paediatric outpatient department. They were weighed and weights recorded. Weight difference between the measured weight and the calculated weight for APLS and Luscombe and Owen formulae was the main outcome measure. The Bland-Altman method was used for detecting accuracy and precision of the formulae. Results. A sample of 314 males and 252 females was analysed. There was no significant (p>0.05) association between gender, actual weights and estimated weights. The mean values of actual weight of the private participants were significantly higher (p<0.05) than those of public participants. Predictive accuracy was superior for the APLS than for the Luscombe formula. Using the mean percentage differences, the APLS again showed a superior advantage in predicting weight over the Luscombe formula. The estimated weight by the APLS formula within 10% error was higher than in the Luscombe formula. All results were statistically significant. Precision was more accurate with the APLS formula than with the Luscombe. Conclusion. The results showed that the APLS formula has a superior advantage over the Luscombe and Owen formula at estimating weight in this population of children. This is in agreement with the Western Cape study by Geduld et al.

Patient perceptions of ICU care: A scoping review M van Ness, F Karachi, S Hanekom* Physiotherapy Interdisciplinary Health Sciences, Stellenbosch University *sdh@sun.ac.za Background. Physiotherapy practice in intensive care unit (ICU)s is changing. Early mobilisation programmes are included and prioritised. Methods and measures to assess physiotherapy effectiveness in the ICU have often been geared to physiological data. It is unclear whether patients’ perspective and satisfaction with care in ICU have been investigated.

SAJCC June 2015, Vol. 31, No. 1

Method. A scoping review was undertaken with the aim of determining how patient perception and satisfaction with critical care is measured. Seven databases were searched using the following keywords in various combinations: physiotherapy or physical therapy, patient satisfaction, perception or patient perception, patient experience, intensive care unit or ICU, critical care, hospitalised adult population, hospital, measurements, measuring and outcome measure. Results. 1 626 articles were independently screened by two reviewers at title, abstract and full text level respectively. The final review included 26 articles. Only two of the studies were conducted in Africa, compared with ten in Europe and six in Northern America, respectively. Nine of the included articles investigated a particular service such as nursing care, emergency care and physiotherapy with regards to patient perception and satisfaction. Only one article, published in 2008, investigated patient perception and satisfaction in physiotherapy. Various outcome measures were identified in this review that measure perception and/or satisfaction. However, there is currently no validated and reliable instrument to assess patient satisfaction with care in the ICU. Conclusion. A gap in the literature was identified for patient perceptions regarding physiotherapy care in the ICU. The results will be used to inform the planning of a primary qualitative study. Knowing and understanding the patients’ perception and satisfaction with care, ensures the professional development in the critical care field, and improving the quality of care.

Articulating the nature of clinical specialist nurse practice J Bell,* D van Rooyen, P Jordan Stellenbosch University, Nelson Mandela Metropolitan University *jbell@sun.ac.za Background. People sharing stories of their encounters in critical care environments revealed that some nurses were perceived to have distinctive qualities that influenced an encounter in a positive way. Encounters with these nurses were described as being different and better, with the qualities that characterised anecdotes of ‘different and better’ nursing not experienced by all critical care nurses. Objective. The purpose of this study was to articulate an understanding of the qualities that people who engage with critical care nurses recognise as ‘different and better’ to the norm of nursing practice they encounter in this discipline. Method. Constructivist grounded theory methodology was used. Participants were drawn from patients’ significant others, nurses and medical colleagues. Data generation began with participants contributing through in-depth unstructured individual interviews and creating a naïve sketch. A focused literature review contributed to a final study sample of 74 data items. Data generation was completed through grounded theory data analysis method processes. Results. An inductively derived substantive grounded theory named Being at Ease was constructed from the data. The core concern in recognising different and better nursing practice emerged as ‘being at ease’. This is underpinned by four categories,: ‘knowing self’, ‘skilled being’, ‘connecting with intention’ and ‘anchoring’. Conclusion. Being at Ease adds to our practice narrative through this explanation of how tacit qualities of ‘different and better’ nursing are located as discrete elements within the complex nature of specialist clinical practice.

CASE REPORT

Postoperative internal iliac artery embolisation as salvage therapy for bleeding in an HIV-positive patient with giant cell tumour of bone T van den Heever,1 BSc, MB ChB; Dipl (Anaes), MMed (Sc Crit Care), Dipl (Transfusion Med); C L Barrett,2 MB ChB, Dipl (Transfusion Med); M J Webb,2 MB ChB; MMed (Int Med); Cert (Clin Haematology); M G L Spruyt,1 MB ChB, MMed (Surg), MMed (Crit Care Med); C J Louw,2 MB ChB, MMed (Int Med); Cert (Clin Haematology), PhD (Med Ed) 1 2

Department of Critical Care, Faculty of Health Sciences, University of the Free State, Bloemfontein, South Africa Division of Clinical Haematology, Department of Internal Medicine, Faculty of Health Sciences, University of the Free State, Bloemfontein, South Africa

Corresponding author: T van den Heever (theavdheever@gmail.com)

Giant cell tumour of bone (GCTB) is a highly vascular tumour, sporadically complicated by massive bleeding during surgery. We report a rare case of GCTB in an HIV-positive patient who suffered massive blood loss intra- and postoperatively. The patient was a 46-year-old HIV-positive female with symptoms and signs of a pelvic mass, and ultrasound evidence of an ovarian mass. Surgery was performed, and a highly vascular retroperitoneal mass originating from her sacrum was identified. Massive blood loss occurred, which required aggressive resuscitation and transfusion of blood products. Damage control surgery was performed, and bleeding was ultimately only controlled postoperatively using bilateral internal iliac artery radiological embolisation. The patient suffered acute kidney injury, which was multifactorial in aetiology, which recovered within 6 days. She was discharged from ICU in a stable condition 7 days postoperatively. S Afr J Crit Care 2015;31(1):30-31. DOI:10.7196/SAJCC.177

Giant cell tumour of bone (GCTB) is a benign but locally aggressive skeletal neoplasm. It is relatively rare, and mostly affects young adults in the second decade of life. A slight female predominance with a female to male ratio of 3:2 is typical of GCTB. It may undergo malignant transformation. [1,2] Only 4% of patients with GCTB present with the tumour affecting the pelvis.[3] GCTB is known as a highly vascular tumour.[4] Consequently, intraoperative blood loss may be considerable. Preoperative embolisation has been proven successful in reducing intraoperative blood loss.[4] For embolisation to be successful there should be a normal intact clotting system. The aim of embolisation is to devascularise the tumour in order to reduce the amount of blood loss or to serve as a curative procedure, resulting in partial or complete necrosis of the tumour.[2] Usually a 4 or 5 French diagnostic catheter or a microcatheter is used with a gelatine sponge.[2] The gelatine sponge is used only as a temporary measure, as this is part of a staged procedure prior to surgery. Embolisation should be combined with intralesional curettage.[2] Bone cement (polymethyl methacrylate, PMMA) is recommended to fill the area of the bone that is removed.[2] Only two cases of GCTB have ever been reported in HIV-positive patients, and neither of these cases required a massive transfusion of blood products.[1,5] Approval to report this case was obtained from the Ethics Committee of the Faculty of Health Sciences, University of the Free State, the relevant heads of department and the Head of Clinical Services of the Universitas Academic Hospital Complex in Bloemfontein, South Africa.

SAJCC June 2015, Vol. 31, No. 1

Case report A 46-year-old female patient, who was known to be HIV-positive and treated with combination antiretroviral therapy (cART), was admitted to our intensive care unit (ICU) for postoperative care following a massive blood transfusion in theatre. She had resented to her gynaecologist with a pelvic mass. Abdomino-pelvic ultraÂ sound examination suggested an ovarian mass. IntraÂoperatively, it was discovered that the patient had a highly vascular, malignantappearing retroperitoneal mass involving the sacrum, and not the ovary as expected. Surgical incision using a Ligasure (Covidian, South Africa) curved small-jaw open sealer/divider was complicated by profuse bleeding after approximately two-thirds of the tumour was resected. Attempts to surgically control the bleeding failed, and her abdomen was packed with adrenalin-soaked swabs. Coagulation tests were suggestive of either post-transfusional coagulopathy or disseminated intravascular coagulation. No abdominal pressures were measured, but may have been useful to confirm the diagnosis of abdominal compartment syndrome. The patient received a massive transfusion (16 units of red cell concentrate, 15 units of fresh frozen plasma and 2 units of platelets; cryoprecipitate was not available) intraoperatively, and in addition, tranexamic acid, conjugated equine oestrogen (Premarin) and desmopressin were administered. Hypocalcaemia was corrected. Cell salvage could not be used in this patient to reduce allogeneic blood usage owing to the suspected malignancy. Every effort was made to maintain normothermia, normotension and a physiological pH; despite this the patient became hypothermic, hypotensive and acidotic, and the coagulopathy did not resolve.

As she continued to bleed postoperatively, she received another massive transfusion (7 units of red cell concentrate, 1 600 mL freezedried plasma, 10 units of cryoprecipitate and 1 unit of platelets, as well as 2 000 U of prothrombin complex concentrate and 1 000 U of factor VIII). Despite all the blood products transfused, the patient did not develop transfusion-associated circulatory overload (TACO) or transfusion-related acute lung injury (TRALI). As the bleeding continued, the patient was taken to inter ventional radiology on day 1 postoperatively. A 5 French catheter was inserted, and embolisation of both internal iliac arteries as well as the tumour vessels was performed using an absorbable gelatine compressed sponge (Gelfoam, Pfizer, USA). Bleeding ceased completely and no further blood products were required following embolisation. The patient subsequently developed acute kidney injury with a peak serum creatinine of 380 µmol/L on day 1 following embolisation. This was attributed to hypotension and contrast nephropathy. The serum creatinine returned to a normal baseline within 6 days. There was no clinical or biochemical evidence of tumour lysis syndrome. Histology of the resected mass revealed a GCTB arising from the sacrum. The patient was discharged from the ICU in a stable condition 7 days postoperatively.

Discussion This case demonstrates the value of bilateral internal iliac artery embolisation when control of massive bleeding from a highly vascular sacral tumour is required. In our case, it was used postoperatively, but it has been advocated as a measure to reduce tumour bulk and to reduce perioperative blood loss if used pre operatively.[6] Preoperative embolisation has been well described in hypervascular bone tumours.[7] Embolisation of the internal iliac arteries may be complicated by ischaemia of the pelvic organs, as well as the spine, buttock and thigh. This may present as gluteal necrosis, paraplegia and impotence, bowel ischaemia, and bladder and perineal necrosis.[8] Postembolisation syndrome, which may include symptoms of pelvic pain and cramping, nausea, vomiting, fever, fatigue, myalgias and malaise with leucocytosis, usually presents within 48 hours and is self-limiting, and typically resolves within 7 days.[9] Damage control surgery comprises three stages, namely: (i) control of bleeding and temporary closure of the abdomen; (ii) aggressive resuscitation while attempting to correct the acidosis, hypothermia and coagulopathy; and (iii) return to the operating

theatre for more definitive treatment and closure of the abdominal wall once the patient has been stabilised.[10-12] The patient developed an acute kidney injury, for which she had many risk factors. Her baseline renal function was normal. In spite of the massive blood loss, initial hypovolaemia and hypotension, she had no progressive renal injury, with renal function recovering within 6 days. We believe that this is due to adequate fluid resus citation that she received in the intra- and postoperative period, as well as avoidance of further exposure to nephrotoxic agents.

Conclusion We report a unique coincidental case of an HIV-positive patient with a GCTB complicated by massive intra- and postoperative blood loss and dilutional coagulopathy. Postoperative embolisation of the internal iliac arteries was a life-saving intervention in this patient with uncontrolled blood loss from a highly vascular tumour bed. The patient did not suffer postembolisation syndrome or ischaemic complications from the embolisation. Radiological embolisation should be considered preoperatively in patients with highly vascular bone tumours where massive blood loss is anticipated, but may be considered postoperatively if it was not performed prior to surgery. Acknowledgements. Dr Daleen Struwig, medical writer, Faculty of Health Sciences, University of the Free State, for technical and editorial preparation of the manuscript. References 1. Ares O, Conesa X, Seijas R, Huguet P, González R, Fernández N. [Giant cell tumour of bone in a patient with HIV infection]. Enferm Infecc Microb Clin 2010;28(6):396-397. [http://dx.doi. org/10.1016/j.eimc.2009.09.003] 2. Thomas DM, Desai J. Giant cell tumor of bone. http://www.uptodate.com/patients/content/ topic.do?topicKey=~ahzlhQH65OZROGa (accessed 12 February 2012). 3. Turcotte RE. Giant cell tumour of bone. Orthoped Clin North Am 2006;37(1):35-51. [http:// dx.doi.org/10.1016/j.ocl.2005.08.005] 4. Yasko AW. Giant cell tumour of bone. Curr Oncol Rep 2002;4(6):520-526. [http://dx.doi. org/10.1007/s11912-002-0067-2] 5. List AF. Metastatic giant-cell bone tumour in a man positive for HIV. N Engl J Med 1988;318:517. 6. Owen RJT. Embolization of musculoskeletal tumours. Radiol Clin North Am 2008;46(3):535543. [http://dx.doi.org/10.1016/j.rcl.2008.02.002] 7. Nair S, Gobin YP, Leng LZ, et al. Preoperative embolization of hypervascular thoracic, lumbar, and sacral spinalcoloumn tumors: Technique and outcomes for a single center. Interv Neuroradiol 2013;19(3):377-385. 8. Andriole GL, Sugarbaker PH. Perineal and bladder necrosis following bilateral internal iliac artery ligation: Report of a case. Dis Colon Rectum 1985;28(3):183-184. [http://dx.doi. org/10.1007/BF02554240] 9. Society of Obstetricians and Gynaecologists of Canada. SOGC clinical practice guidelines. Uterine fibroid embolization (UFE). Number 150, October 2004. Int J Gynaecol Obstet 2005;89(3):305-318. 10. Open Abdomen Advisory Panel, Campbell A, Chang M, et al. Management of the open abdomen: From initial operation to definitve closure. Am Surg 2009;75(11 Suppl):S1-S22. 11. Barker DE, Green JM, Maxwell RA, et al. Experiences with vacuum-pack temporary abdominal wound closure in 258 trauma and general and vascular surgical patients. J Am Coll Surg 2007;204(5):784-792. [http://dx.doi.org/10.1016/j.jamcollsurg.2006.12.039] 12. Aydin C, Aytekin FO, Yenisey C, et al. The effect of different temporary abdominal closure techniques on fascial wound healing and postoperative adhesions in experimental secondary peritonitis. Langenbecks Arch Surg 2008;393(1):67-73.

SAJCC June 2015, Vol. 31, No. 1

OBITUARY

Max Klein

Max Klein died unexpectedly on 27 January 2015 while riding his bicycle with friends near Stellenbosch. Max grew up in the country districts, and came to Cape Town as a schoolboy at South African College Schools (SACS). He went to medical school at the University of Cape Town, where his experience included a bout of meningococcal meningitis, which profoundly affected his view of patient care. Early in his medical career he was a Wellcome Research Fellow in the neonatal intensive care unit (ICU) at Groote Schuur Hospital. Together with Vincent Harrison, Boet Heese and Atties Malan, he was involved in early research on neonatal ventilation, and was part of a remarkable study elucidating the significance of grunting in newborn infants with hyaline membrane disease.[1] Vincent Harrison has written of the remarkable originality of thought that Max displayed during this time and throughout his career. Subsequently, he trained in paediatrics

SAJCC June 2015, Vol. 31, No. 1

at the Red Cross War Memorial Children’s Hospital (RCWMCH), before going on to train in respiratory medicine with Attie de Kock at Stellenbosch University (Karl Bremer Hospital). In 1972, he went to the Cardiovascular Research Unit and the Department of Paediatrics at the University of California Medical Centre in San Francisco, USA, on a Lilly International Fellowship, before returning in 1974 to RCWMCH as head of the paediatric ICU and respiratory service. He remained in that role until stepping down as head of the paediatric ICU in 1999. Max played a significant role in the development of paediatric intensive care in South Africa (SA) and beyond. Many senior paediatric intensivists across the world received their exposure to paediatric intensive care under his tuition. Something of his role in paediatric critical care was recognised when he was presented with the gold medal of the World Federation of Pediatric Intensive and Critical Care Societies in 2000 in Montreal. He was also awarded the President’s Award of the Critical Care Society of Southern Africa in recognition of his noteworthy contribution to paediatric (and adult) intensive care in SA. He was a remarkable clinical physiologist, with unique insight into a variety of conditions including croup (for which he developed a scoring system that has been used extensively in SA), oropharyngeal obstruction (for which he developed the system of continuous insufflation of the pharynx), respiratory failure (for which he developed innovative approaches to ventilation), shock (with a clear and lucid approach to diagnosis and management of hypovolaemia), among many other conditions. There are many children who benefitted from his clear insight and innovative thought, while many clinicians were deeply affected by his attention to detail, clinical acumen

and understanding of pathophysiological processes. Perhaps most importantly, he cared deeply for the welfare of his patients and was not prepared to comply with dicriminatory practices. Under his leadership, the tracheostomy and home ventilation service at the RCWMCH developed into a remarkably successful prototype of effective home care of children with complex conditions. He played a significant role in the development of the Critical Care Society of Southern Africa: organising conferences and delivering presentations; participating in council and annual general meetings; and joining in with gusto at celebrations and parties. He played a similar role within the pulmonology community. After his retirement from intensive care, Max continued to head the Department of Paediatric’s pulmonology unit at RCWMCH until his retirement. Thereafter he continued to teach and consult in paediatric respiratory medicine across the country. He was appointed as the first Extraordinary Professor in the Department of Paediatrics and Child Health at the University of Pretoria. Max was not always easy to get on with, and he disagreed vehemently with quite a few people. But he will be remembered for his clinical skills, his commitment to patient and family care and for his remarkable contributions as a teacher and innovator in paediatric critical care and pulmonology. Reference 1. Harrison VC, Heese H de V, Klein M. The significance of grunting in hyaline membrane disease. Pediatrics 1968;41(3):549-559.

A C Argent Department of Paediatrics and Child Health, University of Cape Town and Paediatric Intensive Care Unit, Red Cross War Memorial Children’s Hospital, Cape Town, South Africa

RETHINK

Moderate to severe community acquired infections1 was the most active antibiotic tested overall (SMART 2004-2009) 2 Benefits of a once-daily empiric monotherapy 3,4 Best suited for treatment of community-acquired complicated intra-abdominal infections in South Africa2 Cost-effectiveness Patient convenience and comfort A lower risk of medication errors vs. combination therapy or agents that require multiple doses

Right Spectrum. Smart Choice. SMART = The Study for Monitoring Antimicrobial Resistance Trends References: 1. Brink AJ, Feldman C, Grolman DC, et al. Appropriate Use of the Carbapenems. S Afr Med J 2004;94(10):857- 861. 2. Brink A, Botha R, Poswa X, et al. Antimicrobial susceptibility of Gram-negative pathogens isolated from patients with complicated intra-abdominal infections in South African hospitals (SMART study 2004-2009): impact of the new carbapenem breakpoints. Surg Inf 2012;13(1):1-7. 3. Namias N, Solomkin J, Jensen E, et al. Randomized, Mulricenter, Double-Blind Study of Efficacy, Safety, and Tolerability of Intravenous Ertapenem versus Piperacillin/Tazobactam in Treatment of Complicated Intra-Abdominal Infections in Hospitalized Adults. Surgical Infections 2007; 8(1):15-28. 4. INVANZ® Sterile Powder for Injection approved package Insert – 10 October 2008. S4 INVANZ® Sterile Powder for Injection. Reg. No. 37/20.1.1/0424. Each vial contains ertapenem sodium equivalent to 1 g of ertapenem free acid

For full prescribing information refer to the package insert approved by the Medicines Regulatory Authority.

MSD (Pty) Ltd (Reg. No. 1996/003791/07), Private Bag 3, Halfway House 1685. Tel: (011) 655-3000. www.msd.co.za. Copyright © 2015 Merck Sharp & Dohme Corp., a subsidiary of Merck & Co., Inc., Whitehouse Station, NJ, USA. All rights reserved. AINF-1099836-0000

14944