SAJCC SAJCC Vol 31, No 2 (2015)

SAJCC SOUTHERN AFRICAN JOURNAL OF CRITICAL CARE

November 2015 Vol. 31 No. 2

• Ethical research in the ICU • Moral distress of ICU nurses • Nutrition practices in SA ICUs • Colistin for VAP • Recognising stroke

THE OFFICIAL JOURNAL OF THE CRITICAL CARE SOCIETY OF SOUTHERN AFRICA

SAJCC THE SOUTHERN AFRICAN JOURNAL OF CRITICAL CARE

The Official Journal of the Critical Care Society of Southern Africa November 2015 Vol. 31 No. 2

CONTENTS EDITORIAL 34

Ethical considerations for critical care research B Morrow

ARTICLES 36

Moral distress experienced by intensive care nurses G C Langley, L Kisorio, S Schmollgruber

42 Nutrition support practices in South African ICUs: Results from a nationwide pilot survey

L T Hill

51 Comparison of the efficacy of colistin monotherapy and colistin combination therapies in the treatment of nosocomial pneumonia and ventilator-associated pneumonia caused by Acinetobacter baumannii I Kara, F Yildirim, B Bilaloglu, D Karamanlioglu, E Kayacan, M Dizbay, M Turkoglu, G Aygencel

58 The accuracy of Johannesburg-based ambulance personnel in identifying stroke

D Nel, W Stassen

CASE REPORT

62 Atrial myxoma-related embolism resulting in acute limb ischaemia in a critical care patient

M Knight, R D Wise

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: 072 635 9825 Email: publishing@hmpg.co.za HMPG CEO AND PUBLISHER Hannah Kikaya HMPG EDITOR-IN-CHIEF Janet Seggie CONSULTING EDITOR J P de V van Niekerk EXECUTIVE EDITOR Bridget Farham MANAGING EDITOR Ingrid Nye TECHNICAL EDITORS Emma Buchanan Paula van der Bijl PRODUCTION MANAGER Emma Jane Couzens DTP & DESIGN Carl Sampson HEAD OF SALES AND MARKETING Diane Smith tel. (012) 481-2069 ISSN 1562-8264

BOOK REVIEW

64 Cardiopulmonary Physiotherapy in Trauma: An Evidence-based Approach

A Lupton-Smith

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 Department of Higher Education and Training.

This Journal is also published online at www.sajcc.org.za Critical Care Society of Southern Africa Contact details: Administrative Secretary, P O Box 521, Melville, 2109 Email: criticalcare@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

Ethical considerations for critical care research Ethical lapses are almost never a case of bad people, doing bad things, for no good reason. More often they are good people, doing bad things, for good reasons.

(Marcia Angell, previous Editor-in-Chief (1988 - 2000), New England Journal of Medicine.)

Critical illness carries high morbidity and mortality worldwide, with a disproportionate burden of critical illness in low- and middle-income countries, where access to intensive care is particularly limited.[1] With­ out research in the intensive care unit (ICU) population, we are unlikely to improve our understanding of how to safely and effectively manage a wide range of diseases and injuries, minimise discomfort, reduce organ dysfunction, improve survival, improve quality of life in survivors of critical illness, and ensure rational and equitable use of scarce resources.

Clinical research in the critical care environment is therefore essential to inform best practice (‘evidence-based care’).

The Southern African Journal of Critical Care (SAJCC) is committed to publishing clinical research in critical care, with the proviso that the research is conducted and presented in an ethically appropriate manner. Critical care research has special ethical challenges. Many critically ill patients are essentially captive and entirely dependent on the ICU team for their care.

Their critical illness may mean that their ability to tolerate an adverse event from an experimental intervention, such as excessive blood sampling, could be compromised. In addition, their condition, medication and presence of invasive ventilation may make communication and/or understanding difficult, and affect a patient’s ability to make rational informed decisions.

Most ICU patients are therefore in no position to give informed consent.

Proxy consent by a relative may also be compromised by anxiety for their relative, making them susceptible to pressure from the ICU management team. In this context, a balance is clearly needed between discovering new knowledge and protecting patients from research-related risk of harm.[2] These aspects, among others, make ICU patients particularly vulnerable to research-related exploitation, coercion and risk of harm.[3] It is, however, generally accepted that research can and should be con­ ducted in vulnerable population groups, provided additional safeguards are put in place to minimise the risk of research-related harm and exploitation.

These include the ‘subject-condition’ requirement, whereby participants must have the condition being investigated, and the ‘necessity’ require­ ment, where there must be assurance that the research could not feasibly or appropriately be conducted in a less vulnerable group.[4-6] In order to ensure that the ethical integrity of critical care research is maintained, a number of ethical principles must be considered. The Belmont report[7] presents three fundamental concepts for ethical research: respect for persons, justice, and beneficence (or its corollary non-maleficence). The principle of respect for persons includes the obligation to treat the subject as an autonomous agent, for example by obtaining their informed consent for participation in research.

Beneficence refers to the requirement that researchers act to maximise potential benefits to participants while minimising associated risks.

Emanuel et al.[8] expanded these principles in listing the requirements for ethical research. These are discussed below in the context of ICU research.[9]

Societal value Considering the global burden of critical illness (both acutely and after discharge), most research with the objective to improve critical care

34

SAJCC November 2015, Vol. 31, No. 2

within the community concerned has clear societal benefit, provided such research can be used to improve clinical outcome.

Scientific rigour and validity Poor science equates to poor ethics, although these concepts are sometimes confused. If research is poorly conducted or method­ologic­ ally unsound, it cannot be said to be generalisable. This could place participants at risk of research-related harm without any personal or societal benefit (i.e. it would not be able to inform practice), which shifts the risk-benefit ratio to unfavourable.

In order for a novel intervention to be compared with either placebo or current care, there is the ethical requirement for equipoise – this is defined as a genuine uncertainty in the expert medical community on whether an intervention is beneficial to patients. Equipoise is the ethical principle underlying randomisation to different treatment arms.[9]

Acceptable risk-benefit ratio Similar to clinical decision-making processes, any research interventions for which the benefits equal or exceed the risks are generally considered ethically acceptable. In some cases a small net risk may be permissible in order to generate knowledge to benefit future patients.[2,10]

Informed consent Informed consent upholds the principle of respect for autonomy, and, together with independent review, is one of the cornerstones of ethical research.

Informed consent, which is required for most prospective, interventive research, is sometimes difficult to obtain correctly in the ICU context.

The informed consent process is designed as a mechanism for the participant to protect themselves. In order for consent to be valid, potential participants must: have intact decision-making capacity; be legally competent; be fully informed; be able to communicate a decision (not necessarily verbally); and offer the consent voluntarily, without any implicit or explicit coercion or undue influence.[11] Critically ill or injured patients may not be able to understand the information provided to them or have sufficient decisionmaking ability, owing to, among others, the underlying illness, delirium or medication (e.g. sedatives, opioids). The presence of invasive ventilation does not in itself equate to limited capacity, and in ventilated patients the degree of decision-making capacity should therefore be assessed before consent is requested. When communication is limited, researchers could consider taking assent from the participant themselves (indicating some degree of understanding and agreement), as well as proxy consent, similar to what is done in most paediatric research.[3,6,11] In critically ill patients without the capacity to provide independent consent, proxy consent may be obtained from the person’s legally authorised representative (e.g. from an advance directive), a spouse or family member. The person providing consent must be properly acquainted with the participant and should have no conflicts of interest. Proxy consent is given using either the ‘substituted judgement’ standard, where the decision is based on a good-faith judgement of what the participant would have chosen if he/she were able to make the decision him-/herself, or a decision made in the ‘best interests’ of the participant.[6]

It is important that consent forms are written in plain language, avoiding medical jargon, and that they clearly describe the potential benefits and risks of the study. Consent forms must be provided in a language understood by the person giving consent and a translator must be on hand if the investigator taking the consent is not able to speak the consenter’s language.[11] Some categories of research may be awarded a waiver of the need for informed consent following research ethics committee review, for example where the research holds only minimally increased risks above standard care, where important research would not feasibly be possible without the waiver, and in emergency research situations.[12] Silverman et al.[11] provide a template for informed consent in the ICU, which is a useful resource for researchers to adapt to local context.

Research into emergency conditions, such as cardiac arrest, where prospective consent may not be possible, is particularly challenging.

In some situations deferred consent, where consent is obtained retrospectively, may be permitted by the ethics review board. For example in the FIRST trial conducted in Cape Town, deferred consent was permitted because the two resuscitation fluids studied were both already in clinical use, there was equipoise, and the study outcome was considered of sufficient benefit to society to counterpoise potential subject harm.[13]

Fair selection of participants This relates to the ethical principle of justice. Vulnerable critically ill populations should only be included in research where absolutely necessary, but they should also not be excluded from research which might be beneficial to them and/or future ICU patients.

Independent review All research involving human participants requires independent review by a research ethics committee or institutional review board.

This includes the retrospective collection of data from medical records, prospective observational studies and interventional trials. Articles describing human research that have not undergone ethical review will generally not be considered for publication in peer-reviewed journals, including SAJCC, and are considered unethical. Reports of ICU practice improvement initiatives that have not systematically collected individual patient data for the purposes of research may be considered exempt from ethical review. Similarly, case studies and small case series (usually <5 patients) may be exempt from ethical review, but consent should be obtained from the included patients for the use of their data, and this must be stated in the research paper.

Respect for potential and enrolled subjects Participants should have the right to withdraw from a study or not to participate without affecting the medical care they receive in ICU. This must be explicitly stated in the informed consent document, and during the informed consent process.

Clinical research should be distinguishable from clinical care. Clinical care has a personalised focus directed at helping a particular per­ son in need of expert medical attention. The purpose of clinical research, however, is to develop generalisable knowledge, which may or may not benefit the individual participant.[11] An inability to fully understand the difference between research and clinical practice is termed therapeutic misconception. Participants may conflate research participation with receiving ‘cutting-edge’ care, with the unwarranted belief or hope of benefit. Therapeutic misconception is common in the critical care environment, where most critical care health providers, who are first and foremost clinicians with the intent of providing best care to their patients, frequently fulfil a dual role of care provider and researcher.[6]

Conclusion Clinical research is essential to ensure optimal and safe ICU management. Ethical compliance is fundamental to ensuring that we effectively protect these vulnerable patients and ensure best practice through research, rather than protecting them from research.

B Morrow Department of Paediatrics and Child Health, University of Cape Town, South Africa References 1. Vukoja M, Riviello E, Gavrilovic S, et al. A survey on critical care resources and practices in lowand middle-income countries. Glob Heart 2014;9(3):337-42.e1-5. [http://dx.doi.org/10.1016/j. gheart.2014.08.002] 2. Weijer C. The ethical analysis of risk in intensive care unit research. Crit Care 2004;8(2):85-86. [http://dx.doi.org/10.1186/cc2822] 3. Silverman HJ, Lemaire F. Ethics and research in critical care. Intensive Care Med 2006;32(11):1697-1705. [http://dx.doi.org/10.1007/s00134-006-0305-4] 4. World Medical Association (WMA). Declaration of Helsinki. Ethical Principles for Medical Research Involving Human Subjects. Fortaleza, Brazil: 64th WMA General Assembly, 2013. http://www.wma.net/en/30publications/10policies/b3/ (accessed 30 September 2015).

5. US Department of Health and Human Services. Protection of Human Subjects; Code of Federal Regulations; Code of Federal Regulations. 45 CFR 46, 2009. http://www.hhs.gov/ohrp/ humansubjects/guidance/45cfr46.html (accessed 30 September 2015).

6. Silverman H. Protecting vulnerable research subjects in critical care trials: Enhancing the informed consent process and recommendations for safeguards. Ann Intensive Care 2011;1(1):8-5820-1-8. [http://dx.doi.org/10.1186/2110-5820-1-8] 7. The National Commission for the Protection of Human Subjects of Biomedical and Behavioural Research. The Belmont Report: Ethical Principles and Guidelines for the Protection of Human Subjects of Research. 1979. http://www.hhs.gov/ohrp/humansubjects/guidance/belmont. html (accessed 30 September 2015).

8. Emanuel EJ, Wendler D, Grady C. What makes clinical research ethical? JAMA 2000;283(20):27012711.

9. Luce JM, Cook DJ, Martin TR, et al. The ethical conduct of clinical research involving critically ill patients in the United States and Canada: Principles and recommendations. Am J Respir Crit Care Med 2004;170(12):1375-1384. [http://dx.doi.org/10.1164/rccm.200406-726ST] 10. Miller FG, Wertheimer A. Facing up to paternalism in research ethics. Hastings Cent Rep 2007;37(3):24-34. http://www.jstor.org/stable/4625743 (accessed 30 September 2015).

11. Silverman HJ, Luce JM, Lanken PN, et al. Recommendations for informed consent forms for critical care clinical trials. Crit Care Med 2005;33(4):867-882. [http://dx.doi.org/10.1097/01. ccm.0000159201.08203.10] 12. Gefenas E. Balancing ethical principles in emergency medicine research. Sci Eng Ethics 2007;13(3):281-288. [http://dx.doi.org/10.1007/s11948-007-9029-2] 13. James MF, Michell WL, Joubert IA, Nicol AJ, Navsaria PH, Gillespie RS. Resuscitation with hydroxyethyl starch improves renal function and lactate clearance in penetrating trauma in a randomized controlled study: The FIRST trial (Fluids in Resuscitation of Severe Trauma). Br J Anaesth 2011;107(5):693-702. [http://dx.doi.org/10.1093/bja/aer229]

S Afr J Crit Care 2015;31(2):34-35. DOI:10.7196/SAJCC.2015.v31i2.249

The Critical Care Society of Southern Africa works for the benefit of critically ill patients. Membership is open to all healthcare professionals involved in the management of the critically ill.

Visit the Society’s web page at: www.criticalcare.org.za

SAJCC November 2015, Vol. 31, No. 2

35

ARTICLE

Moral distress experienced by intensive care nurses G C Langley, PhD; L Kisorio, PhD; S Schmollgruber, PhD Department of Nursing Education, Faculty of Health Sciences, University of the Witwatersrand, Johannesburg, South Africa Corresponding author: S Schmollgruber (shelley.schmollgruber@wits.ac.za)

Background. Moral distress is experienced when nurses experience conflict while making an ethical decision. This is magnified when the decisions are about withholding or withdrawing life-sustaining treatment.

Objective. To explore and describe nurses’ experiences of situations that involve end-of-life care and evoke moral distress in the intensive care units (ICUs) of two public tertiary-level hospitals in South Africa (SA), the personal consequences of these situations and the means employed to manage their distress.

Methods. An exploratory, descriptive design was used. A short survey/interview guide was administered to registered and enrolled nurses (N=100) employed in the ICUs from two academic-affiliated, specialist public hospitals.

Results. A total of 65 completed surveys were collected. Of these, 32 responses were judged not to be describing moral distress while 33 clearly described moral distress and were included and analysed by means of initial content analysis. The findings were presented in five major categories: (i) collegial incompetence or inexperience; (ii) resource constraints; (iii) end-of-life issues; (iv) lack of consultation, communication and negotiation; and (v) support.

Conclusion. The study found that nurses experienced considerable moral distress. This is compounded in an environment where gender, professional and social status inhibit the nurses’ assertiveness, ‘voice’ and influence in the healthcare system. Parallels can be drawn between the microcosm of the ICU and the macrocosm of the SA social and ethical character.

S Afr J Crit Care 2015;31(2):36-41. DOI:10.7196/SAJCC.2015.v31i2.235

It has long been recognised that moral distress may be experienced by all health practitioners when a grave decision has to be made, when personal or team conflict arises or when constraints make it difficult or impossible to follow the perceived correct course of action.[1] This distress may be expressed in physical and emotional ways, e.g. guilt, anger, feelings of poor self-worth, frustration, depression, spiritual distress, insomnia, nightmares and absenteeism.

Because active treatment and end-of-life decisions are mostly made by doctors, the medical perspective of cure is frequently emphasised and that of nursing and caring, marginalised. Nurses are involved in the care of critically ill and dying patients, their mandate being to provide patient and family-centred care in intimate and often intense situations such as in the intensive care unit (ICU).[2] In this heightened emotional climate, ethical decisions have to be made that may be contested or divisive within the multidisciplinary healthcare team.[3] End-of-life care (EOL) in particular has long been recognised as engendering moral distress due to the situations inherent in caring for a dying person.[1,4-11] A study conducted in Johannesburg level 1 ICUs revealed that 33% of the nurse participants felt that decisions to withdraw active treatment were made too late.[12] Distress may also be experienced when the expectations of family members are unrealistic or not in the patient’s best interests, when patient safety is compromised[13] or when the medical team overrides the patient’s or family’s wishes. [14] Compassion fatigue is commonly found in staff caring for patients and their families faced with critical illness and death.[15] Furthermore, the ethical sensitivity of the carer appears to be proportionate to the distress experienced.[11] Anger, anxiety, sadness, depression, guilt, frustration and feelings of helplessness and hopelessness may signify distress.[9,16,17]

36

SAJCC November 2015, Vol. 31, No. 2

Physical symptoms may include gastrointestinal effects, headaches, palpitations, insomnia and nightmares.[18,19] Socially, withdrawing from friends and family, and substance abuse have been cited.[19] Professionally, a reluctance to interact with patients’ families, limiting care, and absenteeism or leaving the unit, the hospital or even the profession have been noted as a result of moral distress.[3,9,14,16] Despite the causes of and responses to moral distress being extensively reported, the South African (SA) healthcare system merits further exploration as it is conceivable that the SA context would add unique stressors to the ICU environment. The country is considered to have one of the world’s most unequal economies, with a Gini coefficient measuring 0.58; the population unemployment rate is estimated to be between 25 and 35%.[20] Marked wage inequalities exist within the workplace.

Public provincial hospitals are managed by SA provincial health departments and cater for the majority of the population. The mortality rate in ICUs varies from 20 to 40%.[21] Poverty-related diseases, chronic diseases, a high incidence of violence and injuries are compounded by the presence of HIV/AIDS and tuberculosis.[22,23] Negative media portrayal of nursing, the quality of nurse education and nursing graduates, and the acute shortage of nurses also cause anger and concern among the general public.

A tension exists between the medical staff (appointed jointly by the university and the provincial health department) and the hospital management and provincial authorities who administer the facilities and also employ the nursing staff members. Strained management-employee and nurse-physician relationships and competing expectations from all involved affect the emotional and moral climate.[24] Economic inequality and differences in ethnicity (the majority of nurses are black women), gender, social status, culture, spiritual values, education (most nurses are educated in nursing

colleges v. doctors at university) and professional socialisation are exacerbated in the acute-care environment and add to the potentially fraught situation.[25] Relationships are strained as authorities attempt to implement government policy in a climate of political interference, deteriorating infrastructure, corruption and wasting of public resources. A poor ethical climate is not conducive to moral equilibrium and ethical composure.[25,26]

Method The study was passed by the University of the Witwatersrand Human Research Ethics Committee (M121136), the hospitals and the relevant provincial authorities. After informal discussions in the relevant units, an information letter inviting participation and the data collection instrument were distributed to the total population (N=100, n=100) of registered and enrolled nurses in the ICUs of two major specialist hospitals.

The instrument used comprised open-ended questions requiring narrative descriptions and explanations, and was derived from Corley et al.’s[16] Moral Distress Scale. Those invited to participate could decline participation or withdraw at any time without consequence.

Posting the completed instrument in a sealed box would indicate consent to participate and ensure anonymity. A blank questionnaire indicated a wish to refuse participation. Collection boxes were emptied on a weekly basis over 4 weeks. An offer of counselling with independent counsellors was included should the participants find the participation experience emotionally distressing. Many participants attached extra pages to the instrument to describe complex situations – some commenting that the experience of writing their stories had been cathartic.

Four focus group discussions of 45 minutes - 1.5 hours each with a minimum of six nurse participants (six to nine members per group) were also conducted. Members were invited from the same ICUs in order to expand upon or give greater depth to the written responses obtained in the survey and on the understanding that while in-group confidentiality was urged, anonymity could not be guaranteed. The members were persons who accepted the invitation to participate and no effort was made to establish whether they had also participated in the survey. Verbatim transcripts from these groups with added field notes augmented the written narratives and ensured data triangulation.

Lincoln and Guba’s[27] four criteria for qualitative data gathering, namely credibility, dependability, confirmability and transferability were applied to ensure trustworthiness. As the researchers are considered the ‘instruments’ in qualitative studies,[28] it should be noted that the authors are experienced in the use of qualitative methods at a postgraduate level. All three were known in and had conducted studies in these ICUs. By virtue of their being accepted by the staff, they received assistance from the nursing and medical staff, thus meeting the requirement for prolonged engagement in the setting.

Excerpts from the written narratives are given in this article to illustrate the data presented and to enhance data credibility.

Neutrality is achieved by providing narratives faithful to the participants’ written stories and providing evidence of deviant as well as typical accounts. [29] A definition of moral distress, constructed by the authors from different published sources, was given to guide the participants, namely: A conflict which arises in certain circumstances to do with patient care, which occurs:

• when one knows or believes what the correct thing would be to do but can’t pursue this option OR • when either of two responses might be appropriate to a situation, both of which are not considered ideal.

Demographic information required included only the participants’ qualification: registered nurse with an ICU qualification (diploma or degree in critical care); registered nurse with ICU experience; Registered nurse and ICU student; or enrolled nurse.

Participants were invited to respond to three open-ended questions: • What situation/issue concerning the care of a critically ill or dying patient have you experienced which has caused you to feel morally distressed?

• What did you experience physically, mentally (thoughts), emotionally (feelings) or behaviourally (that is, your actions) which you consider due to the effects of this experience?

• How did you manage the feelings or thoughts in order to reassure/ soothe yourself or regain your composure?

Analysis A total of 65 completed surveys were collected and numbered sequentially. An initial content analysis evaluated the responses.

Thirty-two written responses clearly described emotional distress: sadness or distress in response to death of a patient or when comforting newly bereaved families; they were set aside and not analysed in this study. Thirty-three responses clearly described moral distress where a decision or action was experienced as infringing upon moral beliefs or ethical values held; these were included for analysis.

Data were then augmented with transcript material discussing moral distress from the focus group discussions and field notes that expanded on the focus group process. The data were categorised according to the situations or issues engendering distress, the experiences of the nurses, their actions taken, any ensuing effects and the manner in which the nurses dealt with these effects. This was a recursive process that involved going between tentative categories and returning repeatedly to the data, discussion, and continued refinement and positing of themes. Data were deemed sufficiently saturated when it was obvious that no new information would be obtained and that negative cases had been identified.

Results Findings are presented under the five major themes elicited, which evoked moral distress: (i) collegial incompetence or inexperience; (ii) resource constraints; (iii) lack of consultation, communication and negotiation; (iv) end-of-life issues: maintaining futile care or withdrawing treatment; and (v) support. Where quotes are given to illustrate the theme or subtheme, the source will be given as B (for hospital 1), C (for hospital 2) or FG (for quotes taken from the focus group interview transcripts. Qualification and experience are noted where the information was given in the responses.

Theme 1: Collegial incompetence or inexperience Perceived incompetence of colleagues was a source of considerable moral distress, particularly when deemed to threaten the integrity of the patient. Doctors, ‘agency’ nursing staff and students were among those considered to have insufficient knowledge or experience. This problem was mentioned by all participants in

SAJCC November 2015, Vol. 31, No. 2

37

conjunction with the complaint of a shortage of staff. Nurses felt unable to confront or rectify the situation due to the hierarchical nature of the doctor-nurse relationship, the prevailing shortage of staff or their own inadequacy.

One registered ICU nurse told of nursing students who repeatedly failed to insert an intravenous (IV) line: ‘They should have told me that they didn’t know. That was unfair, I was very worried’ (C5). Another said: ‘… inexperienced nurse, critically ill patient, not even getting support from senior staff ‘ (C13). Two ICU students (C12 and 7) stated that they were inadequately prepared but unable to ask for help as they were the only registered nurses available to care for the patients despite their inexperience.

An ICU registered nurse admitted to neglecting her patient: ‘I got back … his condition had deteriorated, blocked ET tube, SATS less than 80, hypotensive, bradycardia. I was frightened and depressed.

His safety was compromised because of staff shortage. Afterward I wasn’t able to concentrate’ (B22).

Ten of the thirteen responses mentioned doctors causing moral concerns. Generally, doctors were not challenged even when their inexperience was obvious. One participant recounted that the doctor ‘gave scoline (succinylcholine) and then couldn’t intubate. I was running around seeking help from the anaesthetic department’ (C1). It is interesting that the participant did not assert any authority, but deferred to the doctor and did not take over and insert an endotracheal tube personally.

An ICU nurse said ‘… gunshot abdomen and lost so much blood. I advised her to increase the fluids but she wouldn’t listen. I just gave in, I was submissive, just followed orders, I did nothing actually!’ (C3).

The regret and guilt in the narrative are obvious.

Another registered nurse (C8) said that she felt emotionally exhausted when many patients died after being operated on by a particular surgeon. She alerted the hospital management about her concern. An ICU nurse (B18) told of doctors who wouldn’t review the treatment of a confused patient but prescribed sedation (haloperidol) and mechanical restraint. She finally persuaded them to examine the patient: ‘… his vocal chords were paralysed! Some patients die because we fail to diagnose them properly.’

An ICU nurse (B16) felt that she hadn’t been sufficiently insistent when a junior consultant refused to move a badly burned patient to a specialised unit despite the nursing staff’s urging. The patient died.

Another ICU nurse (C7) recounted that a post­operative patient was pale, in pain and bleeding from the IV site. ‘So I asked about his history. He said he used to bleed when he was a child … the doctors didn’t ask’. He exanguinated and she said that she still has nightmares, feels guilty, frets about the patient’s family and doesn’t trust doctors.

An ICU nurse (B5) described how a doctor who had repeatedly failed to insert an arterial line ‘writes a lie inside patient notes – that failure is due to lack of equipment. I felt frustrated!’ She regretted that she didn’t challenge the doctor’s explanation, stop the procedure or call a more experienced physician. Truthfulness is an ethical norm and hallmark of professional behaviour[4] and lying about the reasons for the failure of an intervention is likely to engender moral distress in colleagues. In their Canadian study, Pauly et al.[26] found that competency of self and other healthcare providers emerged as key issues associated with moral distress.

Theme 2: Resource constraints Staff shortages were a common problem that caused both anger and moral distress, because it was perceived that the patients’ lives

38

SAJCC November 2015, Vol. 31, No. 2

were jeopardised. One ICU nurse (C6) explained: ‘Absenteeism, sickness, family responsibilities … I feel emotionally exhausted and guilty. Patient care compromised, no shift leader, no overall supervisor.’ Another senior ICU nurse (B3) wrote: ‘… using nursing assistants or grabbing anyone who calls herself ICU experience (sic) as long as there is someone in front of the patient. You get burned out, lose faith in people in charge; I am accountable, my profession at risk.’

Because of the shortage of staff in provincial hospital units, agency nurses are used to ensure a minimum level of staffing. In some cases the shift leader was the only permanent member of staff and often the only qualified ICU nurse. The responsibility for the whole unit and the practice of others evoked anxiety, anger and a perception that they, as registered staff members, were being abused: ‘… coming to ICU not trained, also lacking experience; one called a cardiac monitor a ventilator! I’m emotionally exhausted seeing patients at risk in unsafe hands!’ (C2). Many nurses ‘moonlight’ to supplement their income and request to work in ICU, relying on the expertise of the permanent staff to cover their inadequacies. The nurse in charge is likely to fear for the patients’ welfare in the hands of unqualified or inexperienced colleagues. Only one participant mentioned that she had complained to the hospital management with good effect.

Increasingly the concept of ‘skills mix’ is used, with subprofessional staff (enrolled nurses) being brought in to care for patients for financial reasons or to augment the number of professional nurses. One ICU nurse (B15) said: ‘… eight beds, enrolled nurses only, some of them not even experienced and the patients (are) sick! I had a critical patient myself. I felt mentally and physically exhausted. I was angry – one nurse accused me of following her around, checking on her!’

Adequate staffing levels with competent, registered nurses has been identified as important in ensuring safe patient care and addressing the intensity of moral distress.[26]

Theme 3: Lack of communication, consultation and negotiation Two participants in the survey group and three in the focus group discussions said that the need to communicate with family members evoked moral distress: ‘… they’re a challenge. They still have hopes and want to hear something different … they’ve been told by the doctor but they still come and ask you. Miracles do happen so you can’t say the patient will die’ (FG2). Many exhibited a strong religious or fatalistic point of view, and spoke of prayer as a means of coping with their distress or rationalised the death of patients by referring to God as the ultimate arbiter of life and death.

However, one study has demonstrated that registered nurses were able to identify situations of futile care more often than doctors.[30] If this is so, moral distress may be experienced as the nurse tries to align his or her deeply felt personal beliefs with their professional knowledge and experience, and with that of the medical staff and the hopes of the family.

The difficulty of caring for the family after the patient’s death is illustrated by one nurse (FG12), who said: ‘the family sit (at the dead person’s bedside), I run around looking for a shroud. There is no time – the surgeon has already asked for a bed.’ In the same transcript another nurse said: ‘while they’re still there, I’m looking for the pump and bringing in the admission that is coming for the bed!’

A focus group participant (FG10) said: ‘Sometimes the nurses disappear – they don’t want to be asked questions.’ This was echoed

by an experienced registered nurse (C14): ‘I have decided to keep quiet, only speak when necessary, they don’t accept the condition.

It is disheartening to try and convince them.’ This avoidance is concerning as a nurse cares for both the patient and the family.

When talking to the patient is not possible, it is the family who need information, reassurance and guidance on the process of dying. Pauly et al.[26] note that moral distress may manifest behaviours such as avoiding or withdrawing from patients.

Another ICU nurse (B6) spoke of the family being angrily divided about the decision to terminate treatment. She also disagreed with the decision but had to explain why she was switching off the ventilator. This situation was particularly difficult because ‘… there is a patient with a good prognosis who needs to be admitted to the same cubicle. This is very difficult. You’ve maybe failed, maybe you were not competent enough’. Gutierrez[14] and Fry et al.[18] assert that persons experiencing moral distress experience a decrease in their self-esteem and self-confidence and tend to blame themselves.

The lack of team communication was a common problem. One focus group participant (FG5) reported: ‘Doctors just go ahead and do their own thing; you don’t matter.’ Another (FG2) in the same focus group agreed: ‘No, you don’t argue’.

Poor communication also included team members ignoring suggestions for treatment. A registered ICU nurse (B19) recounted how she had pressed a doctor to intervene urgently but her advice was discounted. ‘I felt demotivated … my contribution was ignored, the patient could lose his kidney.’ Despite the rebuff, she called the nephrology consultant and asked him to initiate treatment. ‘Doctors must understand that ICU nurses have worked with these patients.’

She was one of the few participants who stated that they had acted autonomously.

In addition to feeling that their advice was ignored, nurses also verbalised that they wanted to be included when decisions were made. ‘I shouldn’t feel like I’m being left out and just being told; I need to understand!’ (FG6).

All but one participant stated that they wished that they had been consulted or included in the discussion leading to the decision to withdraw active treatment; only two of the participants stated that they had been involved but explained that they were included only because they were in a position of authority. Another two wrote that they had requested that they be involved in end-of-life discussions when they were caring for the patient under consideration. This lack of communication may lead to fundamental differences in perception. For example, a teenage girl had sustained irreparable brain damage. Active treatment was terminated without consulting or informing the child’s family but ventilation and cardiac monitoring were continued. Only later was the family asked for consent for organ donation. The nurse (B25) said: ‘Doctors only care for organ donors – this for me is a crisis of conscience’. The decision and management of the case was probably ethical and correct; however, it appears that the nurses were not included in the decision and this nurse felt that they and the family should have been consulted before the decision was made. However, the processes of who negotiates with whom and when also need to be agreed upon, as some nurses do not wish to be part of the deciding team: ‘I’m not God. I wouldn’t like to be included when the family still have hope’ (FG6).

One nurse (B15) felt defeated when the consultant countermanded the doctor’s and her recommendations: ‘So we were blamed for the death; we tried but that is how it is, somebody has to be blamed’.

Schluter et al.[25] comment that unsuccessful attempts to advocate

for patients engender strong feelings of moral outrage and moral distress.

The hierarchical structure of the healthcare system is demonstrated in the above section. Nurses need to be advocates for their patients and the families. They are the professionals who are in a position to observe the patient’s status and manage or adjust the complex interventions. They are also the members of staff who are in a position to interact with and give information and reassurance to family members. However, very few participants in this study felt empowered to act autonomously or challenge the lack of resources, the decision-making process or the medical sovereignty.

Theme 4: End-of-life issues: maintaining futile care or withdrawing treatment Three participants specifically addressed the issue of maintaining futile care in their narratives. This issue was also discussed exten­ sively in the focus group transcripts.

One registered ICU nurse (B23) was sanguine: ‘I sympathise with the family. The patient is a breadwinner but according to the Bible, we must depart.’ Another nurse (FG1) in the focus group interviews said: ‘You’re not adding value, just prolonging suffering. The patient is still going to die.’ Yet another (FG2) said: ‘They don’t want to feel guilty!’

It is moot whether these nurses experienced no emotional distress or whether they were losing their capacity for caring as a result of moral distress.[3] Compassion fatigue is commonly found in staff who are regularly faced with critical illness, death and assisting families to deal with grief.[15] One focus group member (FG3) expressed concern about the medical team’s inability to come to an agreed-upon decision: ‘… orders are not clear: one will say “no escalation”, the other will say “give voluven” (a plasma volume replacement); there is no clear-cut end. You don’t know because there is no handover from one to the other.’

One ICU nurse (C4) recounted her distress when a patient who had been declared brain dead was prescribed IV inotropes by a consultant.

She asked why this was being done but failed to receive a satisfactory answer, ‘delaying the patient and causing more unnecessary labour’.

It is difficult to predict patient survival accurately. Active care given to a patient who has no hope of recovering prolongs the dying process and creates moral distress, raises false hopes for the family and is said to contribute to an extended grieving period.[31] Gutierrez[14] asserts that when doctors give erroneous information about a patient’s prognosis, nurses experience moral conflict and distress.

Schluter et al.[25] state that moral distress is caused by providing poorquality or futile care, unsuccessful advocacy and unrealistic hope.

Three nurses described the distress they felt when a decision was made to hasten death. One ICU nurse (B16) recounted how she was told to give ‘high doses of morphine and Dormicum’. ‘He was intact mentally, just not recovering physically, and aware of being killed or left to die because we’re not going to do anything more for them. This is murder!’ Similarly, a patient who was unable to be weaned from a ventilator was ‘looking with fear; he knew it was a death sentence. He was detached from the ventilator and then bradycardic til asystolic.’

A particularly poignant story was told by an ICU nurse (B14): ‘Patient was fully conscious, but the family asked that treatment be terminated, citing the patient’s “living will”. The consultants “ordered” that the endotracheal tube be removed but wouldn’t do this themselves and delegated the task to a “junior doctor”. It was unfair for the junior doctor to be given an order that will definitely

SAJCC November 2015, Vol. 31, No. 2

39

kill.’ The nurse said: ‘I just switched off my mind. I went off for 2 days but I was hurting.’

A similar narrative told of a young foreign patient who was extensively burned. He remained in hospital, ventilated, for 3 months but could not be weaned off the ventilator. He was finally transferred to specialist unit but soon died. The nurse attributed the delay in transfer to racism and xenophobia, and felt ‘very bitter; we failed the patient and denied him right to treatment’ (B18) The focus group transcripts revealed another response: two nurses who disagreed with the decision to terminate active care told of undermining the process. The first (FG1) said: ‘I will treat the patient – even if it’s not on the chart; even if it’s with a little glucose. I will not slow down.’ In the same group, a nurse (FG5) said they would continue with treatment: ‘… even inotropes, but you forget about the colleague coming on duty. It’s going to be a shock to find out that there was really no blood pressure.’

Two nurses felt that the process of asking the family for consent to harvest organs was traumatic and poorly managed. One registered nurse (B5) said: ‘I was so disturbed … I shouted at the lady who came to enquire about organ donation. I shouted at the doctor – told him to come and explain to the relatives. They must not pressurise.’

Once again, the lack of interdisciplinary communication and agreement regarding the care of the patient and decisions as to the outcomes envisaged is demonstrated. Nurses were angry and resentful, or in some cases resorted to passive (and frightening) means of undermining the process. The outcome of their behaviour might well further disrupt team cohesion and lead to distrust and alienation.

Theme 5: Support Fifteen participants explicitly stated that no support was offered by management; they felt that this should be given and said that the poor support was likely to affect their care of patients: ‘… things that we get exposed to are deep, but there is nothing. You survive, then you become desensitised. You admit another patient, carry on, no debriefing, none expected. Someone should say “How do you feel? In future how should we deal with this?”’ (ICU nurse C15).

One said that counselling had been provided after a particularly traumatic incident but that this hadn’t helped; another one said that counselling had been offered but they were unable to attend on the day of the appointment.

Four derived some comfort from the belief that it was God’s decision as to whether the patient lived or died. One person stated that prayer helped. Most discussed the situation that distressed them with their colleagues.

Participants offered suggestions as to what they believed should be done: ‘Give us stress management’ (B28); ‘… we should be debriefed.

Just to talk to someone and offload’ (FG1); ‘Offer some EOL training.

Junior nurses are the most affected’ (B12).

The effect of witnessing or experiencing extreme stress is a key aetiological factor in post-traumatic stress disorder. Vicarious traumatisation is recognised in those who care for people who have been exposed to a major stressor, and describes the cumulative transformative effect on the caregiver.[32] The effects appear similar to those of post-traumatic stress disorder and affect the sufferer’s frame of reference, identity, sense of safety, trust, self-esteem, intimacy and sense of self-control. The continuing exposure to fraught situations, the effects of violence and trauma, and distress may cause emotional numbing and even result in the caregiver choosing to leave the practice

40

SAJCC November 2015, Vol. 31, No. 2

area or even the profession.[3,9,14,16] Emotional support, regular debriefing and, if required, referral to a professional counsellor may assist the nurse to engage in reflective practice, encourage self-awareness and self-care and offer collegial support to other practitioners in the intensive care arena.

Discussion In this study, considerable distress was evoked by the perceived morality of the situation or the treatment decisions and their influence on nursing care. The situations or issues identified which caused moral distress were: perceived collegial incompetence or inexperience; a shortage of resources, particularly staff; poor communication and consultation practices; and concerns around EOL (maintaining futile care or withdrawing treatment, and support).

Many of the narratives gave details of distress experienced by the participants when decisions were made to withdraw or withhold active treatment, when they felt conflicted about the decisions, when they were expected to institute the process of withdrawing treatment, and when they were told of a decision without having been party to the discussion.

Of note was that, apart from discussions with their colleagues, the nurses perceived that they were totally unsupported. Management was viewed as uncaring and of the 33 written responses, five explicitly stated that there was no support or debriefing offered. More than half of the participants urged that counselling and debriefing sessions be arranged by hospital management. Supervisors were also not seen as helpful, while they in turn were conscious that their ability to supervise staff and students was compromised by the pressure of work and the inadequacy of many of the staff they supervised. No mention was made of multidisciplinary team support, and unit supervisory support was only mentioned by one participant supervisor who stated that she made a point of trying to assist students. Four cited collegial support as helpful and one person stated that she had called a climate meeting to discuss a particularly distressing situation.

God and prayer were also viewed as a means of comfort for both the patient families and the nurse. Many emphasised that simply writing their stories had helped them deal with emotions that had lain dormant for some time.

It appears that participants were profoundly affected by their experiences; they felt guilty, humiliated, inadequate, angry and, in one case, bitter. Many wrote long, detailed narratives about the situation, and one person mentioned that they still dreamt about a particular situation. This accords with the literature surveyed, which describes how distress was experienced as feelings of powerlessness, subordination and compromised efficiency that might engender passivity and blunted moral sensitivity.[33] The possible effects of unresolved moral distress are significant, and include: job dissatisfaction, burnout and turnover of staff. This, in an already stretched environment with poor resources and staff shortages, has grave implications for healthcare.

SA’s history is that of a racially and ethnically divided, patriarchal society with marked cultural, social and educational divisions between the diverse groups. Medicine as a profession still enjoys status not accorded to nursing, with the profession predominantly represented by advantaged, university-educated, white and Asian males; the nursing profession still primarily comprises black women who have diplomas in nursing. A social hierarchical system persists, and culturally, black women are socialised to be submissive and

defer to men or people in authority. This is not peculiar to SA and many studies attest to non-assertive behaviour of nurses in the workplace. [34- 36] Nurse education and socialisation may stress autonomy and patient advocacy but, in the clinical situation, it is often difficult to practise these attributes. Consequently, patterns of passivity or submission to medical authority prevail, leading to tension between what is believed or held important and what is culturally and socially the norm. Narratives and focus group transcripts revealed that patient advocacy was tentative but abandoned when the nurse’s voice opposed that of the doctor. Most felt that they were forced to accept the situation; some addressed the problem after the event by challenging the person they perceived as incompetent or referring the issue to management. Some felt angry or devalued and responded passively or aggressively, but seldom assertively.

Participants wrote of termination or withdrawal of treatment as a cause of moral distress. Many suggested how the decisions or interventions should have been made. Some stated that they were actually able to do what was required but chose not to do so. One or two were able to comment on or question the decision after the occurrence. Recommendations as to the correct treatment, procedure or intervention might be proffered but, when not adopted, no further action was taken – even when this inaction was to the patient’s detriment; as one nurse said: ‘I was submissive, I did nothing!’(C3).

The lack of trust, inadequate communication between disciplines; absence of collegiality, restricted institutional and mutual support and the continued submission of nurses within the ICU is a reflection of a divided society in microcosm.

Conclusion The divisions evident in SA society, with its history of class, race, professional and gender inequality are mirrored and perhaps exacerbated in the ICU. A history of trauma, denigration and disenfranchisement that affects self-esteem, speaking out (and the right to speak), listening and acting shapes this microcosm.

The primacy of life and death issues prevents the forming of ‘normalising’ relationships (collegial, cooperative and collaborative) and addressing the consequences of this history. Traditional patriar­ chal norms and the hierarchical nature of the doctornurse relationship are reinforced, constraining the autonomy and accountability of the nurse and contributing to moral distress and burnout.

Recommendations Support needs to be offered on a regular basis for nurses practising in the ICU. This should be based on a non-directive ‘story-telling’ format and allow for debriefing and affirming in a safe, group context. At the same time, nurses’ self-reflection and self-awareness should be encouraged; this could be done in the group. Gradually, assertiveness education and training could be addressed in workshops, super­ vision meetings with individual staff members, and continuing professional development (CPD) presentations and staff meetings.

Skills could be practised in a safe, supportive group environment.

Encouraging a change in attitudes and behaviour affords a space to ‘try out’ and deconstruct the norm – instituting social and professional transformation. Interprofessional collaboration should be promoted, for example, nurses should attend daily ward rounds and be urged to contribute to the discussion and planning of

patient care. Enhancement of collaboration, nurse autonomy and interdisciplinary respect is likely to promote an appreciation of nursing input and a decrease in the instances of moral distress.

References 1. Jameton A. Nursing practice: The ethical issues. Englewood Cliffs, New Jersey, USA: Prentice Hall, 1994.

2. Gastmans C. Nursing ethics perspectives on end-of-life care. Nurs Ethics 2012;19(5):603-604. [http://dx.doi.org/10.1177/0969733012453234] 3. American Association of Critical-Care Nurses (AACN). Moral distress position statement. Aliso Viejo, California, USA: AACN, 2006.

4. Wilkinson JM. Moral distress in nursing practice: Experience and effect. Nurs Forum 1987;23(1):16-29. [http://dx.doi.org/10.1111/j.1744-6198.1987.tb00794.x] 5. Fowler MD. Moral distress and the shortage of critical care nurses. Heart Lung 1989;18(3):314315.

6. Hamric AB. Moral distress in everyday ethics. Nurs Outlook 2000;48(5):199-201. [http://dx.doi. org/10.1067/mno.2000.110564] 7. Oberle K, Hughes D. Doctors’ and nurses’ perceptions of ethical problems in endof life decisions. J Adv Nurs 2001;33(6):707-715. [http://dx.doi.org/10.1046/j.13652648.2001.01710.x] 8. Van Soeren M, Miles A. Commentary: The role of teams in resolving moral distress in intensive care unit decision-making. Crit Care 2003;7:217-218. [http://dx.doi.org/10.1186/ cc2168] 9. Elpern EH, Covert B, Kleinpell R. Moral distress of staff nurses in a medical intensive care unit.

Am J Crit Care 2005;14(6):523-530.

10. Lützén K, Dahlqvist V, Eriksson S, Norberg A. Developing the concept of moral sensitivity in health care practice. Nurs Ethics 2006;13(2):187-196.

11. Hamric AB, Davis WS, Childress MD. Moral distress in health care professionals: What is it and what can we do about it? Pharos Alpha Omega Alpha Honor Med Soc 2006;69(1):16-23.

12. Langley G, Schmollgruber S, Fulbrook P, Albarran J, Latour J. South African critical care nurses’ views on end-of-life decision-making and practices. Nurs Crit Care 2013;19(1):9-17. [http:// dx.doi.org/10.1111/nicc.12026] 13. Corley MC. Moral distress of critical care nurses. Am J Crit Care 1995;4(4):280-285.

14. Gutierrez KM. Critical care nurses’ perceptions of and responses to moral distress. Dimens Crit Care Nurs 2005;24(5):229-241.

15. Wee D, Myers D. Compassion satisfaction, compassion fatigue and critical incident stress management. In J Emerg Health 2003;5(1):33-37.

16. Corley MC, Elswick RK, Gorman M, Clor T. Development and evaluation of a moral distress scale. J Adv Nurs 2001;33(2):250-256.

17. Corley MC, Minick P, Elswick RK, Jacobs M. Nurse moral distress and ethical work environment.

Nurs Ethics 2005;12(4):381-390.

18. Fry ST, Harvey RM, Hurley AC, Foley BJ. Development of a model of moral distress in military nursing. Nurs Ethics 2002;9(4):373-387.

19. Nathaniel AK. Moral reckoning in nursing. West J Nurs Res 2006;28(4):439-448. [http://dx.doi. org/10.1177/0193945905284727] 20. Keeton G. Inequality in South Africa. J Helen Suzman Found 2014;1(74):26-31.

21. De Beer J, Brysiewicz J, Bhengu B. Intensive care nursing in South Africa. S Afr J Crit Care 2011;27(1):6-10.

22. Delobelle P, Onya H, Langa C, Mashamba J, Depoorter AM. Advances in health promotion in Africa: Promoting health through hospitals. Global Health Promotion 2010;17(2):33-36. [http:// dx.doi.org/10.1177/1757975910363929] 23. McNeilly L. WHO examines disabilities worldwide: Report outlines prevalence, barriers and health care issues. World Beat 2011;16(9):38.

24. McLendon H, Buckner EB. Distressing situations in the intensive care unit: A descriptive study of nurses’ responses. Dimens Crit Care Nurs 2007;26(5):199-206. [http://dx.doi.org/10.1097/01.

DCC.0000286824.11861.74] 25. Schluter J, Winch S, Holzhauser S, Henderson A. Nurses’ moral sensitivity and hospital ethical climate: A literature review. Nurs Ethics 2008;15(3):304-321. [http://dx.doi. org/10.1177/0969733007088357] 26. Pauly B, Varcoe C, Storch J, Newton L. Registered nurses’ perceptions of moral distress and ethical climate. Nurs Ethics 2009;16(5):561-573. [http://dx.doi. org/10.1177/0969733009106649] 27. Lincoln Y, Guba E. Naturalistic inquiry. Beverly Hills, USA: Sage, 1985.

28. Patton MQ. Qualitative Research and Evaluation Methods. 3rd ed. Thousand Oaks, USA: Sage, 2002.

29. Koch T. Establishing rigour in qualitative research: The decision trail. J Adv Nurs 1994;19(5):976986.

30. Frick S, Uehlinger D, Zenklusen M. Medical futility: Predicting outcome of intensive care unit patients by nurses and doctors – A prospective comparative study. Crit Care Med 2003;31(2):456-461. [http://dx.doi.org/10.1097/01.CCM.0000049945.69373.7C] 31. Meltzer LS, Huckabay LM. Critical care nurses’ perceptions of futile care and its effect on burnout. Am J Crit Care 2004;13(3):202-208.

32. Trippany RL, White Kress VE, Wilcoxon SA. Preventing vicarious trauma: What counsellors should know when working with trauma survivors. J Couns Dev 2004;82(1):31-37. [http:// dx.doi.org/10.1002/j.1556-6678.2004.tb00283.x] 33. De Veer AJ, Francke AL, Struijs A, Willems DL. Determinants of moral distress in daily practice: A cross sectional correlational questionnaire survey. Int J Nurs Stud 2013;50(1):100-108. [http:// dx.doi.org/10.1016/j.ijnurstu.2012.08.017] 34. Timmins F, McCabe C. How assertive are nurses in the workplace? J Nurs Manage 2005;13(1):6167. [http://dx.doi.org/10.1111/j.1365-2834.2004.00492.x] 35. Timmins F, McCabe C. Nurses’ and midwives’ assertive behaviour in the workplace. J Adv Nurs 2005;51(1):38-45. [http://dx.doi.org/10.1111/j.1365-2648.2005.03458.x] 36. Butters KJ. A qualitative study of the ethical practice of newly graduated nurses working in mental health. Thesis. New Zealand: Massey University, 2008.

SAJCC November 2015, Vol. 31, No. 2

41

ARTICLE

Nutrition support practices in South African ICUs: Results from a nationwide pilot survey L T Hill, PhD RD (SA) Critical Point Critical Care Nutrition Consultancy, Cape Town, South Africa Corresponding author: L T Hill (laurenhill@mweb.co.za)

Background. Nutrition support of the critically ill can positively affect clinical outcomes. International audit data of nutrition practices in intensive care units (ICUs) suggest that inconsistent application of recommended nutrition support practices (NSPs) occurs. There are no data on NSPs in South African (SA) ICUs. Objective. To perform a national pilot survey of NSPs in private SA adult ICUs. Methods. A descriptive, observational, cross-sectional survey was performed among prescribers of nutrition support in private ICU facilities. Participants were targeted through non-randomised convenience sampling and invited to complete a 51-item electronic questionnaire covering ICU demographics, profile of nutrition prescribers, and nutrition assessment support, delivery and monitoring practices.

Results. Responses were received from 125 practitioners in 60% of representative private hospitals with ICU facilities, mainly general/ mixed ICUs. Forty-six percent of respondents reported structured nutrition support teams and 61% reported that practices were governed by formal nutrition support protocols. Enteral nutrition was reported to be based upon published guidelines by 72% of dietitians, while parenteral nutrition decisions were reportedly based mainly on clinical judgement (43%). For both enteral and parenteral feeding practices, compliance with guidelines was inconsistent. There was a disjuncture between various NSPs as reported by dietitians and by nurses. Nurses generally appeared unaware of published nutrition guidelines.

Conclusion. Various disparities in reported nutrition practices were revealed, suggesting that the organisation and operations of teams and the implementation of protocols informed by published nutrition guidelines may not be well established in private SA ICUs. S Afr J Crit Care 2015;31(2):42-50. DOI:10.7196/SAJCC.2015.v31i2.252

Nutrition support of a critically ill patient is a form of therapy[1] and positively affects important clinical ou­t­ comes. Various studies have shown that it reduces infectious morbidity, duration of mechanical ventilation, length of stay, impaired functional capacity and the overall cost of care.[2-6] Despite this evidence, prevalence of malnutrition in intensive care unit (ICU) patients remains high.[7,8] This suggests that nutrition support practices (NSPs) are still suboptimal in many clinical settings.

Globally, information about NSPs in ICUs is scarce. An extensive survey of NPSs in 20 European countries[9] indicated that NSPs and, in particular, consistent adherence to recommended practices, varied widely throughout the countries surveyed. No data are available on factors related to NSPs in South African (SA) ICUs, which may differ from other parts of the world. In addition, most ICU practice occurs in the private healthcare sector, which is largely run using an open-unit model without a full-time intensivist who maintains executive clinical oversight of all aspects of patient care. Previous SA audit data indicate that 4% of private ICUs were compliant with the closed-unit model.[10] While closed units are reportedly an effective approach to enhanced clinical outcomes owing to integrated care,[10] the private sector is subject to minimal constraints on resources, which may also contribute differently to quality of care.

Furthermore, the typical SA ICU patient differs from that of developed countries in ways that are of relevance to NSPs; patients are generally younger, more likely to enter ICU as a result of trauma, and beset with a different underlying disease burden, most notably HIV and tuberculosis (TB).[11,12] These factors may all influence the manner in which nutrition support is utilised and delivered in SA ICUs, and whether published international nutrition practice guidelines are being followed.

42

SAJCC November 2015, Vol. 31, No. 2

Given these factors and the lack of SA data, the objective of this study was to perform the first national survey of nutrition support and related practices in SA ICUs within the private healthcare structure.

Methods This was a descriptive, observational, cross-sectional survey of nutri­ tion practices of ICU dietitians, doctors and nursing managers working daily in private SA adult ICUs. The study utilised a non-randomised convenience sample of healthcare practitioners practising in the 154 private hospitals with adult ICUs in SA, with the objective to obtain a response from at least one key nutrition prescriber per adult ICU nationwide. Target participants were invited to complete a 51-item online questionnaire covering: (i) ICU and patient demographic characteristics; (ii) professional profile of ICU nutrition support prescribers; (iii) nutrition screening and assessment practices; and (iv) nutrition support prescription, delivery and monitoring practices.

Participants were also invited to share anonymised written nutrition protocols. The electronic survey was managed via a professional online survey company that provided electronic announcements, reminders and unique invitation links to the survey instrument to targeted participants, as well as an open link for any additional responders, and returned anonymised, uncleaned but collated data to the researchers.

Ethical approval to perform the study was obtained from PharmaEthics Independent Research Ethics Committee (ref. number: 150110873), the Research and Ethics Committees of the private hospital groups, the applicable regional managers of the private hospital groups, and nursing managers of individual hosptials. Survey completion was voluntary and data were anonymised.

Statistical analysis was done using Statistica 12 (Statsoft, USA). Results are presented as mean (standard deviation (SD)) or median (interquartile

Nutrition assessment Overall, 82% of ICUs reported that nutrition assessment of patients was routine, with 83% identifying the dietitian as responsible for this task and the remaining identifying either the nurse or doctor, or a joint effort. However, 54% omitted to answer this question. Nutrition assessment was reported by 68% of dietitians

ICU type

n (%)*

Bed capacity, mean (SD)

General/mixed

106 (85)

14.3 (7.9)

Surgical

31 (25)

11.9 (6.4)

Medical

24 (19)

13.7 (5.8)

Cardiothoracic

33 (26)

12.3 (6.2)

Neurological

13 (10)

8 (1 - 8)†

Burns

6 (5)

14 (8 - 30)†

Trauma

15 (12)

6 (6 - 6)†

Overall median

15 (12 - 26)†

Primary diagnosis

n (%)*

Mainly elective surgery

34 (27)

Mainly emergency, non-trauma surgery

9 (7)

Mainly medical, including obstetrics

35 (28)

Mainly mixed trauma

20 (16)

A mix of elective and emergency surgery

20 (16)

Mainly cardiac/cardiothoracic

6 (5)

*Responses exceed 100% as multiple choices were possible. †

Clinical structure for nutrition support

Median (IQR).

70

Hospitals with ICUs

60

Survey responses

Percentage (%)

50 40 30 20 10

t No

rth

W es

St at e Fr ee

o po p

an al pu m

Lim

ga

l at a M

g

Kw aZ ul uN

ut en Ga

Ca p n er rth

Ca p n te r

e

e

e

0

W es

Overall, 46% of respondents reported the pre­sence of a formal nutrition support team in the ICU. However, signifi­cantly more ICU nursing managers (56%) reported the pre­ sence of a nutrition support team compared with reports from dietitians and doctors com­ bined (37%, p=0.03). Questions re­garding professional composi­tion of the nutri­tion support team yielded dis­p ar­a te re­s ults.

Overall, 26% of respondents indicated that the nutrition support team comprised a dietitian, nurse, doctor and pharmacist, 19% by a dietitian and doctor, and 16% by a dietitian, nurse and doctor. However, 31% of respondents reported that the nutri­tion support team comprised a dietitian acting

NSPs

Table 1. Profile of responding ICUs and patients (N=125)

Ca p

A total of 961 targeted electronic survey in­vites were sent to all 154 private hospitals with adult ICUs, and there were 125 (13%) responses. The responses came from 93 distinct ICUs, representing 60% of the private adult ICUs in the country, but were concen­ trated in the three large metropole areas of the Western Cape, Gauteng and KwaZuluNatal provinces. In addition, there were seven responses from unidentified hospitals.

Responses from the nine SA provinces re­flect­­ ed the proportional regional distri­bution of ICUs throughout the country, indicating that this pilot survey is nationally representative of all the private adult ICU facilities nationwide (Fig. 1). The ICU capacity was reported to be a median (IQR) of 15 (12 - 26) beds (Table 1).

Most practitioners reported working mainly in general/mixed ICUs (85%) with mixed gender. The majority of responders (67%) were dedicated to work in a single ICU, while the remaining respondents reported clinical coverage of between two and seven separate ICUs. Surveys were completed by dietitians (48%), ICU nursing managers (48%), intensivists (2%) and specialist physicians (2%). Only one participant shared a nutrition protocol with the researcher.

No

Sample and ICU demographics

te rn

Results

covered both enteral and parenteral nutri­ tion support, while 41% only governed enteral nutrition support and the remainder covered only parenteral nutrition support.

alone in managing nutrition support in the ICU, and 41% of overall responses included the ICU nurse as part of the team. For res­ ponses from the ICU nursing manager sub­ group, only 33% included the ICU nurse as in the nutrition support team, while 42% viewed the dietitian acting alone as the nutri­tion support team. In contrast, 58% of diet­itians included nurses in the team. No doctors included nurses in the professional composition of the nutrition support team.

Only 61% of ICUs reported the presence of formal, written nutrition support protocols governing nutrition support. Of those, 54%

Ea s

range (IQR)) for continuous data, depending on the normality of data. Categorical data are described as frequencies or percentages.

Group differences in frequencies were ana­ lysed using Fisher’s exact test for small samples.

In such cases p<0.05 was taken as statistically significant. Owing to the very small number of doctor responders, this group was combined with dietitians for subgroup analyses.

Provinces

Fig. 1. National representation of ICU facilities.

SAJCC November 2015, Vol. 31, No. 2

43

as being achieved by a combined ABCD (anthropometry, biochemistry, clinical, dietary) approach, while the remainder reported to use various other standard nutrition assessment tools in combination. For subgroup responses, 48% of nurses and 50% of the doctors (n=2) were unsure of the nutrition assessment method used in the ICU. Nutrition decision-making Summary data for nutrition decision-making practices are shown in Table 2 and Fig. 2. The primary decisions about route of nutrition delivery were made jointly by dietitians and doctors (57%), followed by doctors alone (34%). The group overall and the two profession subgroups separately indicated that nursing involvement in this decision was low (3%), although nurses were included in decisionmaking about daily fluid volumes allocated to nutrition in most responses (51%). The dietitians and doctors subgroup reported

using published guidelines (European Society for Parenteral and Enteral Nutrition (ESPEN), American Society for Parenteral and Enteral Nutrition (ASPEN), Canadian or other) to determine nutrition requirements in 21% of responses, with 62% reporting the use of the above-mentioned guidelines together with equation-based and other methods. The remaining responses reported the use of body weight-based methods (8%), and equations such as Harris-Benedict (8%) without reference to guidelines. Of nurses, 42% responded that they were unsure of how nutrition requirements were calculated in the ICU, while 15% reported that body weight-based methods were used and a further 13.5% reported that requirements were based on clinical judgement. These responses were significantly different between nurses and dietitians and doctors combined (Table 2).

The most commonly reported determinant of enteral feed initiation was haemodynamic stability (43%) followed by the first ICU day if no

Table 2. Nutrition support decision-making practices

Decisions regarding route of nutrition support, N

Total group, n (%)

Dietitians and doctors, n (%)

Nurses, n (%)

115

63

52

Made by dietitian and doctor together

66 (57)

39 (63)

27 (52)

Made by specialist doctor alone

39 (34)

19 (31)

20 (40)

Made by dietitian alone

6 (5)

3 (5)

3 (6)

Made jointly by nurses and doctors

3 (3)

2 (3)

1 (2)

Methods of calculating nutrition requirements, N ESPEN/ASPEN or other published guidelines

115

63

52

15 (31)

13 (21)

2 (4)

<0.050 0.003

Clinical judgement

7 (6)

0 (0)

7 (14)

Body weight-based methods

13 (11)

5 (8)

8 (15)

Equations

7 (6)

5 (8)

2 (4)

Indirect calorimetry

2 (2)

0 (0)

2 (4)

Other methods combined with guidelines

48 (42)

39 (62)

9 (17)

Don’t know

23 (20)

1 (2)

22 (42)

Fluid volume allocated to nutrition support, N Decided jointly by multidisciplinary team

113

61

52

58 (51)

30 (48)

28 (54)

Decided by doctor

49 (43)

29 (46)

20 (38)

Decided by dietitian

6 (5)

2 (3)

4 (8)

Determinants of enteral feed initiation, N First day in ICU

114

62

52

10 (9)

2 (3)

8 (15)

Gastric residual volume

3 (3)

1 (1.6)

2 (4)

Haemodynamic stability

49 (43)

31 (50)

18 (35)

Reasonable gastrointestinal function

21 (18)

8 (13)

13 (25)

Presence of bowel sounds

0 (0)

0 (0)

0 (0)

Reasonable nutritional status

2 (2)

1 (1.6)

1 (2)

First day in ICU if no clinical contraindication

30 (26)

19 (31)

11 (21)

Determinants of parenteral feed initiation, N First day in ICU

114

62

52

2 (2)

0 (0)

2 (4)

Gastrointestinal failure, ileus or obstruction

2 (2)

1 (2)

1 (2)

Any gastrointestinal surgery

2 (2)

0 (0)

2 (4)

Poor nutritional status of patient

2 (2)

0 (0)

2 (4)

Unsuccessful enteral nutrition

8 (7)

4 (6)

4 (8)

GIT factors making EN success unlikely

89 (78)

51 (82)

38 (73)

Other factors

9 (8)

6 (10)

3 (6)

*In all cases, where no p-value is shown the intergroup difference is not significant.

44

SAJCC November 2015, Vol. 31, No. 2

Intergroup p-value*

0.040

clinical contraindications were evident (26%). However, 18% of the respondents overall and 25% of nurses indicated that enteral feed initiation was determined by reasonable gastrointestinal function.

The majority (78%) of respondents identified the main determinant of parenteral nutrition initiation as the presence of gastrointestinal factors likely to make enteral feeding unsuccessful, while only 7% indicated this factor to be actually unsuccessful enteral feeding.

The respondents overall and the dietitian and doctor subgroup reported that enteral decisions were governed by a combination of published guidelines and clinical judgement (Fig. 2A). This was significantly different from the nurses’ response, which was that the enteral support decision was based on a multidisciplinary team decision. A total of 23% of respondents reported that enteral nutrition decisions were based solely on the clinical judgement of the dietitian or doctor. In contrast (Fig. 2B), while the majority of dietitians and doctors reported basing parenteral nutrition decisions on published guidelines (58%), 43% of respondents overall identified clinical judgement as the main factor underlying decisions around parenteral nutrition support.

This proportional difference was due to nurses (46%) perceiving a high reliance on clinical judgement for parenteral decisions, while 0% of nursing staff reported published guidelines as a basis for parenteral decision-making.

A 80

*

All

70

Dietitians and doctors

Percentage (%)

60

Nurses

*

50 40 30

* p<0.05 v. nurses

20 10

in

re Un

en to

Ju d

es a

ge

m

nd

su

se fn

em dg ju

m yt ea

lin ar Bo

M

ul

th

tid

gu

id

isc

el

ip

ur

io cis de

oc /d ns ia

tit fd ie en to

m dg e in ica

l ju

en t

s to r

es in el id gu d he lis Pu b

n

0

Cl

Basis for enteral nutrition support decisions

B 60

All

Percentage (%)

50

Dietitians and doctors

40

Nurses

30 20 10

ur e Un s

pa tie T GI

if e in Ro ut

es in

Bo

th

gu

id

el

nt

en t em ju

an d

yt ea ar lin ip

tid isc

ul M

dg

de c m

ns / ia tit

en to fd ie

m dg e

isi

do ct or s

es lin id e gu d he lis Pu b Cl

in

ica

l ju

on

0

Basis for parenteral nutrition support decisions

Fig. 2. Basis for enteral and parenteral nutrition support decisions. (GIT = gastrointestinal tract.)

Enteral nutrition delivery, administration and monitoring Summary data for enteral support practices are shown in Table 3.

Healthcare professions were congruent in reports that enteral feeding was commenced within 24 hours (47%) or 48 hours (22%) of ICU admission. However, 19% of nursing staff reported that timing of enteral feed initiation was dependent on the practice of individual dietitians or doctors. The nasogastric route of delivery was most commonly used (96%). While 41% reported that postpyloric feeding was seldom used, there was a 42% overall reported use of this route in patients having undergone gastric surgery – although significantly more nurses than dietitians and doctors reported this practice (55% v. 31%, p=0.020).

Feeding tube placement depended largely on the dietitian/doctor (82%). While dietitians and doctors reported that the most common method of checking tube position was by chest X-ray (49%), nursing staff reported significantly different bedside methods including use of litmus paper (24.5%, p=0.000 v. 5.0% for dietitians and doctors) and air auscultation (41.5%, p=0.000 v. 6.6% for dietitians and doctors). Pattern of feed delivery was significantly different between dietitians and nurses, with the former reporting mainly 24-hour continuous delivery (59% v. 30%, p=0.010) and nurses reporting mainly continuous delivery with short holds (45% v. 34%, p=0.020). Again, 13% of nurses indicated that delivery routine differed for individual doctors and dietitians.

Other feed management practices (tube and giving set changes, and tolerance checks) were reported significantly differently between dietitians and doctors and nurses (Table 3). For a number of aspects, 30 - 40% of dietitians and doctors responded that they were unsure of the practice.

There were discrepancies in the reported utility of gastric residual volume (GRV) in guiding enteral feeding with a mismatch between 100% of dietitians and doctors who said that GRV was not used at all to guide practice, while 23% later reported that it was used alone or in combination with symptoms (57%) to monitor enteral feed tolerance. In addition, 92% of nurses reported that GRV was the main method used to check enteral feed tolerance. The question on GRV guidance of practice was answered by a disproportionately low proportion of respondents and was not answered by any respondent who had previously indicated that NSPs were governed by formal nutrition protocols in their ICU. No respondents who had indicated enteral feeding protocol use identified a specific GRV threshold that guided practice.

Enteral feed optimisation methods were different between nurses and the combined dietitians and doctors group, but the total research group overall reported using gastric acid suppression and prokinetic agents when clinically indicated (35%) or at the request of nursing staff (21%), while changing the enteral feed was the next most commonly used strategy (19%). Frequency and method of enteral feed monitoring was different between dietitians and doctors compared with nurses, with 12.5% of nurses being unsure of methods used.

Parenteral nutrition delivery, administration and monitoring There was a low initiation of early parenteral nutrition in malnourished patients (7%), and clinical indication overrode guideline-based timing according to most responses (30%) (Table 4). However, 20% overall and 26% of nurses reported parenteral nutrition typically started within 24 hours of ICU admission. The majority (87%) of dietitians and doctors viewed the rationale for parenteral nutrition as the provision of requirements only with unsuccessful enteral support, which was proportionally higher than the percentage of nurses (61%) with the same view (group difference, p=0.0065). Relatively more nurses (19.5%) than dietititians and doctors reported use of parenteral nutri­tion as a supplement to enteral feeds (p=0.030). Frequency and methods SAJCC November 2015, Vol. 31, No. 2

45

Table 3. Enteral NSPs

Typical timing of enteral feed initiation, N

Total group, n (%)

Dietitians and doctors, n (%)

Nurses, n (%)

113

61

52

Within 24 hours of ICU admission

53 (47)

29 (47.5)

24 (46)

Within 48 hours of ICU admission

25 (22)

17 (28)

8 (15)

Within 3 days of ICU admission

16 (14)

13 (21)

3 (6)

Once haemodynamically stable within 5 days

8 (7)

2 (3)

6 (11.5)

Once gastric residual volume below threshold

1 (1)

0 (0)

1 (2)

Depends on individual dietitian/doctor

10 ((9)

0 (0)

10 (19)

Most common route, N Nasogastric Orogastric Main rationale for postpyloric feeding, N

113

61

52

109 (96)

58 (95)

51 (98)

4 (4)

3 (5)

1 (2)

109

60

49

Only when gastric delivery is unsuccessful

7 (6)

2 (3.3)

5 (10)

Following gastric surgery

46 (42)

19 (31)

27 (55)

In pancreatitis

9 (8)

9 (15)

0 (0)

Hardly ever use it

45 (41)

28 (46)

17 (35)

When gastric emptying delayed for long periods Procedure used for tube insertion, N

2 (2)

2 (3.3)

0 (0)

111

58

53

Blind, bedside placement by nurses

6 (5)

5 (9)

1 (2)

Blind, bedside placement by dietitian/doctor

91 (82)

43 (74)

48 (91)

Fluoroscopy-assisted

3 (3)

3 (5)

0 (0)

In the operating theatre

11 (10)

7 (12)

4 (8)

Procedure used for checking feeding tube position

Intergroup p-value*

0.020

114

61

53

Chest X-ray

45 (39)

30 (49)

15 (19)

Auscultation of injected air

26 (23)

4 (6.6)

22 (41.5)

0.000

pH measurement/litmus paper

16 (14)

3 (5)

13 (24.5)

0.000

Aspiration of bile-stained fluid

3 (3)

0 (0)

3 (6)

Unsure

24 (21)

24 (39)

0 (0)

Pattern of feed delivery, N Continuous over 24 hours without any breaks

113

61

53

52 (46)

36 (59)

16 (30)

0.001 0.020

Continuous, with short holds for tolerance check

45 (40)

21 (34)

24 (45)

Continuous daytime, stopped during the night

7 (6)

2 (3)

5 (9)

Depends on individual dietitian/doctor

7 (6)

0 (0)

7 (13)

According to judgement of nurse

3 (3)

2 (3)

1 (2)

Frequency of changing feeding tube, N Only if clinically indicated

112

60

52

60 (54)

26 (43)

34 (65)

Routine daily

14 (13)

10 (17)

4 (7.6)

Unsure

24 (21)

24 (40)

0 (0)

According to judgement of nurse Frequency of changing feed giving set, N

14 (13)

0 (0)

14 (27)

113

61

52

Daily

72 (64)

33 (54)

39 (75)

Every time a new feed package is hung up

18 (16)

8 (13)

10 (19)

Randomly (no specific routine)

5 (4)

1 (1.6)

4 (8)

Unsure

19 (17)

19 (31)

0 (0)

Frequency of checking feed tolerance, N

0.015

114

61

53

Every 4 - 6 hours throughout ICU stay

24 (21)

24 (39)

0 (0)

Every 4 - 6 hours only until feeds established

58 (51)

17 (28)

41 (77)

0.010

0.008

0.000 0.000 Continued ...

46

SAJCC November 2015, Vol. 31, No. 2

Table 3 (continued). Enteral NSPs

Once daily

Total group, n (%)

Dietitians and doctors, n (%)

Nurses, n (%)

Intergroup p-value*

24 (21)

17 (28)

7 (13)

0.040

Unsure

3 (3)

3 (5)

0 (0)

No standard – as requested by doctor

1 (1)

0 (0)

1 (2)

Done by doctor/dietitian on ward round Method of checking feed tolerance, N

2 (2)

0 (0)

2 (4)

112

61

51

Gastric residual volume (GRV)

61 (54)

14 (23)

47 (92)

Gastrointestinal (GI) symptoms

2 (2)

2 (3)

0 (0)

Combination of GRV and GI symptoms

35 (31)

35 (57)

0 (0)

Combined GRV, GI symptoms and intra-abdominal pressure monitoring

10 (9)

6 (10)

4 (8)

Unsure GRV used to guide enteral feeding, N

4 (4)

4 (7)

0 (0)

40

14

26

GRV not used at all to guide practice

17 (43)

14 (100)

3 (11.5)

No specific threshold applied

5 (13)

0 (0)

5 (19)

Different doctors/dietitians use different volumes Methods of optimising enteral nutrition, N

18 (45)

0 (0)

18 (69)

108

59

49

Routine use of gastric acid suppression agents

9 (8)

3 (5)

6 (12)

Routine use of prokinetic agents

17 (16)

9 (15)

8 (16)

0.000 0.000

<0.050

Use of above medications when indicated

38 (35)

27 (46)

11 (22)

0.009

Use of above medications on nurse request

23 (21)

7 (12)

16 (33)

0.008

Change to another enteral feed Frequency of monitoring enteral nutrition support, N

21 (19)

13 (22)

8 (16)

105

59

46

Daily

88 (84)

56 (95)

32 (69.5)

Only as clinically indicated

7 (6)

2 (3.3)

5 (11)

Randomly

4 (4)

1 (2)

3 (6.5)

Depends on individual dietitian/doctor

6 (6)

0 (0)

6 (13)

Methods of enteral nutrition monitoring, N Compliance with enteral feeding protocols

107

59

48

2 (2)

1 (1.6)

1 (2)

Compliance with prescribed product

1 (1)

0 (0)

1 (2)

Compliance with prescribed rate

1 (1)

0 (0)

1 (2)

Compliance with nutritional goals

3 (3)

1 (1.6)

2 (4)

Clinical signs/symptoms of intolerance

5 (5)

0 (0)

5 (10)

Combination of the above methods

88 (81)

56 (95)

32 (67)

Unsure

7 (6)

1 (1.6)

6 (12.5)

0.0005

0.0005

*In all cases, where no p-value is shown the intergroup difference is not significant.

of parenteral nutrition monitoring were different be­t ween the reports of nurses and those of doctors, with 24% of nurses reporting uncertainty about such practice aspects.

Types of nutrition products used Types of enteral products and the frequency of use are presented in Figs 3 and 4. Of the responses from dietitians and doctors, 49% stated that no standard starter feed was used. For nurses, 12% of respondents reported being unsure of the features of different enteral feed types.

The most important criteria for enteral product choice were patient tolerance (97%), clinical appropriateness of product features (89%), followed by availability on the hospital pharmacy formulary (38%). Fac­­ tors considered unimportant for determining product choice were cost (43%), pharmacist preference (73%) and manufac­turer service support (29%). For the nurses subgroup, 85% reported that they were unsure of the criteria for choice of enteral and parenteral feed products.

Parenteral products were found to be mainly premixed industry compounded bags (65%) followed by multichamber bags (30%), while the remaining respondents reported use of both types of parenteral products. Factors considered most important in parenteral product selection were that nutrition requirements were met (91%), clinical appropriateness for the patient (84%) and an appropriate electrolyte profile (78%). Among nurses (42%) but not dietitians (14%), an impor­ tant factor for parenteral product choice was a diagnosis matched to information on product marketing material. For 47% of respondents, cost was not an important criterion for parenteral product choice.

Perceptions and self-reports of skill in NSPs Dietitians and doctors self-rated their competence in ICU nutrition support as expert (21%) and above average (70%). Similarly, nurses’ reported perceptions of dietitians’ nutrition competence was expert (61%) and above average (23%), while their perception of the skill SAJCC November 2015, Vol. 31, No. 2

47

Table 4. Parenteral NSPs

Typical timing of parenteral nutrition initiation, N

Total group, n (%)

Dietitians and doctors, n (%)

Nurses, n (%)

100

53

47

Within 24 hours of ICU admission

20 (20)

8 (15)

12 (26)

As soon as clinically indicated regardless of timing

30 (30)

16 (30)

14 (30)

Within 3 days if enteral feeding unsuccessful

7 (7)

3 (6)

4 (8.5)

After 7 days if enteral feeding unsuccessful

3 (3)

3 (6)

0 (0)

Immediately in malnourished patients

7 (7)

3 (6)

4 (8.5)

When clinically indicated >3 days of unsuccessful enteral

19 (19)

14 (26)

5 (11)

When clinically indicated >7 days of unsuccessful enteral

4 (4)

4 (7.5)

0 (0)

When clinically indicated in malnourished patients

2 (2)

2 (4)

0 (0)

Depends on dietitian/doctor

8 (8)

0 (0)

8 (17)

100

54

46

Provide requirements only when enteral nutrition not possible

75 (75)

47 (87)

28 (61)

Provide requirements in all ICU patients

8 (8)

3 (5.5)

5 (11)

Usual rationale for parenteral nutrition, N

Supplement enteral nutrition in all malnourished patients

1 (1)

0 (0)

1 (2)

Supplement enteral when requirements not met

12 (12)

3 (5.5)

9 (19.5)

Routine to start with parenteral in GIT surgery

4 (4)

1 (2)

3 (6.5)

101

54

47

Typical duration of parenteral nutrition, N 0 - 5 days

53 (53)

26 (48)

27 (57)

7 days

31 (31)

20 (37)

11 (23)

>7 days

16 (16)

8 (15)

8 (17)

>14 days

1 (1)

0 (0)

1 (2)

Frequency of monitoring of parenteral nutrition, N Daily

100

54

46

60 (60)

38 (70)

22 (48)

1 - 3 times weekly

18 (18)

13 (24)

5 (11)

Randomly, as clinically indicated

10 (10)

2 (4)

8 (17)

Unsure

12 (12)

1 (2)

11 (24)

100

53

47

Methods of parenteral nutrition monitoring, N Compliance with parenteral feeding protocol

2 (2)

0 (0)

2 (4)

Compliance with prescribed product

1 (1)

0 (0)

1 (2)

Compliance with prescribed rate

2 (2)

0 (0)

2 (4)

Compliance with nutritional goals

3 (3)

0 (0)

3 (6)

Clinical signs/symptoms of intolerance

1 (1)

1 (2)

0 (0)

Combination of above methods

82 (82)

51 (96)

31 (66)

Unsure

9 (9)

1 (2)

8 (17)

Intergroup p-value*

0.0065

0.0300

0.0250

0.0010

0.0005

*In all cases, where no p-value is shown the intergroup difference is not significant.

of doctors in this discipline was expert (19%) and above average (43%). However, 38% of nurses rated doctors working in their units as below average or totally unskilled in nutrition support of critically ill patients. Nurses’ self-report of their own skills indicated mainly average/ satisfactory (40%) and above average (47%) competence.

Discussion This is the first comprehensive survey of ICU nutrition practices to be performed in SA. The data from this pilot study showed a high selfconfidence in self-reported nutrition support competency among dietitians working in private ICUs, matched by a perception from nursing staff that dietitians are expertly skilled in this practice. In contrast, nurses perceived their own nutritional skill to be lower than that of dietitians, but also reported a considerable lack of perceived

48

SAJCC November 2015, Vol. 31, No. 2

nutrition support competency among doctors working in critical care units. These varied competency levels may be problematic for the integrated delivery of nutrition support in ICU. Critical care nutrition is not a core entry-level skill for health pro­ fessionals in SA and there is no regulatory requirement for post­basic qualification or certification. Varied levels of practitioner expertise be­ tween healthcare professionals qualified in different disciplines empha­sise the need for good multidisciplinary cooperation in order to achieve nutrition support goals, and positively affect clinical outcomes.

In the context of these perceived skill disparities among various healthcare professionals, less than half of the ICUs surveyed reported the operation of a formal nutrition support team within the unit.

There was also a significant discrepancy between reports by nurses and dietitians in this regard, with nurses being significantly more likely

Standard feed Standard with fibre Semi-elemental High protein High energy Disease-specific formulation Immune-supporting

Fig. 3. Types of enteral formulas administered.

100

Never Rarely Sometimes

80

Often

Responses (%)

Always

60

40

20

Im

m

ar d nd

re fib w

da rd

St a

m le i-e

St an

m Se

ith

en ta l

n pr ot ei Hi gh

en er gy Hi gh

m

sp eas Di se

un e

-su

pp

or

ec

tin g

ifi c

fo r

fo rm

ul at io n

ul at io n

0

Enteral feed types

Fig. 4. Frequency of use of enteral formulations. than dietitians to report the presence of a nutrition support team, yet nurse inclusion in such a team was low, as was nurse decision-making involvement in nutrition support. Team structure and composition varied between units, but reports indicated overall that in one-third of cases the dietitian was the only member of the team, and less than half of the team compositions included nursing staff. Furthermore, discrepant responses from dietitians and nurses with regards to the practicalities of NSP, including tube management, delivery patterns and monitoring patterns, as well as reports from both subgroups of uncertainty around such

practices suggest that integrated teamwork is not a matured phenomenon. It is therefore questionable whether truly multidisciplinary nutrition support teams are in fact operation­ al in private ICUs in SA. Low nurse involvement in whatever teams may exist is troubling. Clearly, nurses are expected to implement nutrition support in critical care units. It has been reported that nurses face significant challenges in achieving this, but are frequently blamed for poor nutrition support success. Some of the challenges facing ICU nurses include the fact that nutrition support guidelines do not address relevant nursing issues, or conflict

with nursing guidelines. [13] Nutrition practice guidelines are often not tailored to nursespecific barriers to nutrition delivery, and this is exacerbated by a lack of multidisciplinary collaboration and dietetic input that supports rather than criticises nursing constraints, particularly after working hours. [2] For example, in this survey dietitians and doctors reported a high adherence to published guidelines in ICU NSP, while few nursing staff seemed aware of such guidelines and were more likely to understand (or possibly have observed) nutrition support decision-making as emanating from the clinical judgements of dietitians or doctors. Additionally, nurses seemed to have low involvement in important aspects of feed implementation, such as tube feed placement, while reporting a reliance on their own judgement for various aspects of nutrition support management. Furthermore, differences were seen in nurses’ reported practices and dietitians’ perceptions of nursing practices. Dietitian and nurse subgroups both reported that they were unaware or unsure of what the other was doing with regard to certain aspects of nutrition support. This does not appear consistent with team management of ICU nutritional care, and does not suggest optimal enabling of nurses to implement nutrition support at a high standard.

One strategy to foster quality integration of nutrition support in ICUs is the use of formal, written nutrition support protocols. Large-scale cohort studies[14] have shown that ICUs that use nutrition protocols have superior nutrition adequacy in terms of delivery of requirements, success of enteral nutrition, achievement of early enteral feeding and methods to optimise enteral nutrition. These SA pilot survey data revealed that approximately two-thirds of ICUs utilised formal, written nutrition support protocols, of which only about half governed both enteral and parenteral nutrition support.

Despite reported use of formal protocols and the reported compliance with published nutrition support guidelines, there was a high reported reliance on clinical judgement to guide practice. There was also was a disjunc­ ture between aspects of nutrition practice (rationale, initiation, delivery and monitoring) as reported by dietitians and by nurses. In addition, practice inconsistencies revealed in other sections of the survey instrument suggest that the reported use of nutrition support protocols may not hold true in practice.

Importantly, nurse reports indicated that NSPs varied with different individual dietitians and doctors working in the ICU, suggesting that an overarching, protocolised approach to nutrition support did not occur in units.

SAJCC November 2015, Vol. 31, No. 2

49

For both enteral and parenteral nutrition practices, compliance with recommendations varied in how well they matched with currently available guidelines. For enteral nutrition, initiation decisions and feed delivery pattern generally matched the guidelines, and most units reported adherence to timing of early enteral nutrition given haemodynamic stability of the patient.[1,15] However, rather a high importance was placed on proof of gut function, which may be inappropriate in the critically ill and is not supported by published recommendations.[1] There was a very low utilisation of postpyloric feeding even when enteral feeding was unsuccessful due to extended delays in gastric emptying. This is unfortunate, since postpyloric feeding is a specific strategy to optimise enteral feeding success. [1,15,16] Data also showed excessive use of semi-elemental formulae and very low utilisa­ tion of standard formulations compared with international usage data,[9] and also in contradiction to enteral feed guidelines.[15] The nursing subgroup had an undue reliance on gastric residual volumes as a method of checking feed tolerance, unsupported by the literature,[1] and there was a low compliance with methods to optimise enteral feeding delivery.

This is in line with studies that have previously shown that optimisation of enteral nutrition support is generally more poorly performed than other aspects of nutrition support, such as glucose control or elevation of the head of the bed.[2,16] For parenteral nutrition practices, there was both a high reported reference to published guidelines and use of clinical judgement.

Overall rationale for parenteral use was largely appropriate, but the main determinant of parenteral nutrition initiation was reported to be the presence of gastrointestinal factors likely to make enteral feeding unsuccessful. In contrast, only 7% indicated this factor to be actually unsuccessful enteral feeding. This indicates that parenteral nutrition initia­ tion was determined by the anticipation of lack of success with enteral feeding rather than proven failure of this route of delivery, in contrast to recommendations.[1] Reported utilisation of parenteral nutrition for malnourished patients appeared low. It is unclear whether this is reflective of a lower adherence to parenteral nutrition guidelines,[1,15,17] or of the adequate nutritional status of ICU patients in private care facilities.

These survey data have limitations. Despite targeted sampling of >500 medical doctors known to work in ICUs and coverage of >60% of private hospitals with ICU facilities, the overall response from individual healthcare professionals working in ICUs was low. In particular, the re­s­ ponse from intensivists and other medical specialists was confined to just 4% of the entire sample, suggesting that nutrition may not be a priority feature of critical care management for doctors working in ICU. Given that more than a third of nurses rated the doctors working in their units as unskilled in NSPs, this perhaps is not a surprising result. However, it is impossible to comment on the reasons for the overall low response, or comment on the level of selection bias that may exist within the sample. There was also an overall reluctance among participants to share nutrition protocols with the researcher; therefore, this survey could not test participant responses against unit policies and standardised clinical practices that may have been in existence, or even verify the existence of such protocols as reported. In light of these study shortcomings, this research should be viewed as a pilot survey. A more comprehensive survey – or indeed a more robust study design – might produce different results.

Nevertheless, taken together, the data from this survey do raise questions as to the organisation and operations of formal nutrition teams and the implementation of formal nutrition protocols in private ICUs in SA. Inconsistent responses from the professional subgroups represented suggest that such structures may run less formally and effectively than may be believed or intended by the participants.There are several possible implications of this information for the critical

50

SAJCC November 2015, Vol. 31, No. 2

care nutrition discipline in SA. It may be true that there are barriers to the implementation of international guidelines in the SA context. If this is the case, then more comprehensive cross-sectional data from a more fully representative sample of prescribers and implementers of nutrition support in ICUs are required. This will allow the experience and input of SA healthcare practitioners to shape national nutrition practices and develop SA nutrition practice guidelines, which are appropriate to any unique demographic or resource-related factors affecting nutrition delivery in this setting. Nevertheless, international data have shown that improved multidisciplinary team approaches translate into improved quality of care. Because of the demonstrated perception that dietitians have nutrition skill not matched by that of other health professionals, there is an opportunity to enhance the positioning of the dietitian as the nutrition expert in ICUs, and build a more holistic team approach to nutritional care of the critically ill patient. There may be a need for a regulatory mechanism for practi­tioners to prove postbasic competency in critical care nutrition support. Finally, there is a need for improved knowledge-to-practice translation for all disciplines of healthcare professionals involved with nutritional care of the critically ill patient.

Acknowledgements. This research was made possible with financial aid received from Nutricia, SA and Nutricia, Middle East. The author wishes to acknowledge the scientific input of Kuba Morris, Scientific Officer of Nutricia Middle East, South-East Asia and Africa. The author also wishes to gratefully acknowledge the input of all the respondents to this survey Conflict of interest. The author is a consultant to industry and has consulted to Nutricia, SA and BBraun, SA. References 1. McClave SA, Martindale RG, Vanek VW, et al; ASPEN Board of Directors; American College of Critical Care Medicine; Society of Critical Care Medicine. Guidelines for the provision and assessment of nutrition support therapy in the adult critically ill patient: Society of Critical Care Medicine (SCCM) and American Society for Parenteral and Enteral Nutrition (ASPEN). JPEN J Parenter Enteral Nutr 2009;33(3):277-316. [http://dx.doi.org/10.1177/0148607109335234] 2. Cahill NE, Murch L, Cook D, Heyland DK on behalf of the Canadian Critical Care Trials Group.

Improving the provision of enteral nutrition in the intensive care unit: A description of a multifaceted intervention tailored to overcome local barriers. Nutr Clin Pract 2014;29(1):110117. [http://dx.doi.org/10.1177/0884533613516512] 3. Farber M, Moses J, Korn M. Reducing costs and patient morbidity in the enterally fed intensive care unit patient. JPEN J Parenter Enteral Nutr 2005;29(1 Suppl):S62e9. [http://dx.doi.org/10.11 77/01486071050290s1s62] 4. Heyland D, Schroter-Noppe D, Drover J, et al. Nutrition support in the critical care setting: Current practice in Canadian ICUs – opportunities for improvement? JPEN J Parenter Enteral Nutr 2003;27(1):74e83. [http://dx.doi.org/10.1177/014860710302700174] 5. Thomas DR. Starving in the hospital. Nutrition 2003;19(10):907e8. [http://dx.doi.org/10.1016/ s0899-9007(03)00169-2] 6. Villet S, Chiolero R, Bollmann M, et al. Negative impact of hypocaloric feeding and energy balance on clinical outcome in ICU patients. Clin Nutr 2005;24(4):502e9. [http://dx.doi.org/10.1016/j. clnu.2005.03.006] 7. Sheean PM, Peterson SJ, Gurka DP, Braunschweig CA. Nutrition assessment: The reproducibility of Subjective Global Assessment in patients requiring mechanical ventilation. Eur J Clin Nutr 2010;64(11):1358-1364. [http://dx.doi.org/10.1038/ejcn.2010.154] 8. Shpata V, Prendushi X, Kreka M, Kola I, Kurti F, Ohri I. Malnutrition at the time of surgery affects negatively the clinical outcome of critically ill patients with gastrointestinal cancer. Med Arch 2014;68(4):263-267. [http://dx.doi.org/10.5455/medarh.2014.68.263-267] 9. Roynette CE, Bongers A, Fulbrook P, Albarran JW, Hofman Z. Enteral feeding practices in European ICUs: A survey from the European federation of critical care nursing associations (EfCCNa). E Spen Eur E J Clin Nutr Metab 2008;3:e33ee39. [http://dx.doi.org/10.1016/j.eclnm.2007.10.004] 10. Scribante J, Bhagwanjee S. National audit of critical care resources in South Africa – open versus closed intensive and high care units. S Afr Med J 2007;97(12):1319-1322.

11. Hodgson RE. The management of death in the ICU. The South African perspective. Crit Care Resusc 2006;8:73-75.

12. Mathivha LR. ICUs worldwide: An overview of critical care medicine in South Africa. J Crit Care 2002;6(1):22-23. [http://dx.doi.org/10.1186/cc1449] 13. Marshall AP, Cahill N, Gramlich L, MacDonald G, Alberda C, Heyland DK. Optimising nutrition in intensive care units: Empowering critical care nurses to be effective agents of change. Am J Crit Care 2012;21(3):186-194. [http://dx.doi.org/10.4037/ajcc2012697] 14. Heyland DK, Cahill N, Dhaliwal R, Sun X, Day AG, McClave SA. Impact of enteral feeding protocols on enteral nutrition delivery: Results of a multicenter observational study. JPEN J Parenter Enteral Nutr 2010;34(6):675-684. [http://dx.doi.org/10.1177/0148607110364843] 15. Kreymann KG, Berger MM, Deutz N, et al. ESPEN Guidelines on Enteral Nutrition: Intensive care.

Clin Nutr 2006;25(2):210-223. [http://dx.doi.org/10.1016/j.clnu.2006.01.021] 16. Dhaliwal R, Cahill N, Lemieux M, Heyland DK. The Canadian critical care nutrition guidelines in 2013: An update on current recommendations and implementation strategies. Nutr Clin Pract 2014;29(1):29-43. [http://dx.doi.org/10.1177/0884533613510948.6] 17. Singer P, Berger MM, van den Berghe G, et al. ESPEN Guidelines on Parenteral Nutrition: Intensive care. Clinical Nutrition 2009;28(4):387-400. [ttp://dx.doi.org/10.1016/j.clnu.2009.04.024]

ARTICLE

Comparison of the efficacy of colistin monotherapy and colistin combination therapies in the treatment of nosocomial pneumonia and ventilator-associated pneumonia caused by Acinetobacter baumannii I Kara,1 MD; F Yildirim,2 MD; B Bilaloglu,3 MD; D Karamanlioglu,4 MD; E Kayacan,3 MD; M Dizbay,4 MD; M Turkoglu,3 MD; G Aygencel,3 MD ¹ Intensive Care Fellowship Programme, Department of Anesthesiology and Reanimation, Faculty of Medicine, Gazi University, Ankara, Turkey ² Intensive Care Fellowship Programme, Department of Pulmonary Medicine, Faculty of Medicine, Gazi University, Ankara, Turkey ³ Division of Intensive Care Medicine, Department of Internal Medicine, Faculty of Medicine, Gazi University, Ankara, Turkey ⁴ Department of Infectious Diseases, Faculty of Medicine, Gazi University, Ankara, Turkey Corresponding author: I Kara (driskenderkara@gmail.com)

Objective. To investigate whether there was a difference in mortality, clinical response and bacterial eradication between colistin monotherapy and colistin combination therapies for the treatment of nosocomial pneumonia/ventilator-associated pneumonia (VAP) caused by Acinetobacter baumannii in a medical intensive care unit (ICU).

Methods. This retrospective, observational and single-centre study included all patients who were in the medical ICU of Gazi University Medical Faculty Hospital and diagnosed with nosocomial pneumonia/VAP caused by A. baumannii between January 2009 and September 2014.

Results. The median age of the 134 patients was 68 years and 53.3% were male. The most common causes of admission were respiratory insufficiency (66.7%) and sepsis/septic shock (54.8%). In patients with nosocomial pneumonia/VAP caused by A. baumannii, on median day 5 of admission, colistin monotherapy was used in 23 (21.6%) patients, a carbapenem combination was used in 80 (59.7%) patients, sulbactam-ampicillin combination was used in 42 (31.4%) patients, tigecycline combination was used in 26 (19.4%) patients, and sulbactam-cefoperazone combination was used in 17 (12.7%) patients. Median ICU stay of the patients was 15.5 days, and 112 (83.6%) patients died. Colistin monotherapy and combination therapies had no superiority over each other in clinical response for the treatment of A. baumannii-associated nosocomial pneumonia/VAP. Mortality was found to be higher in patients receiving the colistin-carbapenem combination (64.3% v. 36.4%, p=0.016). Discharge/day-of-death Sequential Organ Failure Assessment score (odds ratio (OR) 2.017, 95% confidence interval (CI) 1.330 - 3.061) and vasopressor use (OR 9.014, 95% CI 1.360 - 59.464) were independent risk factors for ICU mortality.

Conclusion. Colistin monotherapy and combination therapies have no superiority over each other for clinical response in the treatment of nosocomial pneumonia/VAP caused by multidrug-resistant A. baumannii. Colistin-SAM was associated with improved microbiological eradication and colistin-carbapenem combination was associated with increased mortality.

S Afr J Crit Care 2015;31(2):51-58. DOI:10.7196/SAJCC.2015.v31i2.246

Acinetobacter baumannii is an aerobic, Gramnegative, non-motile, lactose-negative, oxidasenegative, catalase-positive coccobacillus, and is a member of the Moraxellaceae family. [1,2] Over the last 2 decades, nosocomial infections caused by multidrug-resistant (MDR) A. baumannii (MDR-AB) have increased globally. [3] Infection by this agent increases mortality, morbidity and costs. Mortality rates can increase by up to 65% when pneumonias are caused by MDR-AB. [4] In critically ill patients who are in intensive care units (ICUs), MDR-AB infections are often difficult to treat since they are susceptible to only a limited number of antibiotics. Colistin, a polymyxin antibiotic, is one of the most common antibiotics used to treat MDR-AB infections. It is an important treatment option not just for MDR-AB infections, but also for other MDR Gram-negative bacterial infections. [5] Colistin was first introduced in 1952 and was used routinely until the 1980s. Its use was then abandoned for a period owing to serious side-effects such as nephrotoxicity and neurotoxicity; however, it

then became popular again with the emergence of life-threatening MDR Gram-negative infections. Colistin, which is almost the only available treatment option for MDR Gram-negative infections, is often used in combination therapies in order to reduce resistance and improve efficacy. [6,7] The primary objective of the study was to determine whether there was a mortality difference between colistin and colistin combination therapies in MDR-AB nosocomial pneumonia/ ventilator-associated pneumonia (VAP). Secondary objectives were to investigate clinical response, bacterial eradication and nephrotoxicity in colistin monotherapy and colistin combinations in MDR-AB associated pneumonia/VAP.

Methods This retrospective, observational, single-centre study was conducted in the medical ICU of Gazi University Medical Faculty Hospital.

Patients who were diagnosed with MDR-AB-associated nosocomial pneumonia/VAP between January 2009 and September 2014 were

SAJCC November 2015, Vol. 31, No. 2

51

included. Patients’ demographic characteristics, underlying diseases, other infection sites, dosage of colistin and antibiotics used in colistin combinations, treatment durations, clinical responses, bacterial eradication and mortality rates were recorded. The study was approved by the Ethics Committee of Gazi University Medical Faculty (date: 12 January 2015; serial no: 25901600-79; decision number: 05).

The study included patients: who were diagnosed with nosocomial pneumonia/VAP caused by MDR-AB as detected in endotracheal aspirates, bronchoalveolar lavage or sputum cultures; who received colistin/colistin combination therapy for at least 72 hours; and who received proper and efficient treatment for other concomitant infection sites and agents. If there was more than one nosocomial pneumonia/ VAP episode caused by MDR-AB during the follow-up of the patient, data of the first infection episode were included in the study.

The study excluded patients: who received colistin therapy for <72 hours; in whom colistin use was contraindicated; and who were under the age of 18.

The Centers for Disease Control and Prevention criteria were used to diagnose nosocomial pneumonia/VAP.[8] The severity of the patient’s clinical condition at admission was determined using Acute Physiology and Chronic Health Evaluation (APACHE) II and Sequential Organ Failure Assessment (SOFA) scores, and the severity of underlying diseases was established using the Charlson Comorbidity Index (CCI).[5]

Criteria for evaluating clinical response to colistin therapies Clinical responses were evaluated at the end of the treatment.

• Complete clinical response (cure): The improvement of all symptoms, signs and laboratory values related to the infection (such as fever, hypothermia, tachypnoea, tachycardia, leukocytosis, leukopenia, C-reactive protein and procalcitonin).

• Partial clinical response: The improvement of only a portion of the symptoms, signs,and laboratory values observed at the beginning of the infection.

• Treatment failure: The non-improvement or worsening of the symptoms, signs and laboratory values related to the infection despite the antimicrobial treatment.[5,6] • Microbiological eradication: Throughout treatment, microbio­ logical sampling was conducted every 4 days. Eradication was con­sidered as bacterial absence in two consecutive samples from respiratory tract secretions during or at the end of the treatment.[9] Respiratory tract specimens taken from the ICU patients were sent to the Infectious Diseases and Clinical Microbiology Laboratory of our hospital. Here, the samples were inoculated into blood agar and eosin methylene-blue lactose sucrose agar, and the antibiotic susceptibilities were studied in Mueller-Hinton agar medium using the Kirby-Bauer Disk Diffusion method in accordance with the standards of Clinical and Laboratory Standards Institute for microorganisms growing after 24 - 48 hours. In addition to the classical methods for bacteria identification, BBL Crystal Enteric/ Non-Fermentative ID and BBL Crystal Gram-Positive ID kits (Becton Dickinson, USA) were used. The agents/microorganisms and antibiogram results were recorded.

Renal toxicity Renal toxicity was defined according to the RIFLE (risk, injury, failure, loss and end-stage) criteria, and a basal serum creatinine

52

SAJCC November 2015, Vol. 31, No. 2

(SCr) value ≤114.9 µmol/L was considered normal renal function.

Nephrotoxicity was defined as an increase in basal SCr value >50% or the need for renal replacement therapy during the treatment.[5,10]

Statistical analysis Statistical analysis was conducted using SPSS version 17 (IBM, USA). Continuous data were expressed in median and interquartile range (IQR) values; categorical data were expressed in numbers and percentages. Continuous data of the patients who survived and died at the end of the study were compared using the Mann-Whitney U-test, and categorical data were compared using the χ2 test. Logistic regression analysis was conducted with the significant data from the univariate analysis in order to determine independent risk factors for mortality. Additionally, logistic regression analysis was conducted to determine the effect of colistin monotherapy and combination therapies on mortality. Statistical significance level was considered for p<0.05.

Results The study included 134 patients, median age 68 (58.0 - 76.3) years and 53.3% male. Median admission SOFA score was 7 (5 - 11), and median discharge/date-of-death SOFA score was 14 (8 - 17). Admission to ICU was mostly due to respiratory insufficiency (66.7%) and sepsis/septic shock (54.8%). The most common comorbidity was hypertension (52.6%). A total of 118 (87.4%) and 122 (90.4%) patients received mechanical ventilation (MV) support at admission and during the A. baumannii infection, respectively. A total of 112 (83.6%) patients died during the study period. The discharge/day-of-death SOFA score and procalcitonin value were higher in those who died (15 v. 4, p<0.0001 and 0.72 ng/mL v. 3.35 ng/mL, p<0.0001) (Table 1).

Mortality was lower in patients who received non-invasive MV (NIMV) support during infection (p=0.010), but higher in patients who: received MV support at admission (p=0.002); received invasive MV (IMV) support at admission (p<0.0001); received MV support during infection (p<0.0001); received IMV support during infection (p<0.0001); received vasopressor support; had a central venous catheter and arterial catheter during infection (p<0.0001); developed septic shock (p<0.0001); received sedation (p=0.046); had higher CCI (p=0.049); and had haematological malignancy (p=0.017) (Tables 1 and 2).

In our study, 98.5% of A. baumannii strains were susceptible to colistin. Colistin-carbapenem combination therapy was administered to 60% of the patients, and mortality rates were higher (p=0.016) in this group. Microbiological eradication was achieved in 11/42 patients receiving colistin-sulbactam/ampicillin combination therapy, and mortality was lower in these patients (p=0.007) (Table 3).

The patients were divided into two groups, namely clinically responsive (complete response/cure or partial response) and treatment failure, according to our definitions. There was no difference in clinical response between patients treated with colistin and colistin combination therapies (Table 4).

Mortality was higher in patients with: higher procalcitonin and lower albumin levels prior to the colistin/combination therapy (p=0.007 and p=0.001, respectively); lower mean arterial blood pressure levels (p<0.0001); and higher procalcitonin levels, higher pulse rates and higher creatinine levels after colistin/combination therapy (p<0.0001, p=0.009 and p=0.004, respectively). Mortality was lower in patients with higher haemoglobin and albumin levels after colistin/combination therapy (p<0.001 and p<0.0001, respectively) (Table 5).

Table 1. General characteristics of the patients Characteristics

Total (N=134), n (%)

Survived (N=22), n (%)

Died (N=112), n (%)

p-value

Age (years), median (IQR)

68 (58.0 - 76.3)

66 (47.5 - 75.0)

68.5 (60.3 - 77.8)

0.20

Sex (male)

72 (53.7)

9 (41.0)

63 (56.3)

0.24

APACHE II, median (IQR)

23 (19.0 - 27.0)

21.5 (18.8 - 26)

24 (19 - 27)

0.21

Admission SOFA, median (IQR)

7 (5 - 11)

6 (3 - 12)

8 (6 - 11)

0.09

Discharged/day-of-death SOFA, median (IQR)

14 (8 - 17)

4 (2 - 6)

15 (12 - 17)

0.0001

Admission procalcitonin (ng/mL), median (IQR)

1.9 (0.5 – 8.0)

0.8 (0.3 - 5.5)

2.0 (0.6 - 8.1)

0.25

Discharged/day-of-death procalcitonin (ng/mL), median (IQR)

2.4 (0.7 - 9.6)

0.7 (0.3 - 1.5)

3.4 (1.3 - 10.1)

<0.0001

GCS, median (IQR)

10 (7 - 14)

11 (8 - 15)

10 (7 - 14)

0.36

Emergency service

56 (41.8)

7 (31.8)

49 (43.8)

0.35

Internal medicine ward

55 (41)

9 (40.9)

46 (41.1)

1

Other wards

18 (13.4)

4 (18.2)

14 (12.5)

0.48

Other hospital

5 (3.7)

2 (9.1)

3 (2.7)

0.15

Diabetes mellitus

40 (29.8)

5 (22.7)

35 (31.3)

0.61

Cardiovascular diseases

45 (33.6)

6 (27.3)

39 (34.8)

0.62

Hypertension

71 (53.0)

9 (40.9)

62 (55.4)

0.25

Solid malignancy

23 (17.2)

2 (9.1)

21 (18.8)

0.27

Haematological malignancy

24 (18.0)

-

24 (21.4)

0.02

Chronic pulmonary diseases

29 (21.6)

5 (22.7)

24 (21.4)

0.89

Chronic kidney diseases

41 (30.6)

5 (22.7)

36 (32.1)

0.46

Liver failure

7 (5.2)

-

7 (6.3)

0.23

Neurological diseases

22 (16.4)

3 (13.6)

19 (17.0)

0.70

3 (2.0 - 4.0)

2 (1.0 - 4.0)

3 (2.0 - 4.8)

0.049

Respiratory insufficiency

90 (67.2)

13 (59.1)

77 (68.8)

0.46

Sepsis/septic shock

74 (55.2)

10 (45.5)

64 (57.1)

0.35

Cardiovascular disorders

14 (10.4)

3 (13.6)

11 (9.8)

0.75

Neurological disorders

7 (5.2)

2 (9.1)

5 (4.5)

0.37

Metabolical disorders

5 (3.7)

1 (4.5)

4 (3.6)

0.83

Gastrointestinal/hepatological disorders

9 (6.7)

1 (4.5)

8 (7.2)

0.31

Renal insufficiency

30 (22.4)

5 (22.7)

25 (22.3)

0.97

Postarrest

4 (3)

1 (4.5)

3 (2.7)

0.64

Postoperative respiratory insufficiency

1 (0.7)

-

1 (0.9)

-

Trauma

Admission site

Comorbidites

CCI, median (IQR) Main diagnosis at admission

1 (0.7)

-

1 (0.9)

-

Length of stay in hospital (days), median (IQR)

29 (17.0 - 43.8)

31 (19.0 - 41.5)

28 (16.0 - 47.5)

0.50

Length of stay in ICU (days), median (IQR)

15.5 (8.0 - 30.3)

16.5 (8.0 - 27.5)

15 (8.0 - 30.8)

0.74

Mortality at the 28th day

83 (61.9)

0 (0)

83 (74.1)

<0.0001

Mortality at the 90th day

109 (81.3)

0 (0)

109 (97.3)

<0.0001

GCS = Glasgow Coma Scale; IQR = interquartile range.

Nephrotoxicity developed in 26 (19.3%) patients on median day 6 of colistin therapy.

Mortality was lower in patients with a RIFLE score of ‘no risk’ (p=0.009) and higher in patients with a RIFLE score of ‘failure’ (p=0.004) at the end of colistin therapy (Table 5).

A logistic regression analysis was con­ ducted using data significant on univariate analysis: discharge/day-of-death SOFA score, invasively mechanically ventilated during infection, receiving vasopressors, use of colistin-carbapenem combination, microbiological eradication with colistin-

sulbactam/ampicillin, RIFLE score of ‘failure’ at the end of colistin therapy, post-treatment mean blood pressure and post-treatment albumin levels. Discharged/ day-of-death SOFA score and receiving vasopressor support were independent risk factors (p=0.001, odds ratio (OR) 2.02, 95%

SAJCC November 2015, Vol. 31, No. 2

53

Table 2. Characteristics of patients during intensive care follow-up Characteristics

Total (N=134), n (%)

Survived (N =22), n (%)

Died (N =112), n (%)

p-value

MV on admission

118 (88.1)

15 (68.2)

103 (92.0)

0.002

IMV on ICU admission

112 (83.6)

12 (54.5)

100 (89.3)

0.0001

Duration of IMV (days), median (IQR)

12 (6.0 - 30.6)

8.5 (3.3 - 28.0)

12 (7.0 - 30.8)

0.330

NIMV on ICU admission

26 (19.4)

5 (22.7)

21 (18.8)

0.660

Duration of NIMV (days), median (IQR)

2 (1.0 - 7.0)

7 (5.0 - 8.0)

2 (1.0 - 4.3)

0.016

Duration of total MV (days), median (IQR)

12 (7.0 - 30.0)

9 (5.0 - 19.0)

12.5 (7.0 - 30.3)

0.260

MV at time of infection

122 (91.0)

15 (68.2)

107 (95.5)

0.0001

Duration of MV (days), median (IQR)

8 (4.0 - 18.3)

8 (4.0 - 13.0)

9 (4.0 - 19.0)

0.690

IMV during A. baumannii infection

109 (81.3)

11 (50.0)

98 (87.5)

0.0001

NIMV during A. baumannii infection

12 (8.9)

5 (22.7)

7 (6.3)

0.010

Tracheostomised

28 (20.9)

3 (13.6)

25 (22.3)

0.360

Vasopressor support

102 (76.1)

3 (13.6)

99 (88.4)

0.0001

Central venous catheterisation

117 (87.3)

14 (63.6)

103 (92.0)

0.0001

Arterial catheterisation

104 (77.6)

4 (18.2)

100 (89.3)

0.0001

Septic shock

90 (67.2)

1 (4.5)

89 (79.5)

0.0001

Sedation

44 (32.8)

3 (13.6)

41 (36.6)

0.0460

Other characteristics

confidence interval (CI) 1.33 - 3.06 and p=0.020, OR 9.01, 95% CI 1.37 - 59.46, respectively) for ICU mortality (Table 6).

Logistic regression analysis was conducted based only on treatments in order to investigate the effects of colistin monotherapy and combination therapies on mortality. Receiving colistincarbapenem therapy was shown to be an independent risk factor for mortality (p=0.020, OR 3.28, 95% CI 1.27 - 8.48).

Discussion In ICUs, MDR-AB infection is an important health issue, and is associated with 8 - 43% mortality. [1,11,12] Pneumonia is the most common type of infection; its mortality rate can increase up to 65%[4,13] when attributed to MDR-AB. In our ICU, the mortality rate was 83.6% in 134 patients with MDR-AB-associated pneumonia/VAP. The higher mortality rates of the present study compared with those in the literature were attributed to comorbid diseases, infections and the advanced age of our patients.

Based on European surveillance data from 2009, A. baumannii infections in ICUs occured in cases of pneumonia at a rate of 21.8%, in bloodstream infections at a rate of 17.1%, and in urinary system infections at a rate of 11.9%.[3,14,15] Based on Gazi University Medical Faculty Hospital’s Infection Control Committee data of 2002, the incidence of A. baumannii as an agent for infections in all ICUs was 27.2%, and 77% of these cases are pneumonia/VAP. A study by Dizbay et al.,[11] which was conducted in our hospital, found that 80.5% of A. baumannii isolates from VAPs in our ICUs were MDR. It is becoming more difficult to treat these infections all over the world due to the emerging antibiotic resistance of A. baumannii. In the 1990s, A. baumannii resistance to ceftazidime was 32 - 45%, and 28 - 64% to amikacin, 4 - 94% to ciprofloxacin, and 0 - 2% to imipenem. [2] However, resistance has increased to 85% for ceftazidime, 90% for ciprofloxacin, 38 - 71% for imipenem, and 2.7 - 12% for tigecycline over the last decade. In contrast, colistin resistance is rarely reported. [2] In the study by Dizbay et al.,[11] the

54

SAJCC November 2015, Vol. 31, No. 2

rate of antibiotic resistance in A. baumannii infections was 95.5% for ciprofloxacin, 72.7% for cefepime, 80.3% for imipenem, 17.2% for meropenem, 68.2% for cefoperazone/sulbactam, 30.3% for netilmicin, 25.8% for tigecycline, and 0% for colistin.[11] Therefore, colistin is the most frequently used antibiotic for A. baumannii infections in our hospital.

Colistin has a bactericidal effect based on concentration. It has low pulmonary penetration, but intravenous colistin is effective in MDRAB pneumonia infections; however, the adequacy of monotherapy is doubtful.[1] Although MDR-AB strains are generally susceptible to colistin, colistin monotherapy may cause in vitro heteroresistance.

Combination therapies are recommended to avoid such resistance.

Studies comparing colistin monotherapy and combined therapies provide inconsistent results, with insufficient evidence supporting the superiority of combined therapies.[3,5,16] Colistin is usually combined with sulbactam, cephalosporins, carbapenems, piperacillintazobactam, monobactams, aminoglycosides, fluoroquinolones, rifampin, tetracyclines and tigecycline.[1,4,5,17] Two studies have achieved a positive clinical response of 73.0 80.8% and a microbiological response of 94.9% when using colistin in the treatment of MDR Gram-negative infections.[18,19] Kallel et al.[7] achieved a positive clinical response of 60% with colistin in the treatment of pneumonia associated with resistant A. baumannii and Pseudomonas aeruginosa. The clinical response rate varies between 38 and 57% and the microbiological response rate varies between 45 and 69% in pneumonia treatment with colistin therapy.[4] One study, which compared colistin doses in patients with A. baumannii infection, found a clinical response of 70.8% and a microbiological response of 62.5%.[6] Another study found an efficacy of 45 - 88% for colistin.[20] One prospective study conducted in 28 hospitals compared colistin, sulbactam, tigecycline and tetracycline monotherapies and their combinations. Monotherapy was administered to 67.3% and combination therapy was administered to 32.7% of 101 patients included in the study. The study concluded no difference between

Table 3. Antibiotic susceptibility, treatment modalities, clinical response and bacterial eradication Characteristics

Total (N=134), n (%)

Survived (N =22), n (%)

Died (N =112), n (%)

p-value

Day of A. baumanii isolation after admission, median (IQR)

5 (2.8 - 11.0)

4.5 (1.0 - 9.5)

5 (3.0 - 11.0)

0.23

Duration of colistin use (days), median (IQR)

9 (4 - 17)

11 (7 - 18)

9 (4 - 17)

0.46

Antibiotic susceptibility of A. baumanni Colistin

132 (98.5)

22 (100)

110 (98.2)

0.53

Tigecycline

60 (44.8)

8 (36.4)

52 (46.4)

0.48

Sulbactam/ampicillin

18 (13.4)

2 (9.1)

16 (14.3)

0.51

Aminoglycosides

43 (32.1)

6 (27.3)

37 (33)

0.80

Carbapenems

6 (4.5)

-

6 (5.4)

0.27

Quinolones

8 (5.9)

2 (9.1)

6 (5.4)

0.50

Trimethoprim/sulphamethoxazole

5 (3.7)

1 (4.5)

4 (3.6)

0.83

Antibiotic treatment Colistin monotherapy

23 (17.2)

3 (13.6)

20 (17.9)

0.63

Colistin-carbapenems

80 (59.7)

8 (36.4)

72 (64.3)

0.02

Colistin-sulbactam/ampicillin

42 (31.3)

7 (31.8)

35 (31.3)

1

Colistin-tigecycline

26 (19.4)

2 (9.1)

24 (21.4)

0.18

Colistin-cefoperazone/sulbactam

17 (12.7)

2 (9.1)

15 (13.4)

0.45

Duration of antibiotic treatment (days), median (IQR) Colistin monotherapy

4 (2.0 - 7.0)

4 (2.0 - 7.0)

4 (2.0 - 9.5)

0.93

Colistin-carbapenems

7 (3.0 - 11.0)

14 (8.5 - 15.0)

6 (3.0 - 9.5)

0.01

Colistin-sulbactam/ampicillin

7.5 (4.0 - 12.0)

8 (6.0 - 8.0)

7 (4.0 - 13.0)

0.93

Colistin-tigecycline

5 (2.0 - 12.0)

8.5 (6.0 - 11.0)

4 (2.0 - 12.0)

0.53

Colistin-cefoperazone/sulbactam

6 (3.0 - 12.0)

3 (2.0 - 4.0)

6 (3.5 - 13.5)

0.12

Bacterial eradication rates Colistin monotherapy (n=23)

3 (2.2)

-

3 (2.7)

0.44

Colistin-carbapenems (n=80)

7 (5.2)

-

4 (3.6)

0.05

Colistin-sulbactam/ampicillin (n=42)

11 (8.2)

5 (22.7)

6 (5.4)

0.007

Colistin-tigecycline (n=26)

2 (1.5)

1 (4.5)

1 (0.9)

0.20

Colistin-cefoperazone/sulbactam (n=17)

-

-

-

-

Table 4. Effectiveness of antibiotics in the treatment of patients

Antibiotic

Clinical responsiveness (complete and partial response), n (%)

Treatment failure, n (%)

Colistin monotherapy (n=23)

6 (26.1)

17 (73.9)

Colistin-carbapenem (n=80)

21 (26.2)

59 (73.8)

Colistin-sulbactam/ampicillin (n=42)

14 (33.3)

28 (66.7)

Colistin-tigecycline (n=26)

7 (27.0)

19 (73.0)

Colistin-cefoperazone/sulbactam (n=17)

5 (29.4)

12 (70.6)

Colistin-sulbactam (n=53)

19 (35.8)

34 (64.2)

monotherapy and combined therapies in sepsis caused by resistant A. baumannii.[3] In our study, colistin monotherapy was used at a rate of 17.2% and combination therapies were generally preferred. Clinical response rate varied between 26.1 and 33.3% in colistin and its combinations, which is very low. No significant difference was found in

clinical response between colistin and its combinations.

Although there are many studies on the combined use of colistin in the literature, very different results have been reported with similar combinations. There are some studies that report no additional benefit provided by combination therapy,

while others report superiority for combined therapy.[5] One study did not find any difference in 28-day mortality between the groups that received colistin-sulbactam, colistin-tigecycline, and colistin-carbapenem. [4] Another study found higher treatment efficacy and survival rates in combined therapy group than monotherapy group; however, there was no significant difference when comparing the combination groups among themselves.[5] In contrast, two randomised controlled trials found no difference in mortality and cure rates despite the higher microbiological eradication rate with combined therapy, when comparing colistin and colistin-rifampicin for treating VAP and several severe infections.â&#x20AC;&#x201A;[3,21,22] Another study reported no clinical or statistical difference between monotherapy and combined therapy for the treatment

SAJCC November 2015, Vol. 31, No. 2

55

Table 5. Characteristics of patients, renal functions and colistin nephrotoxity Total (N=134), n (%)

Survived (N=22), n (%)

Died (N=112), n (%)

p-value

Dialysis before colistin treatment

57 (42.5)

6 (27.3)

51 (45.5)

0.16

Dialysis during colistin treatment

69 (51.5)

7 (31.8)

62 (55.4)

0.06

Colistin-related nephrotoxicity

26 (19.4)

2 (9.1)

24 (21.4)

0.18

ICU day that colistin nephrotoxicity developed

6 (3.0 - 11.5)

10 (10.0 - 10.0)

6 (3.0 - 12.0)

0.47

Daily dose of colistin (mg), median (IQR)

150 (150.0 - 287.5)

150 (150.0 - 300.0)

150 (150.0 - 275.0)

0.69

Total dose of colistin (mg), median (IQR)

1 350 (450.0 - 2 681.0)

1 275 (0 - 2 587.5)

1 350 (450.0 - 2 793.7)

0.46

No risk

33 (24.6)

7 (31.8)

26 (23.2)

0.41

Risk

34 (25.4)

7 (31.8)

27 (24.1)

0.43

Injury

17 (12.7)

1 (4.5)

16 (14.3)

0.21

Failure

24 (17.9)

2 (9.1)

22 (19.6)

0.24

Lost

2 (1.5)

-

2 (1.8)

0.53

End stage

24 (17.9)

5 (22.7)

19 (17.0)

0.52

No risk

30 (22.4)

8 (36.4)

22 (19.6)

0.09

Risk

23 (17.2)

5 (22.7)

18 (16.1)

0.45

Injury

14 (10.4)

2 (9.1)

12 (10.7)

0.82

Failure

34 (25.4)

2 (9.1)

32 (28.6)

0.06

Lost

6 (4.8)

-

6 (5.4)

0.267

End stage

27 (20.1)

5 (22.7)

22 (19.6)

0.74

No risk

19 (14.2)

7 (31.8)

12 (10.7)

0.009

Risk

21 (15.7)

6 (27.3)

15 (13.4)

0.10

Injury

16 (11.9)

2 (9.1)

14 (12.5)

0.65

Failure

40 (29.9)

1 (4.5)

39 (34.8)

0.004

Lost

9 (6.7)

1 (4.5)

8 (7.1)

0.66

End stage

29 (21.6)

5 (22.7)

24 (21.4)

0.89

Leukocytes (/µL)

9 900 (6 630 - 14 605)

10 975 (8 240 - 16 887)

9 880 (5 980 - 14 410)

0.21

Haemoglobin (g/L)

88.0 (77.0 - 104.0)

98.5 (85.5 - 106.0)

86.0 (76.0 - 104.0)

0.05

C-reactive protein (mg/L)

1 315 (415 - 1 945)

1 660 (845 - 2 810)

1 230 (250 - 1 930)

0.23

Creatinine (µmol/L)

145.8 (70.7 - 265.2)

92.8 (57.4 - 322.6)

167.9 (70.7 - 265.2)

0.27

Albumin (g/L)

24.0 (20.4 - 28.0)

26.5 (24.8 - 32.0)

23.0 (20.0 - 27.0)

0.001

Procalcitonin (ng/mL)

1.4 (0.5 - 5.2)

0.5 (0.1 - 2)

1.6 (0.6 - 6.3)

0.007

Respiratory rate (/min)

26.0 (22.0 - 30.0)

25.5 (22.0 - 29.3)

26.0 (22.0 - 30.0)

0.84

Temperature (°C)

36.7 (36.5 - 37.5)

36.8 (36.7 - 37.6)

36.7 (36.4 - 37.5)

0.53

Mean arterial pressure (mmHg)

65.0 (57.5 - 73.0)

73.5 (67.75 - 86.5)

63.0 (55.0 - 72.0)

0.0001

Heart rate (/min)

110.0 (98.0 - 125.0)

107.5 (94.3 - 112.5)

110.0 (98.0 - 172.0)

0.12

Leukocytes (/µL)

10 420 (7 195 - 15 355)

9 820 (7 570 - 11 320)

10 765 (6 825 - 15 945)

0.55

Haemoglobin (g/L)

82.5 (72.0 - 95.0)

95.0 (84.7 - 108.5)

80.0 (71.0 - 93.2)

0.001

C-reactive protein (mg/L)

870 (230 - 1 350)

440 (201 - 1 375)

910 (230 - 1 350)

0.51

Creatinine (µmol/L)

185.6 (97.2 - 307.6)

97.2 (53.0 - 179.0)

203.3 (101.6 - 313.8)

0.004

Albumin (g/L)

21 (18 - 25)

27 (25 - 30)

20 (17 - 23)

0.0001

Procalcitonin (ng/ml)

2.4 (0.5 - 9.6)

0.4 (0.1 - 0.7)

4.1 (1.3 - 11.0)

RIFLE on ICU admission

RIFLE at the begining of colistin treatment

RIFLE at the end of colistin treatment

Laboratory and vital signs Before colistin treatment, median (IQR)

After colistin treatment, median (IQR)

0.0001 Continued ...

56

SAJCC November 2015, Vol. 31, No. 2

Table 5 (continued). Characteristics of patients, renal functions and colistin nephrotoxity Total (N=134), n (%)

Survived (N=22), n (%)

Died (N=112), n (%)

p-value

Respiratory rate (/min)

24.0 (20.0 - 28.0)

22.0 (20.0 - 25.3)

24.0 (20.0 - 28.0)

0.13

Temperature (°C)

36.6 (36.2 - 37.2)

36.6 (36.4 - 36.7)

36.6 (36.2 - 37.3)

0.52

Mean arterial pressure (mmHg)

52 (40.0 - 71.5)

78 (72.8 - 87.3)

45 (39.0 - 62.0)

0.0001

Heart rate (/min)

109.0 (88.0 - 129.0)

94.5 (85.75 - 106.0)

112.0 (89.0 - 132.0)

0.009

Table 6. Independent risk factors affecting ICU mortality Parameter

p-value

OR (95% CI)

Discharge/day-of-death SOFA score

0.001

2.02 (1.33 - 3.06)

Vasopressor support

0.022

9.01 (1.37 - 59.46)

of MDR-AB infections, and indicated that use of monotherapy would be reasonable for at least treating less severe infections. [3] The same study suggested that drug choice can be customised according to drug access, drug interactions, side-effects and costs, since there is no difference between the drugs in resistant A. baumannii pneumonia reported to be hospital acquired or ventilator associated.[4] For A. baumannii, the colistin-carbapenem combination is a frequently used combination owing to its high synergistic effect, low antagonism and low resistance. [23] This was also the most commonly used combination in the present study (80 patients). However, carbapenem-resistant A. baumannii strains and other Gram-negativeresistant infections increase with the use of carbapenem.[5,14] Previous studies from Turkey have reported 55 - 70% carbapenem resistance for A. baumannii.[11,24,25] In the present study, carbapenem resistance was 95.5%. In A. baumannii infections resistant to carbapenem, mortality is higher than those with susceptibility to carbapenem. [12] The present study also found higher mortality in patients who received colistin-carbapenem combination. This increased mortality may have resulted either from the failure to achieve the desired synergistic effect after administration to severely infected patients or from the addition of infections by other resistant microorganisms developing under carbapenem.

Sulbactam is an antibiotic with bactericidal or bacteriostatic effects that is used in MDR-AB infections. It is not recommended to use alone for the treatment of severe infections. In vitro studies with imipenem-resistant A. baumannii strains report a high synergistic effect of sulbactam with colistin. [1,5,17] Since sulbactam alone has not been available in Turkey, combined preparations of sulbactam/ ampicillin or cefoperazone/sulbactam in high doses have been used.

In our hospital, a formula with sulbactam alone has been used for the previous year. The present study also found higher microbiological eradication with a colistin-sulbactam-ampicillin (SAM) combination.

However, it did not have an effect on clinical response. A previous study compared colistin and colistin-sulbactam in VAP infections caused by MDR-AB in ICU, and found both similar clinical response and microbiological response on day 5 of treatment and at the end of the treatment. In conclusion, no difference was reported between combination therapy and monotherapy.[26] The known risk factors for A. baumannii infections include advanced age, severe underlying diseases, immunosuppression, major traumas, burns, invasive procedures, catheters, MV support, surgical treatments, prolonged hospital stay, and inadequate

and improper antibiotic therapies.[1,4,27] Additionally, factors such as underlying diseases, varied infection severity depending on the patient, diagnostic delays and initiation of treatment may also cause difficulties in evaluating the efficacy of combined therapies in resistant A. baumannii infections.[5] The present study evaluated APACHE II, CCI, SOFA and RIFLE scores, and found a higher discharged/day-of-death SOFA score associated with increased mortality. Furthermore, a higher CCI score and the presence of underlying haematological malignancy were established as important factors affecting mortality. One study, which employed colistin combinations in Gram-negative infections, found no difference between treatment regimens, but found mortality was significantly associated with age, CCI score and severity of acute disease.[28] Similar to the present study, other researchers have found higher mortality in patients with malignancy and chronic renal damage.[4] Colistin-related nephrotoxicity rates varying between 0 and 33% have been reported in previous studies. Colistin nephrotoxicity usually occurs within the first week of use. Age, severity of the underlying disease, co-existence of septic shock, use of other nephrotoxic agents, time of exposure to colistin and cumulative dose are reported to be associated with colistin nephrotoxicity. [6,20,29] In the present study, colistin-related nephrotoxicity occurred at a rate of 19.2% and on median day 6. No correlation was found between exposure time or cumulative time and nephrotoxicity. A direct correlation was not demonstrated between nephrotoxicity and mortality; however, mortality was lower in the ‘no risk’ group and higher in the ‘failure’ group based on RIFLE scoring at the end of colistin therapy.

Study limitations The present study had some limitations. It was a small-sample, single-centre, retrospective study, which limits generalisability.

We did not include trauma patients and surgical patients, since the study was conducted in a medical ICU. The co-existence of other infections with A. baumannii infection in the patients posed a challenge, particularly for interpreting the clinical outcome and mortality. It became difficult to interpret the results because of intergroup transitions between colistin monotherapy and combination therapies, or other treatment changes.

Conclusion In conclusion, the present study has shown that colistin mono­ therapy and combination therapies have no superiority over each other for clinical response in the treatment of nosocomial pneumonia/VAP caused by MDR-AB; colistin-SAM was associated with improved microbiological eradication and colistin-carba­ penem combination was associated with increased mortality.

References 1. Karageorgopoulos DE, Falagas ME. Current control and treatment of multidrug-resistant Acinetobacter baumannii infections. Lancet Infect Dis 2008;8(12):751-762. [http://dx.doi. org/10.1016/S1473-3099(08)70279-2]

SAJCC November 2015, Vol. 31, No. 2

57

2. Vila J, Pachón J. Therapeutic options for Acinetobacter baumannii infections. Expert Opin Pharmacother 2008;9(4):587-599. [http://dx.doi.org/10.1517/14656566.9.4.587] 3. López-Cortés LE, Cisneros JM, Fernández-Cuenca F, et al. Monotherapy versus combination therapy for sepsis due to multidrug-resistant Acinetobacter baumannii: Analysis of a multicentre prospective cohort. J Antimicrob Chemother 2014;69(11):3119-3126. [http://dx.doi.org/10.1093/jac/dku233] 4. Khawcharoenporn T, Pruetpongpun N, Tiamsak P, Rutchanawech S, Mundy LM, Apisarnthanarak A. Colistin-based treatment for extensively drug-resistant Acinetobacter baumannii pneumonia.

Int J Antimicrob Agents 2014;43(4):378-382. [http://dx.doi.org/10.1016/j.ijantimicag.2014.01.016] 5. Batirel A, Balkan II, Karabay O, et al. Comparison of colistin-carbapenem, colistin-sulbactam, and colistin plus other antibacterial agents for the treatment of extremely drug-resistant Acinetobacter baumannii bloodstream infections. Eur J Clin Microbiol Infect Dis 2014;33(8):13111322. [http://dx.doi.org/10.1007/s10096-014-2070-6] 6. Yılmaz GR, Baştuğ AT, But A, et al. Clinical and microbiological efficacy and toxicity of colistin in patients infected with multidrug-resistant gram-negative pathogens. J Infect Chemother 2013;19:57-62. [http://dx.doi.org/10.1007/s10156-012-0451-2] 7. Kallel H, Bahloul M, Hergafi L, et al. Colistin as a salvage therapy for nosocomial infections caused by multidrug-resistant bacteria in the ICU. Int J Antimicrob Agents 2006;28(4):366-369. [http://dx.doi.org/10.1016/j.ijantimicag.2006.07.008] 8. Centers for Disease Control and Prevention. Ventilator-associated Pneumonia (VAP) Event 2014. http://www.cdc.gov/nhsn/PDFs/pscManual/10-VAE_FINAL.pdf (accessed 4 January 2015).

9. Kuo SC, Lee YT, Yang SP, et al. Eradication of multidrug-resistant Acinetobacter baumannii from the respiratory tract with inhaled colistin methanesulfonate: A matched case-control study.

Clin Microbiol Infect 2012;18(9):870-876. [http://dx.doi.org/10.1111/j.1469-0691.2011.03682.x] 10. Bellomo R, Ronco C, Kellum JA, Mehta RL, Palevsky P, Acute Dialysis Quality Initiative workgroup.

Acute renal failure-definition, outcome measures, animal models, fluid therapy and information technology needs: The Second International Consensus Conference of the Acute Dialysis Quality Initiative (ADQI) Group. Crit Care 2004;8(4):204-212. [http://dx.doi.org/10.1186/cc2872] 11. Dizbay M, Altuncekic A, Sezer BE, Ozdemir K, Arman D. Colistin and tigecycline susceptibility among multidrug-resistant Acinetobacter baumannii isolated from ventilator-associated pneumonia. Int J Antimicrob Agents 2008;32(1):29-32. [http://dx.doi.org/10.1016/j.ijantimicag.2008.02.016] 12. Lemos EV, de la Hoz FP, Einarson TR, et al. Carbapenem resistance and mortality in patients with Acinetobacter baumannii infection: Systematic review and meta-analysis. Clin Microbiol Infect 2014;20(5):416-423. [http://dx.doi.org/10.1111/1469-0691.12363] 13. European Centre for Disease Prevention and Control (ECDC). Annual Epidemiological Report.

Reporting on 2009 Surveillance Data and 2010 Epidemic Intelligence Data. Stockholm, Sweden: ECDC, 2011.

14. Falagas ME, Kopterides P. Review risk factors for the isolation of multi-drug-resistant Acinetobacter baumannii and Pseudomonas aeruginosa: A systematic review of the literature. J Hosp Infect 2006;64(1):7-15. [http://dx.doi.org/10.1016/j.jhin.2006.04.015] 15. Gaynes R, Edwards JR, National Nosocomial Infections Surveillance System. Overview of nosocomial infections caused by Gram-negative bacilli. Clin Infect Dis 2005;41(6):848-854. [http://dx.doi.org/10.1086/432803]

16. Falagas ME, Rafailidis PI, Ioannidou E, et al. Colistin therapy for microbiologically documented multidrug-resistant Gram-negative bacterial infections: A retrospective cohort study of 258 patients.

Int J Microbial Agents 2010;35(2):194-199. [http://dx.doi.org/10.1016/j.ijantimicag.2009.10.005] 17. Michalopoulos A, Falagas ME. Review treatment of Acinetobacter infections. Expert Opin Pharmacother 2010;11(5):779-788. [http://dx.doi.org/10.1517/14656561003596350] 18. Koomanachai P, Tiengrim S, Kiratisin P, Thamlikitkul V. Efficacy and safety of colistin (colistimethate sodium) for therapy of infections caused by multidrug-resistant Pseudomonas aeruginosa and Acinetobacter baumannii in Siriraj Hospital, Bangkok, Thailand. Int J Antimicrob Agents 2007;11(5):402-406. [http://dx.doi.org/10.1016/j.ijid.2006.09.011] 19. Markou N, Apostolakos H, Koumoudiou C, et al. Intravenous colistin in the treatment of sepsis from multiresistant Gram-negative bacilli in critically ill patients. Critical Care 2003;7(5):78-83. [http://dx.doi.org/10.1186/cc2358] 20. Cheng CY, Sheng WH, Wang JY, Chen YC, Chang SC. Safety and efficacy of intravenous colistin (colistin methanesulphonate) for severe multidrug-resistant Gram-negative bacterial infections.

Int J Antimicrob Agents 2010;35(3):297-300. [http://dx.doi.org/10.1016/j.ijantimicag.2009.11.016] 21. Aydemir H, Akduman D, Piskin N, et al. Colistin vs. the combination of colistin and rifampicin for the treatment of carbapenem-resistant Acinetobacter baumannii ventilator-associated pneumonia.

Epidemiol Infect 2013;141(6):1214-1222. [http://dx.doi.org/10.1017/S095026881200194X] 22. Durante-Mangoni E, Signoriello G, Andini R, et al. Colistin and rifampicin compared with colistin alone for the treatment of serious infections due to extensively drug-resistant Acinetobacter baumannii: A multicenter, randomized clinical trial. Clin Infect Dis 2013;57(3):349-358. [http:// dx.doi.org/10.1093/cid/cit253] 23. Zusman O, Avni T, Leibovici L, et al. Systematic review and meta-analysis of in vitro synergy of polymyxins and carbapenems. Antimicrob Agents Chemother 2013;57(10):5104-5111. [http:// dx.doi.org/10.1128/AAC.01230-13] 24. Gür D, Hascelik G, Aydın N, et al. Antimicrobial resistance in Gram-negative hospital isolates: Results of the Turkish HITIT-2 surveillance study of 2007. J Chemother 2009;21(4):383-389. [http://dx.doi.org/10.1179/joc.2009.21.4.383] 25. Kurtoğlu MG, Opuş A, Kaya M, Keşli R, Güzelant A, Yüksekkaya Ş. Bir eğitim ve araştırma hastanesinde klinik örneklerden izole edilen Acinetobacter baumannii suşlarında antibakteriyel direnç. Ankem Derg 2011;25:35-41.

26. Kalin G, Alp E, Akin A, Coskun R, Doganay M. Comparison of colistin and colistin/sulbactam for the treatment of multidrug resistant Acinetobacter baumannii ventilator-associated pneumonia. Infection 2014;42(1):37-42. [http://dx.doi.org/10.1007/s15010-013-0495-y] 27. Maragakis LL, Perl TM. Acinetobacter baumannii: Epidemiology, antimicrobial resistance, and treatment options. Clin Infect Dis 2008;46(8):1254-1263. [http://dx.doi.org/10.1086/529198] 28. Crusio R, Rao S, Changawala N, et al. Epidemiology and outcome of infections with carbapenemresistant Gram-negative bacteria treated with polymyxin B-based combination therapy. Scand J Infect Dis 2014;46(1):1-8. [http://dx.doi.org/10.3109/00365548.2013.844350] 29. DeRyke CA, Crawford AJ, Uddin N, Wallace MR. Colistin dosing and nephrotoxicity in a large community teaching hospital. Antimicrob Agents Chemother 2010;54(10):4503-4505. [http:// dx.doi.org/10.1128/AAC.01707-09]

The accuracy of Johannesburg-based ambulance personnel in identifying stroke D Nel,1 BTEMC; W Stassen,1,2 MPhil EM 1 2

Department of Emergency Medical Care, Faculty of Health Sciences, University of Johannesburg, South Africa Aeromedical Division, ER24 Emergency Medical Services, Johannesburg, South Africa

Corresponding author: W Stassen (stassen88@gmail.com)

Background. Stroke is a potentially life-threatening, time-dependent event that requires urgent management to reduce morbidity and mortality. It has been suggested that earlier recognition by ambulance personnel and transport to stroke centres may significantly reduce treatment delays. For this reason it is vitally important that ambulance personnel are able to accurately diagnose stroke.

Methods. A series of vignettes were created that included images, video and audio displaying either signs or symptoms of stroke or those of another condition. Ambulance personnel were asked to review each vignette and state whether the patient described was suffering from a stroke or not. Further investigation was sought by requesting each individual to motivate their answer, mentioning upon what their diagnosis was based.

Results. A total of 40 basic life support (BLS) and intermediate life support (ILS) personnel from different sites diagnosed 280 vignettes.

BLS personnel were able to diagnose stroke with a sensitivity of 85.3% and a specificity of 89.9% (positive predictive value (PPV) 86.7%, negative predictive value (NPV) 88.8%), while ILS achieved a sensitivity of 98.2% and specificity of 94.0% (PPV 91.7%, NPV 98.8%). The combined sensitivity and specificity were 91.5% and 92.0%, respectively (PPV 89.2%, NPV 93.8%). In order to aid their diagnosis, only 5% of BLS and 18.34% of ILS utilised validated stroke screening tools.

Conclusion. Despite not using validated screening tools, the ambulance personnel sampled in this study were able to identify stroke with high accuracy. Further studies should be considered to identify how these diagnoses were reached in order to identify training needs.

S Afr J Crit Care 2015;31(2):58-61. DOI:10.7196/SAJCC.2015.v31i2.247

58

SAJCC November 2015, Vol. 31, No. 2

Sub-Saharan Africa is currently experiencing a triple burden of disease, and an increase in lifestyle diseases such as stroke is becoming apparent. [1,2] Within Africa, the prevalence of stroke has been esti­m ated to be 81.2 (range 13.2 - 94.9)/100 000 person years. [3] Morbidity and mortality following stroke may be devastating, placing severe strain on healthcare infrastructure.

Studies have shown that reduced time from insult to definitive treatment can greatly reduce the chances of dependency and death.[4] It is understood that stroke must be identified early and the patient must be taken to the closest appropriate stroke facility for definitive treatment options.[4] Often, prehospital personnel are the first healthcare professionals to come into contact with patients presenting with signs of stroke.[4,5] From this, their role in the initial identification and diagnosis of stroke and appropriate referral to a stroke centre becomes clear.[4,6] Early identification may significantly reduce the time to treatment, and finally translate to improved outcome.[7] By allowing for preferential transport to stroke centres, patients may receive timely fibrinolysis or endovascular therapy to reduce cerebral tissue necrosis. [4-6] Locally, ambulance personnel are trained by means of a series of short courses. Formal stroke assessment is only taught within the intermediate life support (ILS) curriculum, unlike the international standard where emergency call-takers are expected to identify stroke immediately on first call to an emergency call centre.[4,6] Prehospital personnel have been reported to identify stroke with an accuracy of between 61 and 83%; however, in the presence of signs and symptoms that mimic stroke, a tendency towards overdiagnosis is seen.[7] In order to improve accuracy, two main prehospital stroke screening and diagnostic tools have been developed. By applying the Los Angeles Prehospital Stroke Scale (LAPSS) and the Cincinnati Prehospital Stroke Scale (CPSS), the accuracy may improve to sensitivities and specificities

of up to 74 - 97% and 72 - 94%, respectively.[7-9] Even though utilising these diagnostic tools is recommended by international guidelines,[4,6] local data for the use and accuracy of prehospital stroke identification are not available.

In order to address this paucity in evidence available, the primary objective of this preliminary study was to determine the accuracy of basic life support (BLS) and ILS identification of stroke, and the rationale behind making these diagnoses. A secondary objective was to understand upon what these diagnoses were based.

Methods A prospective, cross-sectional approach was employed by creating a series of seven non-validated vignettes that included images, video and audio displaying either signs or symptoms of stroke or those of other conditions. Each of the vignettes was created by including all elements of the LAPSS and CPSS into each case. This was confirmed independently during review by three emergency care practitioners.

Table 1 shows the details of each vignette. Due to resource constraints, a non-probability convenience sample of 40 ambulance personnel was drawn from four different private emergency medical service stations within the Johannesburg region of Gauteng Province, South Africa (SA). The stations are located in the southern, eastern, northern and western parts of the greater Johannesburg area. Further investigation was done by requesting each individual to motivate their answer in an open-ended response, describing upon what their diagnosis was based.

Ambulance personnel were required to view a PowerPoint (Microsoft, USA) presentation of the vignettes (Table 1) and then mention whether stroke was present or absent and motivate how they reached the diagnosis. Three vignettes were positive for stroke diagnosis according to the LAPSS and CPSS, while four vignettes described patients with hypoglycaemia and dehydration, ptosis,

Table 1. Vignettes Stroke (based on LAPSS/CPSS)

Case description

Media used

Case 1

A 70-year-old male patient with a history of hypertension is presented. The patient has acute onset of depressed level of consciousness with anisocoria, aphasia and unilateral facial droop. The patient also has hemiplegia.

Anisocoria pictured.

Facial droop pictured.

Yes

Case 2

A 40-year-old female patient with dyspnoea is presented. She presents with a productive cough and a physical examination in keeping with a lower respiratory tract infection. Potentially mimicking facial droop, ptosis is seen.

Ptosis is pictured.

No

Case 3

A 52-year-old female patient with a history of intense headache suddenly presents with slurred speech and depressed level of consciousness.

Hemiparesis and facial droop is appreciable. Positive pronator drift is found.

Facial droop is pictured.

Slurred speech is played.

Yes

Case 4

A 60-year-old male patient with depressed level of consciousness and general weakness is presented. The patient has signs of acute gastroenteritis (with nausea, vomiting and diarrhoea) and is a known diabetic. The patient is hypoglycaemic.

He has normal facial symmetry and appears acutely malnourished.

The patient is pictured.

No

Case 5

A 28-year-old female patient with focal signs is presented. The patient is known to suffer from depression and take selective serotonin-reuptake inhibitors. She presents with tardive dyskinesia. All other findings are normal.

Normal facies pictured.

Embedded video of tardive dyskinesia shown.

No

Case 6

An 89-year-old frail patient with a history of atrial fibrillation is presented. The patient has an acute history of facial asymmetry and aphasia. Hemiparesis is also appreciable.

Assymetrical facies is shown.

Yes

Case 7

A 32-year-old patient involved in a motor vehicle accident is presented.

The patient presents with ataxia, slurred speech and bilateral mydriasis. The patient admits to alcohol use.

Pupils are pictured.

An embedded video of ataxia is shown.

No

SAJCC November 2015, Vol. 31, No. 2

59

tardive dyskinesia and ataxia secondary to alcohol intoxication.

The data collection tool also collected demographic data on each participant.

Demographic data were analysed descriptively and reported in this manner (Table 2). The answers provided for each of the vignettes were marked according to the validated stroke scales (LAPSS and CPSS) by the principal investigators, and accuracy was determined by using standard formulae for sensitivity, specificity and negative and positive predictive values (NPVs and PPVs). Both the LAPSS and the CPSS were used as gold standard, i.e. a positive diagnosis for stroke in both scales was considered positive for stroke. The free text data were analysed to identify the diagnostic factors used to reach each diagnosis.

Each volunteer participant was required to provide written informed consent. Ethical approval was obtained from the University of Johannesburg’s Human Research Ethics Committee (REC-01-1192014). All data collection forms were kept anonymous.

Table 2. Demographic information Demographics

BLS

ILS

Total

Male, n

18

8

26

Female, n

2

12

14

Average age (years)

28

29

28.5

Average work experience (years)

3

9

6

Table 3. Stroke identification of BLS personnel LAPSS/CPSS: Stroke, n

LAPSS/CPSS: No stroke, n

Total, n

BLS: Stroke

52

8

60

BLS: No stroke

9

71

80

Total

61

79

140

Results A total of 43 questionnaires were distributed; three questionnaires were not completed as the participants were on duty and needed to service an emergency call. Only 40 questionnaires were therefore eligible for data analysis. A total sample of 280 vignettes was therefore analysed.

Tables 3 and 4 show the accuracies of stroke identification for BLS and ILS personnel, respectively. BLS were able to accurately diagnose stroke with a sensitivity of 85.3% and a specificity of 89.9% (PPV 86.7%, NPV 88.8%), while ILS achieved a sensitivity of 98.2% and a specificity of 94.0% (PPV 91.7%, NPV 98.8%). The combined sensitivity and specificity were 91.5% and 92.0%, respectively (PPV 89.2%, NPV 93.8%).

Table 5 presents the factors on which stroke diagnoses were based. In general, there was a tendency by the sample at both levels of care (BLS and ILS) to rely on specific signs and symptoms, rather than using validated stroke screening tools such as CPSS and LAPSS. Only 5% of BLS and 18.3% of ILS utilised these tools to aid their diagnosis. In one instance, the application of the stroke scale was done incorrectly.

Participants incorrectly attributed presentations of patients with stroke to other causes such as patients being ‘intoxicated’ or ‘drunk’ on ethanol or patients having suffered a ‘head injury.’ Slurred speech was attributed to intoxication instead of stroke in certain instances; as was change in level of consciousness. Interestingly, a deranged level of consciousness was described as a sign that excludes stroke as a diagnosis. Left-sided hemiparesis was attributed to ‘myocardial infarction’ rather than stroke. This was more likely in the presence of abnormal heart rates and rhythms within the vignette.

Discussion The study presented certain scenarios of various patients either having suffered from stroke (as identified by LAPSS and CPSS) or suffering from a condition that may mimic a stroke in presentation.

BLS participants performed worse than the ILS participants. This was expected due to the fact that the ILS generally had more experience (mean 9 v. 3 years) and had more extensive medical training. In SA, both of these levels of care are educated by means of short-course certification. BLS personnel undergo a 6-week course that covers elementary aspects of prehospital care – this is the entry level for ambulance personnel. After 1 000 practical patient hours as BLS, personnel are eligible for entry into the ILS course. The ILS course is

60

SAJCC November 2015, Vol. 31, No. 2

Table 4. Stroke identification of ILS personnel LAPSS/CPSS: Stroke, n

LAPSS/CPSS: No stroke, n

Total, n

ILS: Stroke

55

5

60

ILS: No stroke

1

79

80

Total

56

84

140

Table 5. Frequency of reasons cited for stroke diagnosis Diagnostic factor

BLS (%)

ILS (%)

Hemiparesis

61.7

80.0

Abnormal pupils

33.3

45.0

Chronic hypertension

60.0

45.0

Confusion

10.0

15.0

Current hypertension

28.3

25.0

Decreased level of consciousness

36.7

28.3

Facial asymmetry

41.7

68.3

Hyperglycaemia

12.5

12.5

Patient age

3.3

10.0

Positive prediction by stroke scales

5.0

18.3

Slurred speech

55.0

63.3

Stroke scales

5.0

18.3

over 18 weeks and teaches further skills such as intravenous therapy, basic electrocardiography and manual defibrillation.

Similar studies on stroke accuracy in SA could not be found.

However, data from international studies may be comparable.

Studies have shown that EMS personnel can identify stroke with an accuracy of between 61 and 83%.[5,10,11] A similar study conducted in Australia reported sensitivities of accurate paramedic stroke diagnosis of 93% and specificities of 87%, which is similar to the accuracies identified by the current study.[12] When using a validated tool (such as the LAPSS), the sensitivity and specificity may increase to 91% (95% CI 76 - 98%) and 97% (95% CI 93 - 99%), respectively.[13] It is important to remember that many areas within SA remain underserviced by advanced life support, with 80% of all emergency care personnel being BLS.[14] As first medical contact, these personnel

need to be proficient at identifying stroke and expediting transport to stroke-ready hospitals.[4] BLS may identify stroke accurately with a sensitivity of 85.3% and a specificity of 89.9% (PPV 86.7%, NPV 88.8%), but only use a validated stroke assessment tool 5% of the time. The literature suggests that diagnostic accuracy may increase should validated screening tools be used to aid diagnosis. Using these tools has been shown to consistently improve accuracy of stroke diagnosis. [7-9] One of the manners in which the utilisation of a stroke diagnostic tool may be improved is by educational interventions.[5,15-17] Yet, educating those in rural areas may be a costly exercise as they lack access to continuous professional development opportunities.[18] An online training programme has been suggested to counteract this, and has been shown to be of benefit.[19] By allowing for earlier screening, transport may be expedited (including aeromedical transport),[4] which might lead to shorter treatment times and a reduction in the burden of morbidity and mortality.[4,17] Interestingly, the respondents (despite having high accuracy rates) incorrectly attributed clinical presentations to diag­ noses. Positive clinical signs for stroke were often misattributed to non-stroke conditions or simply to age-related pathologies and found to be normal. This might show a critical need for development of the clinical reasoning abilities of local prehospital providers. The majority of the participants did not use any form of validated stroke screening tool to identify or rule out stroke. Shortcomings in the clinical decisionmaking of ambulance personnel have also been identified as one of the biggest threats to patient safety in the prehospital setting.[20] Clinical reasoning and decision-making may be improved by way of simulation-based training.[21-23]

Study limitations This preliminary study was largely observational and provides foundational data for future study. The role of sampling bias cannot be excluded. The sample was drawn from a single private service, and this detracts from the external validity. Internal validity is also compromised as the vignettes, even though they were developed and reviewed by senior emergency care practitioners, remained unvalidated.

Conclusion It would be interesting to repeat this study using validated vignettes on a national, more heterogeneous sample. A project to determine the current knowledge of ambulance personnel on diagnostic stroke tools should be undertaken. A high accuracy of stroke identification was appreciable in this sample. This accuracy was higher than that of other studies cited, despite local personnel not utilising validated

screening tools. It is concerning that clinical reasoning was inaccurate at times, and presentations were attributed to unrelated pathologies or comorbid conditions. More training should be considered to promote the use of these validated tools to aid in the diagnosis and referral of stroke patients.

References 1. Reed DM. The paradox of high risk of stroke in populations with low risk of coronary heart disease. Am J Epidemiol 1990;131(4):579-588.

2. Mbewu A. The burden of cardiovascular disease in sub-Saharan Africa. S Afr Heart J 2009;6(1):4-10.

3. Adeloye D. An estimate of the incidence and prevalence of stroke in Africa: A systematic review and meta-analysis. PLoS One 2014;9(6):e100724. [http://dx.doi.org/10.1371/journal. pone.0100724] 4. Jauch EC, Saver JL, Adams HP, et al. Guidelines for the early management of patients with acute ischemic stroke. Stroke 2013;44:870-947. [http://dx.doi.org/10.1161/STR.0b013e318284056a] 5. Watkins CL, Leathley MJ, Jones SP, Ford GA, Quinn T, Sutton CJ. Training emergency services’ dispatchers to recognise stroke: An interrupted time-series analysis. BMC Health Serv Res 2013;13:318. [http://dx.doi.org/10.1186/1472-6963-13-318] 6. Powers WJ, Derdeyn CP, Biller J, et al. 2015 AHA/ASA Focused Update of the 2013 Guidelines for the Early Management of Patients With Acute Ischemic Stroke Regarding Endovascular Treatment. Stroke 2015. [http://dx.doi.org/10.1161/STR.0000000000000074] 7. Bray JE, Martin J, Cooper G, Barger B, Bernard S, Bladin C. Paramedic identification of stroke: Community validation of the Melbourne ambulance stroke screen. Cerebrovasc Dis 2005;20(1):28-33. [http://dx.doi.org/10.1159/000086201] 8. Hurwitz AS, Brice JH, Overby BA, Evenson KR. Directed use of the Cincinnati Prehospital Stroke Scale by laypersons. Prehosp Emerg Care 2005;9(3):292-296. [http://dx.doi. org/10.1080/10903120590962283] 9. Chen S, Sun H, Lei Y, et al. Validation of the Los Angeles Pre-hospital Stroke Screen (LAPSS) in a Chinese urban emergency medical service population. PLoS One 2013;8(8):e70742. [http:// dx.doi.org/10.1371/journal.pone.0070742] 10. Smith WS, Isaacs M, Corry MD. Accuracy of paramedic identification of stroke and transient ischemic attack in the field. Prehosp Emerg Care 1998;2(3):170-175. [http://dx.doi. org/10.1080/10903129808958866] 11. Smith WS, Corry MD, Fazackerley J, Isaacs SM. Improved paramedic sensitivity in identifying stroke victims in the prehospital setting. Prehosp Emerg Care 1999;3(3):207-210. [http://dx.doi. org/10.1080/10903129908958938] 12. Bray JE, Coughlan K, Barger B, Bladin C. Paramedic diagnosis of stroke: Examining long-term use of the Melbourne Ambulance Stroke Screen (MASS) in the field. Stroke 2010;41(7):1363-1366. [http://dx.doi.org/10.1161/STROKEAHA.109.571836] 13. Kidwell CS, Starkman S, Eckstein M, Weems K, Saver JL. Identifying stroke in the field. Prospective validation of the Los Angeles Prehospital Stroke Screen (LAPSS). Stroke 2000;31(1):71-76. [http://dx.doi.org/10.1161/01.str.31.1.71] 14. Health Professions Council of South Africa. Professional Board for Emergency Care, iRegister. http://systems.hpcsa.co.za/iregister (accessed 7 Oct 2015).

15. Thomas C, Mackey E. Influence of a clinical simulation elective on Baccalaureate nursing student clinical confidence. J Nurs Educ 2012;51(4):236-239. [http://dx.doi.org/10.3928/01484834-20120224-03] 16. McNamara MJ, Oser C, Gohdes D, et al. Stroke knowledge among urban and frontier first responders and emergency medical technicians in Montana. J Rural Health 2008;24(2):189-193. [http://dx.doi.org/10.1111/j.1748-0361.2008.00157.x] 17. Nam HS, Park E, Heo JH. Facilitating stroke management using modern information technology.

J Stroke 2013;15(3):135-143.[http://dx.doi.org/10.5853/jos.2013.15.3.135] 18. Christopher L. An investigation into the non-compliance of Advanced Life Support practitioners with the guidelines and protocols of the professional board for emergency care practitioners. Research Report. Durban; Durban University of Technology, Department of Emergency Care and Rescue, 2007.

19. Stassen W, Wylie C, Holgate R. An online learning programme improves traumatic brain injury guideline adherence in a South African Helicopter Emergency Medical Service. Af J Emerg Med 2015;(in press).

20. Jensen J. Paramedic clincial decision-making: Results of two Canadian studies. Int Paramed Pract 2011;1(2):63-71. [http://dx.doi.org/10.12968/ippr.2011.1.2.63] 21. Aggarwal R, Mytton OT, Derbrew M, et al. Training and simulation for patient safety. Qual Saf Health Care 2010;19:i34-43. [http://dx.doi.org/10.1136/qshc.2009.038562] 22. Ten Eyck RP, Tews M, Ballester JM, Hamilton GC. Improved fourth-year medical student clinical decision-making performance as a resuscitation team leader after a simulation-based curriculum.

Simul Healthc 2010;5(3):139-145. [http://dx.doi.org/10.1097/SIH.0b013e3181cca544] 23. Hagiwara MA, Kängström A. Effect of simulation on the clinical competence of Swedish ambulance nurses. Australas J Paramed 2014;11(2).

SAJCC November 2015, Vol. 31, No. 2

61

CASE REPORT

Atrial myxoma-related embolism resulting in acute limb ischaemia in a critical care patient M Knight,1 MA, MB ChB (Hons), MRCEM, MRCP; R D Wise,1,2 MB ChB, FCA (SA), Cert Crit Care (SA), MMed (Anaes), Dip Obst (SA), Dip PEC (SA) 1 2

Department of Anaesthetics, Critical Care, and Pain Management, Edendale Hospital, Pietermaritzburg, South Africa Perioperative Research Group, Discipline of Anaesthesiology and Critical Care, Nelson R Mandela School of Medicine, University of KwaZuluNatal, Durban, South Africa

Corresponding author: R D Wise (robertwise@webafrica.org.za)

This report presents an unusual case of limb ischaemia in the critical care setting, the cause of which was elucidated on echocardio­ graphy. Evaluation of the case highlights the importance of appropriate and timely investigation, in particular the role of bedside echocardiography. Although atrial myxomas are uncommon, a thorough investigation of patients presenting with acute peripheral ischaemic events should be undertaken to facilitate the diagnosis of this treatable condition.

S Afr J Crit Care 2015;31(2):62-63. DOI:10.7196/SAJCC.2015.v31i2.233

We report an interesting case of a patient presenting with a constellation of symptoms and signs, all of which may be attributable to an underlying left atrial myxoma. The availability of bedside echo­cardio­ graphy in the intensive care unit (ICU) permitted the timely evaluation and diagnosis of this rare cardiac lesion.

Case report A 41-year-old male suffered a sudden loss of consciousness at home while eating dinner. He had no complaints of preceding symptoms and had been well earlier that day. He remained unconscious and was transferred to a regional hospital where he was found to have signs of left ventricular failure. His initial management, including tracheal intubation, occurred in the Emergency Department, after which he was transferred to the ICU for ventilatory support.

It was discovered in his history and in discussion with his family that he had recently been admitted to hospital following a non-ST elevation myo­cardial infarction (NSTEMI) and was awaiting outpatient cardiology follow-up. A provisional diagnosis of dilated cardiomyopathy secondary to excess alcohol intake had been made based on clinical and radiographic assessment; however, echocardiographic evaluation had not yet been performed owing to resource limitations. Over the preceding month he had also been complaining of progressive dyspnoea.

Clinical examination on this admission revealed signs of chronic alcohol consumption including parotid enlargement and hepatomegaly. He was noted to have a laterally displaced, thrusting apex beat and a grade 3/6 early diastolic murmur at the lower left sternal edge, with no other added sounds. His electrocardiogram showed lateral T-wave inversion and cardiac enzymes were markedly raised, in keeping with a further NSTEMI. Throughout this, however, he remained haemodynamically stable. Despite a low Glasgow Coma Scale score, he did not have any localising neurological signs, or signs or history of recent toxin ingestion.

On investigation, he did not display any metabolic or electrolyte derangements at the time of his admission; however, thrombocytopenia and a prolonged PT were noted in keeping with chronic alcohol use.

He was treated for an NSTEMI but could not be transferred to the local tertiary hospital’s coronary care unit owing to a bed shortage. Two

62

SAJCC November 2015, Vol. 31, No. 2

days after ICU admission acute ischaemia of his right leg developed; an emergency above-knee amputation took place without complication.

A bedside transthoracic echocardiogram (TTE) performed in ICU following the operation identified a 4 cm × 3 cm highly mobile polypoid mass in the left atrium that could be seen prolapsing into the mitral valve orifice (Fig. 1), in keeping with an atrial myxoma. Two smaller, immobile masses attached to the atrial septum were also seen, which most likely represented atrial thrombi. These were absent on subsequent studies. Further echocardiographic evaluation revealed dilated and poorly contractile ventricles and severe aortic regurgitation.

The patient’s condition continued to deteriorate. He developed an acute kidney injury and his level of consciousness failed to improve des­ pite the discontinuation of sedative agents. A computed tomography scan of his head identified several infarcts in the territories of the left middle cerebral artery, the left and right posterior cerebral artery, and lacunar infarcts in the left basal ganglia and right thalamus. He was considered unsuitable for surgical resection of the tumour and unfortunately he died with multiorgan failure 6 days after his ICU admission.

Fig. 1. Apical four-chamber view showing a well-circumscribed left atrial mass.

Discussion Although rare, with an annual incidence of 0.5 per million, [1] atrial myxomas are the most common of the benign cardiac tumours, accounting for approximately half of identified cases. Seventy-five percent of myxomas occur in women and the majority of patients are aged between 30 and 60 years at diagnosis.[2] Most cases are sporadic; however, familial cardiac myxomas may occur, in association with conditions such as the Carney Complex. This autosomal dominant condition results from a germline mutation in the PRKAR1A gene involved in regulating cell proliferation, and is characterised by skin pigmentation, endocrinopathies, and endocrine and non-endocrine tumour formation, including cardiac tumours.[3] Myxomas are thought to derive from multipotential mesenchymal cells, and have a variable macroscopic appearance ranging from soft and friable to smooth and bosselated.[4] Most tumours are found in the atria, with 75% originating from the fossa ovalis region of the left atrium. Fifteen to twenty percent of cases are detected in the right atrium and only 3 - 4% are found in either the left or right ventricle.

While between 2 and 12% of patients remain asymptomatic,[1,5] most present with one or more of the triad of constitutional symptoms, intracardiac obstruction and embolic phenomena.[6] Embolism occurs in ~30 - 40% of cases and may involve any internal organ; however, cerebral arteries are most commonly affected.[6] The diagnosis of myxoma is occasionally made after histological assessment of embolic material reveals myxomatous tissue, with subsequent investigation leading to identification of the primary cardiac tumour.[7] Constitutional symptoms are common, reported in approximately one-third of patients, and include weight loss, fever, fatigue and symptoms mimicking connective tissue disorders.[4] These have been attributed to the expression of interleukin-6 by the myxoma and typically resolve following tumour resection.[8] Syncope and sudden death caused by cardiac myxomas have been reported.[9] Large left atrial tumours impairing intracardiac blood flow can produce variable clinical features, from slowly progressive orthopnoea and paroxysmal nocturnal dyspnoea, to acute pulmonary oedema, syncope or sudden death.[10,11] Sudden death resulting from very small tumours is more likely to be related to embolism to the coronary or cerebral arteries.

In the case described here, obstruction of the mitral valve by the myxoma may have provoked syncope and features of left ventricular failure. The history of presumed dilated cardiomyopathy secondary to chronic alcohol use may have distracted from the possibility of other aetiologies, and led to a delay in diagnosis, although dual pathologies may well have existed in this case. Tumour embolism, or dislodgement of the associated left atrial thrombi, may also have contributed to the initial deterioration of the patient, and are most likely to be the cause of the acute lower limb ischaemia and multiterritorial cerebral infarcts.

Unfortunately, the embolic material was not removed during the operation, preventing histological confirmation. Several case reports have implicated atrial myxoma as the cause of acute myocardial infarction, and this may also have been a contributing factor in this case.[12,13] Aortic regurgitation has been found in association with atrial myxomas where the tumour pedicle is attached to the atrial septum close to the aortic root.[14] Tractional forces applied to the aortic root as the tumour moves towards the left ventricle during diastole have been implicated in the aortic incompetence.

TTE is a valuable investigation, allowing identification, localisation and measurement of cardiac tumours, and should be considered in all cases of suspected cardiogenic embolism. Where myxoma is identified without preceding embolism, the subsequent risk can be evaluated based on

echocardiographic morphology. Polypoid and polylobulated tumours have been associated with increased risk of embolism and therefore may merit more urgent surgery. [15,16] Transoesophageal echocardiography is more sensitive than TTE (95.2% and 100%, respectively);[17] however, the limited availability, particularly in the developing world, restricts its use as a diagnostic tool. Indeed, since most district-level hospitals in South Africa lack echocardiography services and the sensitivity of clinical examination and plain chest radiography for the diagnosis of cardiac tumours is low, many cases are likely to go undiagnosed.

Treatment of cardiac myxoma is by surgical resection via median sternotomy and is generally curative. Rates of tumour recurrence among sporadic cases are between 1 and 3%, but up to 20% in familial cases. Operative mortality is low in most studies. In a single centre French study, Pinede et al.[5] followed 112 patients postoperatively for a median of 3 years and reported only four deaths, which is consistent with mortality rates in other series.[18]

Conclusion Atrial myxoma is rare and can present with a wide range of symptoms.

A low threshold for TTE in suspected cases may facilitate diagnosis and expedite definitive surgical management that carries with it an excellent prognosis. In the case of cardiomyopathy with embolic phe­ nomena, urgent TTE should be sought to identify potential causes such as thrombi, endocarditis and, as in this case, atrial myxoma.

Learning points • Atrial myxomas have a variable clinical presentation and should be included in the differential diagnosis of confirmed or suspec­ ted cardiogenic embolism.

• Transthoracic echocardiography is a highly sensitive investigation in cases of suspected cardiac myxoma.

• Surgical resection of atrial myxomas is generally curative and carries with it an excellent prognosis, highlighting the importance of early diagnosis.

References 1. MacGowan S, Sidhu P, Aherne T, et al. Atrial myxoma: National incidence, diagnosis and surgical management. Ir J Med Sci 1993;162(6):223-226.

2. Sarjeant J, Butany J, Cusimano R. Cancer and the heart: Epidemiology and management of primary tumors and metastases. Am J Cardiovasc Drugs 2003;3(6):407-421.

3. Mabuchi T, Shimizu M, Ino H, et al. PRKAR1A mutation in patients with cardiac myxoma. Int J Cardiol 2005;102(2):273-277. [http://dx.doi.org/10.1016/j.ijcard.2004.05.053] 4. Butany J, Nair V, Ather N, et al. Cardiac tumours: Diagnosis and management. Lancet Oncol 2005;6(4):219-228. [http://dx.doi.org/10.1016/S1470-2045(05)70093-0] 5. Pinede L, Duhaut P, Loire R. Clinical presentation of left atrial myxoma: A series of 112 consecutive cases. Medicine (Baltimore) 2001;80(3):159-172. [http://dx.doi.org/10.1097/00005792200105000-00002] 6. Reynen K. Cardiac myxomas. N Engl J Med 1995;333(24):1610-1617.

7. Silverman J, Olwin J, Graettinger J. Cardiac myxomas with systemic embolization: Review of the literature and report of a case. Circulation 1962;26:99-103. [http://dx.doi.org/10.1161/01.

CIR.26.1.99] 8. Mendoza C, Rosado M, Bernal L. The role of interleukin-6 in cases of cardiac myxoma: Clinical features, immunologic abdnormalities, and a possible role in recurrence. Tex Heart Inst J 2001;28(1):3-7.

9. Modi K, Venkatesh P, Agnani S, et al. Sudden death in a patient with left atrial myxoma: Report of two cases and review of literature. BJMP 2010;3(2):318.

10. Fisicaro A, Slavich M, Agricola E, et al. Acute pulmonary oedema caused by a giant atrial myxoma. Case Rep Med 2013;2013:904952. [http://dx.doi.org/10.1155/2013/904952] 11. Nogueira D, Bontemp D, Menardi A, et al. Left atrial myxoma as the cause of syncope in an adolescent. Arq Bras Cardiol 2003;81(2):206-209.

12. Sankar N, Vaidyanathan R, Prasad G, et al. Left atrial myxoma presenting as acute inferior wall infarction: A case report. J Card Surg 2006;21(5):478-479. [http://dx.doi.org/10.1111/j.15408191.2006.00282.x] 13. Panos A, Kalangos A, Sztajzel J. Left atrial myxoma presenting with myocardial infarction: Case report and review of the literature. Int J Cardiol 1997;62(1):73-75.[http://dx.doi.org/10.1016/ s0167-5273(97)00178-2] 14. Bourdillon P, Monro J, Johnson A. Left atrial myxoma with aortic regurgitation. Br Heart J 1978;40(5):575-578. [http://dx.doi.org/10.1136/hrt.40.5.575] 15. Scott N, Veinot J, Kwan-Leung C. Symptoms in cardiac myxoma. Chest 2003;124(6):2408.

16. Acebo E, Val-Bernal F, Gómez-Román J, et al. Clinicopathologic study and DNA analysis of 37 cardiac myxomas: A 28-year experience. Chest 2003;123(5):1379-1385.

17. Engberding R, Daniel W, Erbel R, et al. Diagnosis of heart tumours by transoesophageal echocardiography: A multicentre study in 154 patients. Eur Heart J 1993;14(9):1223-1228.

18. Centofanti P, Di Rosa E, Deorsola L, et al. Primary cardiac tumours: Early and late results of surgical treatment in 91 patients. Ann Thorac Surg 1999;68(4):1236-1241.

SAJCC November 2015, Vol. 31, No. 2

63

BOOK REVIEW

Cardiopulmonary Physiotherapy in Trauma: An Evidence-based Approach Edited by Heleen van Aswegen and Brenda Morrow. London: Imperial College Press, 2015. ISBN 978-1-78326-651-7 Trauma is one of the leading causes of mortality and morbidity world­ wide, affecting both adult and paediatric populations. The objective of this book is to provide physiotherapists at all stages of their careers with current evidence-based information to guide their physiotherapeutic management of both adults and children with trauma-related injuries.

Information has been drawn from published research as well as the clinical expertise of the contributors to provide readers with a comprehensive book that details the physiotherapist’s role and management of traumarelated injuries.

Chapter 1 examines the physiological response to trauma, deepening understanding of the physiological basis of the problems trauma patients often face. The following chapter highlights the important anatomical differences between children and adults and the clinical implications thereof.

Of particular relevance to the South African setting is the chapter on immunosuppressive diseases and trauma. This comprehensive chapter covers physiotherapy modalities, exercise prescription and both subjective and objective outcome measures that may be used in the management of trauma patients. Commonly encountered traumatic injuries such as burn, head, spinal cord, multiple orthopaedic and blunt penetrating injuries are presented in the following chapters. The final chapter examines the quality of life of trauma (and critical illness) survivors and its management.

The clinical chapters provide a concise but thorough review of the mechanism of injury, and medical and surgical management. The contributors present the specific objectives of physiotherapy, general precautions and contraindications, recommended treatment and outcome measures and, importantly, specific paediatric considerations.

Each chapter also contains a clinical case study for each of adults and children, providing the reader with an example of the clinical application of the theory presented. The paediatric considerations detailed in each chapter are useful, especially to those working in this field. Key takehome points are highlighted throughout in ‘Key Message’ boxes and the use of summary tables for detailed information helps the reader with quick referencing. While the focus of this book is primarily on the acute management of trauma-related injuries, where information is beyond the scope of the book, the reader is directed to additional resources.

The information provided in this book is practical and presented in such a manner that physiotherapists, whether qualified for one or many years or undergraduate students, will come away with comprehensive, evidenced-based understanding of the physiotherapist’s role and management in adults and children with trauma. It can also be used as a reference source for physiotherapists who have to treat patients with injuries that they do not normally encounter, and who require a quick refresh on how to approach and implement treatment.

This well-written book provides with reader with comprehensive, practical information for a holistic, evidence-based approach to the management of both children and adults with traumatic injuries. It is a worthwhile investment for anyone working in the trauma and hospital setting and with an interest in cardiopulmonary physiotherapy.

Alison Lupton-Smith University of Cape Town; Wessel de Kock Physiotherapists aluptonsmith@gmail.com

64

SAJCC November 2015, Vol. 31, No. 2

IMPORTANT DATES: 15 November 2015 Abstract Submissions Opens

30 March 2016 Early Registration Opens

04 April 2016 Abstract Submission Closes

12 July 2016 Early Registration Closes

04 May 2016 Notice of Approved Abstracts

01 August 2016 Registration Closes (On-site registration to apply)

THE OFFICIAL JOURNAL OF THE CRITICAL CARE SOCIETY OF SOUTHERN AFRICA