Anaesthetics

Recent changes in perioperative patient management — including well integrated pre-admission clinics, the “bundle of care” initiative and clinical handover using electronic medical records — mean that surgical patients should have a smooth perioperative journey.

Bruce P Waxman FRACS, FRCS, FACS

An intensive care unit is only as good as the care and decision making provided at 2 am. What is needed is an environment that promotes optimal decision making 24 hours a day.

Sean L Kelly BMed, FACEM · Roger D Harris MB BS, FACEM, FCICM · Anthony R Burrell MB BS, FANZCA, FCICM

Unless future well designed studies show an actual benefit for patients, hospitals and health care, policymakers should resist any attempts to enforce this potentially expensive and ineffective practice.

Adam M Deane MB BS, PhD, FCICM · Mary G White MB BCh, BAO, FFARCSI

To the Editor: The donation after cardiac death (DCD) procedure used by Coulson and colleagues is ethically problematic and is not, as claimed, re-implementation of the practice followed before brain-death organ donation protocols were introduced. First reported in 1992, the more precise term is controlled donation after cardiac death or controlled non-heart-beating organ donation.2 What is controlled is the timing, mode and criteria of death, ...

Judith R Kennedy · Michael C Kennedy

To the Editor: We congratulate Coulson and colleagues for sharing their single-centre experience of donation after cardiac death (DCD).1 Our recently published Queensland DCD data2 reflect their experience and highlight the efficacy of DCD in augmenting the pool of donors with organs available for transplantation. The number of Australian DCD organ donors is increasing. There were 19 in 2007 and 86 in 2009. In 2011, of the 337 total ...

Raj Kumar · Kiran Shekar · John F Fraser

In reply: The Alfred Hospital’s donation after cardiac death (DCD) program was developed in full knowledge of ethical concerns such as those raised by Kennedy and Kennedy. All are covered under “DCD guideline development and implementation” in the methods section of our article.1 Specifically, no patient was denied a chance of survival, because all would have undergone withdrawal of cardiorespiratory support even if donation were ...

Tim G Coulson · David Pilcher

A tale of anxiety, innocence and trust in the anaesthetic roomAs I entered the busy paediatric ward, I scanned the room for my patient — they’re often not easy to find among the other children, the visiting parents and siblings, and the staff, and are rarely anywhere near, let alone in, their assigned beds. I was eventually directed to a small child sitting quietly on his ...

Geoffrey C Mullins MB BS, FANZCA, FRCPC

Objective: To determine the increase in intensive care unit (ICU) bed availability that would result from the use of the New South Wales and Ontario Health Plan for an Influenza Pandemic (OHPIP) triage protocols.Design, setting and patients: Prospective evaluation study conducted in eight Australian, adult, general ICUs, between September 2009 and May 2010. All patients who were admitted to the ICU, excluding ...

on behalf of the Influenza Pandemic ICU Triage (iPIT) Study Investigators

To the Editor: Kapur and colleagues correctly assert that the Declaration of Montréal states that “access to pain management is a fundamental human right”.1 The Declaration was proposed by International Association for the Study of Pain (IASP) delegates to the International Pain Summit and was approved by the IASP Council. However, Kapur et al incorrectly assert that the Declaration leaves the position of diagnosis uncertain, and that “the only specific treatment modality mentioned is opioid therapy”. Article 3 of the Declaration recognises: The right of all people with pain to have access to appropriate assessment and treatment of the pain by adequately trained health care professionals.2 Footnote 6 includes the need for “educational programs regarding pain assessment and treatment in all of the health care professions”. Thus, a label of “chronic pain” does not lead to a right to opioid treatment without proper assessment and consideration of treatment options. Footnote 6 refers to a range of treatment options, such as pain medications, including opioids and other essential medications for pain, and best-practice interdisciplinary and integrative nonpharmacological therapies, with access to professionals skilled in [their] safe and effective use ...2 It would be hard to describe the carefully worded articles, obligations and footnotes of the Declaration as “dogma, moral coercion or forays into jurisprudence”.1 At its General Assembly in 2011, the World Medical Association (WMA) supported measures to improve access to pain management. Dr Mukesh Haikerwal, Chair of the WMA, said: Physicians and other health care professionals have an ethical duty to offer proper clinical assessments to patients with pain and to offer appropriate treatment.3 Further, a WMA resolution asserted that people facing pain had a right to appropriate pain management, including effective medications such as morphine. Denial of pain treatment violated the right to health and might be medically unethical.3 It is puzzling that Awerbuch feels that describing chronic pain as a disease means that “the patient becomes the sole arbiter of whether he or she is ill. The prescribing doctor has no means by which to objectively determine treatment outcomes”.4 In the article to which he refers,5 and in other reports including Australia’s National Pain Strategy,6 the “disease” of chronic pain is described in terms of physical, psychological and environmental factors that require assessment by health professionals with adequate knowledge and training. This is not just a labelling exercise.

Michael J Cousins

Your guide to paediatric anaesthesia. Craig Sims, Chris Johnson. Sydney: McGraw-Hill 2011 (374 pp, $129.95). ISBN 9780071000222. THIS GUIDE aims to provide expert practical information across the breadth of paediatric anaesthesia without getting bogged down by details. The vast clinical experience of the editors (and their team of contributors) ensures that these aims are fulfilled in an informed, practical way. Chapters covering basic science and techniques relating to anaesthesia and resuscitation for children are complemented by others discussing anaesthesia for a comprehensive range of surgical subspecialities as well as for specific paediatric conditions. Care has been taken to include discussion of current issues that have generated controversy. The target audience includes trainees wanting a pithy summary of paediatric anaesthesia to inform their clinical experience and assist in preparation for exams. Both editors are examiners for the Australian and New Zealand College of Anaesthetists, which means they understand how high the bar is for those exams. The general anaesthetist with an interest in paediatrics wanting an easy-to-read update will also be well served. The book reads very comfortably, often with the tone of the consultant who has an interest in teaching, but the style does not allow for exhaustive explanation, detailed debate over alternative approaches or extensive referencing. However, it is sure to be stimulating for the career paediatric anaesthetist. The search for academic defensibility has led to some bloated, heavily referenced texts that often leave the reader wondering what they would actually do when confronted with a clinical scenario. The source of much clinical teaching is expert opinion, as a consultant works with a trainee. This volume gives expert opinion a voice. Neither online resources nor standard tomes on the topic address the needs of those this volume aims to assist. It is a great resource for trainees and anaesthetists with a paediatric interest and promises to be a springboard for discussion among paediatric anaesthetists.

Ian M McKenzie

Proactive recognition and response, pre-emptive management Despite improvements in medical therapy and improved outcomes, patients admitted to acute care hospitals may suffer clinical deterioration and potentially avoidable morbidity. Recent reports show that patients admitted to hospital suffer cardiac arrests at a rate between 0.66 and 11.2 per 1000 admissions,1,2 and the approximately 80% mortality rate due to in-hospital cardiac arrests has not changed since the description of cardiopulmonary resuscitation more than 50 years ago.3,4 In this article, we review the epidemiology of in-hospital cardiac arrests and the evolution of the medical emergency team (MET) model to prevent them. In addition, we outline lessons learned from the MET model and the development of a national approach for recognising and responding to deteriorating patients. Finally, we contend that another paradigm shift is required to further improve patient outcomes via a more proactive and pre-emptive approach that includes improved detection and recognition of deterioration as well as structured management of patients who are at high risk of deterioration. The old paradigm: cardiac arrestWith introduction of cardiac arrest teams and Code Blue calls around 1970 came the first ubiquitous hospital protocol for escalating the care for deteriorating patients.3 This system is activated when a patient is pulseless and apnoeic, has no recordable blood pressure and is unresponsive. However, several studies published between 1990 and 2002 reported that in-hospital cardiac arrests were not sudden or unexpected, but were preceded by instability that was sub-optimally responded to.5-7 Up to 84% of patients had abnormal vital signs or new problems before cardiac arrest,5-7 and most were in asystole or had pulseless electrical activity.8,9 The current paradigm: METsIn 1995, the first description of the MET as a model for responding to patient deterioration was published.10 METs are composed of an intensive care, general medicine or emergency medicine registrar, and a critical care nurse. The advent of the MET delivered a mechanism for escalation of care earlier in the course of clinical deterioration than cardiac arrest. Criteria for a MET call are based on deranged vital signs and altered conscious state.10,11 A “staff member worried” criterion is often included to empower ward staff to call for assistance and bypass the traditional hierarchical model of escalation. Such criteria provide semi-objective thresholds for when ward staff should escalate care for deteriorating patients, and provide an expected and structured institutional response. The MET call is usually activated using a public announcement and pagers, and the team is expected to arrive within about 5 minutes. This system is the first patient-centred, organisation-wide system to be implemented across hospitals throughout the world. Over the past 16 years,10 METs have been introduced in at least 100 Australian hospitals.12 Although evidence of the effectiveness of METs is controversial,13,14 the MET model is currently the prevailing paradigm for identification of deteriorating ward patients and rapid escalation of care for them. Research into the characteristics and outcomes of MET patients has provided epidemiological data on deteriorating ward patients. In a MET-naive hospital, it was reported that 4.5% of ward patients fulfilled MET criteria on two separate days.15 Studies of mature METs report that 2.6%–5.6% of patients hospitalised for more than 24 hours receive MET review.16 Thus, about one in 25 ward patients will experience clinically significant deterioration. Research into MET patients has also shown that vital sign measurement is inconsistent and often incomplete, deterioration is variably recognised, and staff do not always activate a MET call when patients fulfil the criteria.17 In addition, activation is often delayed, and this is associated with increased mortality.18,19 Two surveys of ward nurses suggest that there is underappreciation of the significance of physiological derangements associated with MET criteria.20,21 A number of studies have reported that patients subject to MET review have a mortality rate of about 25%22-25 — higher than the 11% in-hospital mortality rate of patients admitted to intensive care units (ICUs),26 and much higher than the 2.5% in-hospital mortality rate for overall hospital admissions.27 Finally, about one-third of MET calls involve end-of-life care issues.24,25,28 These often occur out of hours, when parent teams are not in the hospital. As the patients are acutely unwell, they are often unable to participate in discussions regarding their preferences for active treatment, or conservative or palliative care. Accordingly, these decisions are typically left up to covering doctors, the MET and family members, and often need to be made expeditiously. A new paradigm: proactive, pre-emptiveThe MET model has emphasised the tenet that early intervention improves outcome (Box 1). Thus, the mortality of patients subject to MET review (25%) is lower than that of cardiac arrest patients (80%) but higher than that for patients admitted to ICUs. However, the MET model is reactive, as it requires a patient to deteriorate before it can be activated. We contend that a new paradigm is needed to promote a more proactive approach to recognising and responding to deterioration, and to provide a pre-emptive, structured, standardised approach to managing patients who are known to be at risk of deterioration (Box 2). This paradigm shift is supported by the National consensus statement: essential elements for recognising and responding to clinical deterioration,29 which describes the clinical systems necessary to provide appropriate care to deteriorating patients and the organisational supports needed to ensure that these systems operate effectively (Box 3). Predicting risk and preventing deteriorationNot all patients admitted to hospital will suffer a complication, but some will suffer sudden and unexpected deterioration. Increasing evidence suggests that patients can be risk profiled at, or even before, hospital admission. A hospital outcome predictive equation (HOPE) was recently developed using six variables that are associated with increased risk of death: age, sex, diagnosis at admission, emergency admission, aged care resident and inter-hospital transfer.27 In surgical patients, most postoperative complications are medical.30 It remains to be determined whether hospital-wide strategies or models of medical comanagement can improve patient outcomes based on such risk profiling. Similar criticisms were levelled at the MET model of care almost 10 years ago.31 Advance care planningAdvance care planning is particularly important in the context of the ageing Australian population, increasing numbers of comorbidities in hospitalised patients, and increasing complexity and intensity of medical therapies. Patients need to receive therapies that are beneficial to them, and do not cause significant harm or discomfort, and their wishes and choices regarding such care must be considered. Experience with the MET model suggests that many deteriorating patients receive suboptimal end-of-life care planning, and that this planning is done emergently by the MET. Advance care planning should be conducted when a patient is cognitively intact and able to contribute meaningfully to discussions, so that their preferences are known when deterioration is detected. In a recent trial, an advance care planning program improved the rate at which patients received preferred care, improved patient and family satisfaction, and did not increase patient mortality.32 Detecting and recognising early deteriorationImproved detection of deterioration requires tailoring of vital signs measurement to a level commensurate with a patient’s diagnosis, acuity, proposed treatments and change in status with time.29 In some patients, this may include the use of continuous automatic electronic monitoring systems that record observations and alert staff when trigger points are reached.33 Research shows marked variation in the quality and ease of use of vital sign charts.34 Enhanced design of observation charts has been associated with improved documentation of vital sign sets and recognition of deterioration.35 Improving management and follow-up of early deteriorationImproved training in the recognition of and response to early deterioration is needed. A multifaceted approach to this type of intervention — involving a newly designed observation chart, a track-and-trigger system and an education package — has been shown to be associated with increased MET activation, reduced unplanned ICU admission and reduced in-hospital mortality.36 Optimising the outcomes of MET patientsAs delayed MET activation worsens patient outcomes,18,19 research is needed to better understand the decision-making processes of ward staff with regard to escalating care for deteriorating patients. Specifically, it is important to understand the circumstances under which staff call their parent unit, as opposed to requesting external assistance. Auditing the clinical causes of MET calls may assist in the development of strategies to improve the outcomes of patients who fulfil MET criteria. The concept of “MET syndromes” has been raised,37 and it has been shown that sepsis, heart failure, pulmonary oedema and arrhythmias (especially atrial fibrillation) may be associated with about 50% of MET calls.38 Implementing checklists, guidelines or bundles of care for common causes of MET review may help to standardise management and improve outcomes of patients after a MET call has been activated. A simulation-based training course for MET members has been shown to be associated with improved team performance, but the effect of such training on MET patient outcomes has not yet been tested.39 ConclusionIn the 50 years since the components of cardiopulmonary resuscitation were described, outcomes of in-hospital cardiac arrest have changed little. The failure of cardiac arrest teams to substantially improve outcomes, in conjunction with research showing that cardiac arrests are often preceded by signs of deterioration, facilitated a paradigm shift to encourage detection and escalation of care before the onset of circulatory arrest. Although the effectiveness of the MET approach continues to be debated, METs have highlighted the need for even earlier detection and escalation of care for deteriorating patients. Another paradigm shift is now needed to improve prediction, detection and recognition of deterioration, and to optimise subsequent escalation of care. 1 Criteria for escalation of care and rates of mortality at various phases of patient deterioration Hospital inpatient27 Abnormal vital signs15 MET call22-25 Cardiac arrest1-4 Event rates (per 1000 admissions) — 138 10–50 0.66–11.2 Criteria for escalation Heart rate (beats/min) — ≤ 50 or > 120 < 40 or > 130 0 Systolic blood pressure (mmHg) — < 100 < 90 or > 200 0 Respiratory rate (breaths/min) — ≤ 10 or > 28 < 8 or > 35 0 Other — — Fall in conscious state or urine output Unresponsive In-hospital mortality rate 2.5% 14.6% Approx 25% Approx 80% 2 Phases of patient deterioration and strategies that may improve outcomes at each phase MET = medical emergency team. ICU = intensive care unit. 3 Essential elements for the care of deteriorating patients29 Clinical processes Measurement and recording of observations Escalation protocols Rapid response systems Communication processes Organisational prerequisites Organisational supports Education Evaluation and monitoring Use of new technology

Daryl A Jones FRACP, FCICM, MD · Nicola J Dunbar BSc(Hons), PhD, MPS · Rinaldo Bellomo FRACP, FCICM, MD

Fast facts: chronic and cancer pain. 2nd ed. Michael J Cousins, Rollin M Gallagher. Oxford: Health Press, 2011 (167 pp, $25.00). ISBN 97819059832859. Australian Michael Cousins and American Rollin Gallagher are both internationally recognised pain medicine physicians, active in research, teaching and clinical work. Which is why, when my review copy of their book arrived in the post, I was surprised to find it to be so small. It made me recall Mark Twain’s famous comment: “I wrote you a long letter because I did not have the time to write a short one”, and hoped that the necessary time and thought had gone into this little handbook. I was not disappointed. The book explains the new insights into the physiology of persistent pain in a simple fashion. It is well illustrated and provides an up-to-date guide to various treatments. The chapters cover trigeminal neuralgia, complex regional pain syndrome, diabetic and post-herpetic neuropathic pain, central pain, musculoskeletal pain, visceral pain and headache. It is ideally suited to all medical practitioners treating pain including hospital medical officers, general practitioners and various medical specialists. It would also be useful to nurses, allied health clinicians, teachers and medical trainees. The book is balanced in its viewpoint, emphasising the multifactorial nature of persistent pain conditions, and that treatments require both medical as well as restorative approaches to physical function and psychological aspects. The chapter on cancer pain is well done. It offers a good overall summary of pathophysiology of cancer pain and the complications of therapy associated with pain. It summarises pharmacological, non-pharmacological and interventional strategies to manage cancer pain very succinctly. However, it comprises only a small section of the book and fails to cover palliative care or the existential nature of cancer pain and suffering. The introduction reminds readers that “regardless of specialty, clinicians will see patients with persistent pain”. I agree with this statement and, in my opinion, there is a need for concise texts about pain. The authors have definitely tapped into this need with this book.

Carolyn A Arnold

To the Editor: We write in response to the letter by Awerbuch1 and agree with many of his points. He has raised an interesting issue regarding the assertion that chronic pain is itself a disease,2 suggesting it would then logically follow that the patient becomes the final arbiter of whether he or she has the “disease” and hence which treatment may or may not be appropriate. A disquieting development in this regard is the recent Declaration of Montréal, produced at the International Pain Summit of the International Association for the Study of Pain in September 2010.3 This declaration states that access to pain management should be considered a fundamental human right. The position of diagnosis is uncertain, and the only specific treatment modality mentioned is opioid therapy. While we are sure that the Declaration is noble in intent, where does it place a clinician who has concerns about prescribing opioids to a patient who demands them? It adds the legal threat of a breach of human rights if the patient is disaffected with a doctor’s decision on opioid prescribing. As Awerbuch and others4,5 have stated, the public health consequences of prescribed opioids are not trivial. Assessing the appropriate circumstances for their use requires an understanding of clinical evidence and due care, not dogma, moral coercion or forays into jurisprudence.

Dilip Kapur · Phillip B Cornish · Carol A Snellgrove · David A Cherry

Clinical record A 17-year-old girl presented to her general practitioner with a 1-week history of fever, arthralgia, general malaise and dry cough. She had a history of systemic onset juvenile idiopathic arthritis (SOJIA), diagnosed at age 2 years and treated with aspirin, and had been in remission for 13 years. She was initially treated by her GP with oral roxithromycin, and admitted to hospital 3 days later with worsening of her symptoms. Her admission chest x-ray (Box 1) revealed bilateral perihilar infiltrates. A diagnosis of severe community-acquired pneumonia was made and broad spectrum antibiotics were commenced, including vancomycin, moxifloxacin, and oseltamivir. Despite this treatment, her condition deteriorated. On Day 3 of admission she required endotracheal intubation and circulatory support with noradrenaline 18–50 μg/kg/min, and was admitted to the intensive care unit (ICU). She remained hypotensive, with a mean arterial pressure of 50 mmHg, and with sinus tachycardia of 132 beats/min, and subsequently required renal replacement therapy. She remained febrile for the first 3 days after admission (temperature range 37.5°C–39°C), and her temperature settled to normal after appropriate therapy was initiated. Investigations included a computed tomography scan of her abdomen, which showed hepatosplenomegaly, and liver function tests, which showed elevated conjugated bilirubin (36 mmol/L; reference range [RR], < 4 mmol/L), γ-glutamyl transferase (84 U/L; RR, < 24 U/L), lactate dehydrogenase (3540 U/L; RR, 150–280 U/L), alanine aminotransferase (108 U/L; RR, 10–30 U/L) and aspartate transaminase (388 U/L; RR, < 30 U/L). Other abnormal parameters were her haemoglobin level (93 g/L; RR, 120–160 g/L), platelet count (65 x 109/L; RR, 150–400 109/L), white cell count (2.9 x 109/L; RR, 4.5–13 x 109/L), international normalised ratio (2.1; RR, 0.9–1.2) and fibrinogen level (0.9 g/L; RR, > 2.5 g/L). Elevated inflammatory markers included C-reactive protein (301 mg/L; RR, < 5 mg/L) and serum ferritin (50 500 μg/L; RR, 7–140 μg/L). A transthoracic echocardiogram showed a left ventricular ejection fraction of 60%, a mild reduction in right ventricular contractility, and a right ventricular systolic pressure of 48 mmHg. A full screen for sepsis was performed, including a nasopharyngeal aspirate, bronchoalveolar lavage, blood cultures and serological testing; all were unremarkable. Other immunological tests performed were for Mycoplasma pneumoniae antibodies, Streptococcus pneumoniae urinary antigen, Legionella pneumophila urinary antigen, herpes simplex virus, cytomegalovirus, Epstein–Barr virus (EBV) IgM and IgG, influenza A and B, H1N1 influenza RNA, respiratory syncytial virus, parainfluenza and adenovirus DNA, Q fever (Coxiella burnetti) IgM and IgG, and serological tests for hepatitis, dengue fever IgM and Leptospira IgM, all of which were non-reactive. Urinalysis revealed a white blood cell count of 140 x 106/L (RR, < 10 x 106) and an erythrocyte count of > 500 x 106/L (RR, < 10 x 106/L) with no microbial growth on culture. A bone marrow aspirate with trephine was performed on Day 4 (Day 2 in the ICU), and EBV DNA was detected in the resulting sample using a qualitative DNA test. The patient’s failure to improve, in combination with her past history of SOJIA, hepatosplenomegaly and a very high ferritin level led to a preliminary diagnosis of macrophage activation syndrome (MAS). Immunosuppressive treatment in the form of high-dose methylprednisolone (10 mg/kg daily) and intravenous immunoglobulin (1 g/kg daily) were commenced. Inotrope and ventilatory requirements improved within 24 hours of this treatment. On Day 9 of admission she was extubated, and was discharged home on high-dose steroids 8 days later with no complications. Her bone marrow aspirate histological examination showed haemophagocytosis which confirmed the diagnosis of MAS (Box 2). Macrophage activation syndrome (MAS) is a severe, potentially fatal condition associated with paediatric rheumatic diseases. It is a form of secondary haemophagocytic lymphohistiocytosis (HLH) with uncontrolled activation and proliferation of well differentiated macrophages and T-lymphocytes.1,2 The central pathophysiological abnormality in HLH is cytokine dysfunction, resulting in uncontrolled accumulation of activated T-lymphocytes and activated histiocytes (macrophages) in many organs. High levels of cytokines are found in these patients due to ineffective natural killer cells and T-lymphocytes (a positive feedback loop started by ineffective T-cells, triggering an unopposed release of cytokines that attracts further T-cells).3 Hyperactivated macrophages cause damage in different tissues, and this is thought to be the origin of the high serum ferritin levels characteristic of these conditions. The cause of the macrophage activation is multifactorial. Ineffective cytotoxic immunity is due to underactive natural killer cells and reduced perforin production. Secondary HLH can be precipitated by infection, drugs, malignancy and rheumatic diseases. SOJIA has been linked to polymorphisms in genes controlling production of cytokines, such as tumour necrosis factor. This could in turn be a cause for MAS, leading some authors to postulate that MAS and SOJIA could be part of the same disease.4 Clinical manifestations include fever (91%–100%), hepatomegaly (90%–92.3%), splenomegaly (77%–84%), lymphadenopathy (41%–62%), neurological symptoms (47%) and rash (43%).5,6 The central nervous system is commonly affected,7 and patients may present with agitation, seizures, coma, respiratory failure from adult respiratory distress syndrome, and multiorgan failure.7 MAS has a mortality of up to 22%.8 To our knowledge, this is the only reported case of MAS in a patient nearing adult age and with a long inactive rheumatic disease period. MAS is well described in the paediatric population with SOJIA, and has a median age of 5 years at the time of presentation.9 The mean time between initial diagnosis of SOJIA and presentation with MAS is 4 years.8 Our patient was aged 17 years at the time of presentation with MAS, and had not had symptoms of, nor required treatment for, SOJIA for 13 years. Diagnosis was described by Ravelli and colleagues,7 and is based on clinical findings such as organomegaly and laboratory findings such as hypofibrinogenaemia, thrombocytopenia and elevated serum liver enzymes. A bone marrow aspirate can aid diagnosis in uncertain cases, the pathognomonic feature being the presence of well differentiated macrophages actively phagocytosing haemopoietic cells.8 Infections, medications and malignancies have all been identified as triggers for MAS. EBV DNA was found in the patient’s bone marrow aspirate, and EBV has been identified as one of the more common triggers for MAS.5,7 It is likely that EBV was the trigger for this patient’s MAS. Treatment is based on immunosuppression using steroid therapy. Cyclosporin has been used as the first-line treatment, or used in combination with corticosteroids. Other treatment options include plasma exchanges or intravenous immunoglobulins.10 This patient presented with fever, prominent respiratory failure and cardiovascular collapse. Her initial systemic symptoms and radiological findings suggested severe sepsis, most likely respiratory in origin, her SOJIA had been in remission for over 10 years, and there was no neurological involvement. These factors made clinical suspicion of MAS difficult, but her hepatosplenomegaly and highly raised ferritin levels suggested the diagnosis, which was supported by her laboratory test results (leucopenia, thrombocytopenia, hypofibrinogenaemia and abnormal liver function test results) and bone marrow aspirate histological findings (Box 2). Despite MAS predominantly presenting in a paediatric population with active disease, this case report emphasises the need to exercise diagnostic vigilance in treating young adult patients with multiorgan failure and a distant history of rheumatic disease. Lessons from practice Multiorgan dysfunction and vasodilatory shock may not be of infectious origin. If a patient fails to improve when treated with broad spectrum antimicrobials, alternative diagnoses should be sought. Although macrophage activation syndrome (MAS) primarily affects children with active rheumatic disease close to the time of diagnosis, it may occur in adults after a prolonged disease-free period. Diagnosis of MAS is based on history, clinical examination and laboratory findings. Demonstrating haemophagocytosis in a bone marrow aspirate can aid uncertain diagnosis. 1 Chest x-ray of the 17-year-old patient, taken on admission, showing bilateral perihilar infiltrates 2 Macrophage (long arrow), containing a red blood cell (arrow head), seen in the bone marrow aspirate taken from the 17-year-old patient (May–Grünwald–Giemsa stain x 100)

David Gutierrez MD · Louis Guy MB BS · Veera S Katikireddi MB ChB(Hons), MRCP(UK) · Jason P Butler MMedSci, FRACP, FRCPA · John Gowardman FRACP, FCICM

Objective: To reduce the rate of central line-associated bacteraemia (CLAB).Design: A collaborative quality improvement project in intensive care units (ICUs) to promote aseptic insertion of central venous lines (CVLs). A checklist was used to record compliance with all aspects of aseptic CVL insertion, with maximal sterile barrier precautions for clinicians (“clinician bundle”) and patients (“patient bundle”). CLAB was identified and reported using a standard surveillance definition.Participants and setting: Patients and clinicians in 37 ICUs in New South Wales, July 2007 – December 2008.Main outcome measures: Compliance with aseptic CVL insertion; rates of CLAB.Results: 10 890 CVL checklists were reviewed for compliance with the clinician and patient bundles: compliance with aseptic CVL insertion improved significantly (P < 0.001). The CLAB rate dropped from 3.0 to 1.2 per 1000 line-days (P < 0.001). Regardless of CVL type, the relative risk (RR) of CLAB in patients with CVLs inserted by clinicians not compliant with the clinician bundle was 1.62 times greater (95% CI, 1.1–2.4; P = 0.018) than the RR with CVLs inserted by clinicians compliant with both bundles. Compliance with both the bundles was associated with a 50% reduction in risk of CLAB (RR, 0.5; 95% CI, 0.4–0.8; P = 0.004).Conclusions: Compliance with all aspects of aseptic CVL insertion significantly reduces the risk of CLAB. A difficulty we experienced was that most ICUs lacked the organisation and staff to support quality improvement and audit.

Anthony R Burrell MB BS, FANZCA, FCICM · Mary-Louise McLaws DipTropPH, MPH, PhD · Margherita Murgo BN, MN(Critical Care) · Eda Calabria LLB, BBus · Annette C Pantle MB BS, MPH, FRACMA · Robert Herkes MB BS, FRACP, FCICM

To the Editor: We read with interest the recent cover article by Bhonagiri and colleagues detailing increased standardised mortality rates for patients admitted to Australian intensive care units (ICUs) out of hours.1 Have the authors considered a secondary analysis controlling for night-time staffing ratios? Aside from staffing levels, human factors such as the effects of fatigue and sleepiness due to circadian rhythm misalignment and sleep deprivation should be considered and further investigated. Studies have consistently shown that human cognitive and motor performance is substantially worse at night than during the day,2 due to the influence of the endogenous circadian clock. A recent study of emergency medicine registrars at an Australian tertiary hospital showed a 21% decrease in performance at night, in clinical scenarios of fellowship examination standard.3 Sleep during the day is of poorer quality and shorter duration than sleep during the night,4 resulting in chronic sleep restriction that exacerbates the impairment occurring in night work.5 Consecutive night shifts are associated with an increasing risk of accidents and injuries.6 In doctors, including consultants, inadequate sleep (fewer than 6 hours) is associated with harmful effects on patient outcomes.7 Despite our cultural beliefs, doctors are likely to suffer from the same biological effects of sleep loss and circadian rhythm disturbance as suffered by other humans. ICUs may provide one of the best hospital environments in which to study medical error, because of defined and limited personnel compared with non-ICU ward areas. ICUs also usually have better record-keeping and error-recognition systems, because of higher staff:patient ratios. Possible risk mitigation strategies include increasing the number of night staff, implementing oversight mechanisms for specified procedures and decisions, and ensuring that rosters and shift lengths are designed according to evidence linking sleep disturbance with poor patient outcomes. Such evidence should form the basis of any regulatory frameworks.

Dev A S Kevat · Andrew R Davies · Peter A Cameron · Shantha M W Rajaratnam

New results from an Australian collaborative randomised trial will change practice Every year in Australia, 1000 people are admitted to intensive care units with severe traumatic brain injury, mostly as a result of motor vehicle accidents.1 Despite high standards of prehospital and medical care, 50% of these people either die or survive with severe lifelong disability.2 Most of the survivors with severe disability are young men aged in their mid 20s;2 they cannot return to work and will never be able to live independently. The cost of severe traumatic brain injury in human terms is huge, and in economic terms has been recently independently calculated to be $4.8 billion every year in Australia.1 In the United States, the annual economic burden of traumatic brain injury is $60 billion.3 Half of patients with severe traumatic brain injury have haematomas that may require neurosurgical evacuation; management of these patients before and after surgery is not controversial. However, about 10% of patients have diffuse brain injury,4 and also have persistent brain swelling that cannot be effectively controlled in intensive care units by airway management, mechanical ventilation, intracranial pressure monitoring, heavy anaesthesia, resuscitation fluids, osmotherapy, and drainage of cerebrospinal fluid from ventricular catheters. Over the past decade, management of these patients has been shifting in many centres from barbiturate therapy, and sometimes hypothermia, to decompressive craniectomy5 — a well established neurosurgical procedure in which a large piece of skull bone is removed (stored for 1–2 months, and then replaced) to enable better control of intracranial pressure and a more rapid decrease of sedation, with the ultimate goal of improving patients’ long-term functional outcomes. About 20 case series of adult patients from many countries have reported mainly favourable results from this procedure.6 However, all suffer from a lack of prospective randomised control groups and of objective blinded outcome assessment at defined time points. To objectively assess decompressive craniectomy for the first time and to establish its true efficacy in adult patients with traumatic brain injury, intensivists and neurosurgeons in Australia, New Zealand and Saudi Arabia collaborated between 2003 and 2010 to conduct the Decompressive Craniectomy (DECRA) randomised trial. The results of the DECRA trial have recently been published.7 This was the first randomised trial of any neurosurgical technique to be completed in complex adult neurotrauma patients. Patients were randomly assigned to receive decompressive craniectomy plus standard care (maximised intensive care therapies, which could include barbiturate anaesthesia) or standard care alone. The study found that early decompressive craniectomy clearly and dramatically achieved its short-term goals: it decreased intracranial pressure, markedly decreased medical therapies required for intracranial pressure, shortened mechanical ventilation time, and shortened stay in the intensive care unit by 5 days compared with the standard care group. The dramatic short-term effects meant that the study was difficult to complete because neurosurgeons and intensivists were aware of group allocation, could see the early benefits, and increasingly preferred to conduct early surgery rather than recruit patients to the trial. However, despite these positive short-term effects, the findings in relation to patients’ eventual functional outcome, measured carefully at 6 months, were startling and the reverse of what had been expected. There were 19% more patients with poor functional outcomes (odds ratio, 2.21; 95% CI, 1.14–4.26; P = 0.02) and 23% more survivors with severe disability in the decompressive craniectomy group compared with patients who received standard care alone. The cause of these poor outcomes is speculative. They may have related to the surgery itself or to surgical complications (including hydrocephalus), but surgical complications seem an unlikely explanation given that the rates overall were less than those reported in published case series. A more likely explanation is that “axonal stretch” that occurred during swelling of the brain outside the skull through the craniectomy defect8 exacerbated brain injury in a way that has not previously been described in humans and was not anticipated. Multimodal magnetic resonance imaging may enable confirmation of this. We might now coin the term “brain volutrauma”, analogous to the “lung volutrauma” observed in mechanically ventilated patients with acute lung injury. A second ongoing randomised trial — the Randomised Evaluation of Surgery with Craniectomy for Uncontrollable Elevation of Intracranial Pressure (RESCUEicp) — is addressing similar questions in the United Kingdom and Europe. The RESCUEicp trial also includes patients with brain haematomas and intervenes at a slightly higher intracranial pressure threshold,9 so the results will be complementary to the DECRA trial and are likely to be equally important. What is unequivocal is that functional outcomes of future patients with severe diffuse traumatic brain injury in Australia and New Zealand can be substantially improved by choosing best intensive care medical and surgical therapies, based on established guidelines,10 without early decompressive craniectomy, despite increases in intracranial pressure that may be concerning to clinicians. The role of decompressive craniectomy as a rescue therapy in the context of very high intracranial pressure has not been resolved by this study, but is now clearly uncertain. It is also unequivocal that the Australian health care system will save many tens of millions of dollars annually when the less invasive treatment regimen is chosen. Although careful cost–benefit analyses are urgently required (and can be done using existing databases), it is already clear from independent costing analyses, which confirmed that most of the traumatic brain injury cost burden relates to rehabilitation costs of survivors with severe disability,1 that the savings to Australia of choosing intensive medical therapies instead of craniectomy in appropriate patients will be greater than $100 million annually. Savings in the US and Europe are likely to be proportionately greater. Finally, the DECRA trial illustrates with new clarity that carefully designed randomised controlled trials are the only way to correctly advance clinical practice, even in complex, critically ill patients, and despite the clear inherent difficulties of such trials.11 Recent calls for “comparative effectiveness research” in neurosurgery,12 based on aggregations of non-randomised databases, do not represent a valid substitute. In the DECRA trial, a collaborative network of clinician-investigators (the Australian and New Zealand Intensive Care Society Clinical Trials Group), which was supported by the National Health and Medical Research Council (NHMRC) and co-funders, unfettered by industry, and motivated by a strong desire to improve the quality of outcomes in critically ill patients, enabled the trial to be supported to completion and to influence clinical practice for the better. These are the essential ingredients needed to make meaningful improvements in health care.

D James Cooper MD, FRACP, FCICM · Jeffrey V Rosenfeld MD, MS, FRACS

We report a case of compassionate use of a haemoglobin-based oxygen carrier in a severely injured Jehovah’s Witness patient, for whom survival was considered unlikely. Severe anaemia and cardiac hypoxia were reversed after slow infusion of this agent. No vasoactive side effects were associated with the treatment, possibly due to the slow infusion, and the patient survived. (MJA 2011; 194: 471-473) Clinical recordA healthy 32-year-old woman was a passenger in a vehicle involved in a high-speed collision with a truck, and she was entrapped for 2 hours. Initially, her heart rate was 100 beats/min, blood pressure was 90/50 mmHg, respiratory rate was 28 breaths/min and oxygen saturation measured by pulse oximetry (SpO2) was 92% on air. Her Glasgow Coma Scale score was 9 (eye opening, 2; verbal response, 1; motor response, 6) and her pupils were equal and reactive to light. Her family indicated that she was a Jehovah’s Witness and would not accept the units of blood that had been transported to the accident site. Paramedics performed endotracheal intubation, immobilisation, left femoral splinting and resuscitation with a 7000 mL crystalloid infusion, and applied dressings to wounds. The patient was transported by helicopter to The Alfred’s trauma centre. On arrival at the trauma centre, she was ventilated, her heart rate was 120 beats/min, blood pressure was 62/26 mmHg and SpO2 was 79% on 100% fraction of inspired oxygen (FiO2). Following discussions with the family, the trauma team agreed not to treat the patient with packed red cells, platelets or fresh frozen plasma. Thoracostomies for a right tension pneumothorax and a left pneumothorax were performed, which increased her blood pressure to 105/65 mmHg and SpO2 to 100% on FiO2 100%. Intercostal drains were inserted and connected to a cell salvage device. She was administered 1400 mL of succinylated gelatin, 10 units of cryoprecipitate and 5 mg of recombinant factor VIIa. Bleeding was controlled with direct pressure, and a scalp wound was closed. Ultrasonography demonstrated a moderate pericardial effusion with systolic right ventricular collapse and free intraperitoneal fluid. Electrocardiography demonstrated sinus tachycardia with no ST segment changes. Initial blood tests showed a haemoglobin (Hb) level of 67 g/L (reference range [RR], 113–159 g/L), activated partial thromboplastin time of 44.3 s (RR, 26–38 s), international normalised ratio of 1.9 (RR, 1.0–1.3), serum fibrinogen level of 3.2 μmol/L (RR, 5.9–11.8 μmol/L, lactate level of 3.9 mmol/L (RR, 0.6–2.2 mmol/L) and serum creatinine level of 55 μmol/L (RR, 60–105 μmol/L). Imaging showed a fractured right orbit and maxilla, bilateral rib fractures, a grade 4 splenic laceration, a likely jejunal injury with intramural haematoma, a left distal humerus fracture, a comminuted open left femoral shaft fracture, an unstable T12/L1 fracture dislocation (60% off-ended), and multilevel spinous process and transverse process fractures. Laparotomy and fixation of the patient’s thoracolumbar injury were deferred because of the likelihood of associated bleeding. Instead, she underwent splenic embolisation, external fixation of her open left femoral shaft fracture and debridement of her left humerus injury. She received 1000 mL of 4% albumin and 1000 mL of crystalloid fluid during these procedures, and 10 mg of intravenous vitamin K afterwards. On postoperative admission to the intensive care unit, her Hb level was 36 g/L and her coagulation profile was normal. Low-dose noradrenaline was required to support her blood pressure until Day 2. Her urine output over the first 24 hours was 4500 mL. A follow-up transthoracic ultrasound showed abatement of the pericardial effusion. An abdominal computed tomography (CT) scan with oral contrast excluded jejunal injury. To protect renal function, intravenous contrast was not used. Placement of an inferior vena cava filter was deferred because of anatomical distortion secondary to the thoracolumbar injury. Several strategies were used to manage the patient’s anaemia. Sedation minimised metabolic demand. A ventilation cycle of 2 hours of 90% FiO2, followed by 2 hours of 90% SpO2 and then 20 hours of 95% SpO2 was used. This was employed to maximise oxygen delivery while minimising shunt from absorption atelectasis and to promote erythropoiesis. Recombinant erythropoietin (36 000 units daily for 6 days), folic acid (5 mg daily continued until discharge), vitamin B12 (1 mg daily for 6 days) and a single iron infusion of 500 mg were administered to maximise haematopoiesis. Menses was inhibited with progesterone. Blood testing was performed using paediatric-sized samples. Pneumatic calf compressors were applied and regular lower-limb sonography was performed to exclude venous thrombosis. The trauma team considered using a synthetic haemoglobin-based oxygen carrier (HBOC) to increase oxygen delivery to the patient’s tissues. On Day 3, OPK Biotech (Cambridge, Mass, USA), the Therapeutic Goods Administration (TGA), the Australian Quarantine and Inspection Service and airline carriers were contacted to determine availability and import permissions. HBOC-201 was supplied by OPK Biotech without charge. Informed consent for use of HBOC-201 was obtained from the patient’s family. Approval for emergency compassionate use of HBOC-201 was obtained from The Alfred Ethics Committee on Day 4. The published and unpublished in-vivo and in-vitro research into HBOC-201 was reviewed at a multidisciplinary meeting, and its use was agreed to. Ten 250 mL units of HBOC-201 were imported under Category A of the TGA’s Special Access Scheme. By Day 5, the patient’s Hb level had dropped to 29 g/L and her serum troponin I level was 0.33 μg/L (RR, < 0.10 μg/L), indicating cardiac hypoxia (Box). An electrocardiogram showed widespread ST depression and an episode of non-sustained ventricular tachycardia was documented. Survival with this degree of metabolic demand, the associated anaemia, and resultant end-organ hypoxia was considered unlikely. Following advice from experienced United States physicians, 3 units of HBOC-201 were administered on Day 5, and a further 2 units were administered on Day 6 with ascorbic acid (1 g twice daily continued until discharge). Each unit of HBOC-201 was infused over 8 hours to minimise any adverse effects related to volume overload, vasoactivity or methaemoglobin. Intravenous glyceryl trinitrate was the agreed treatment in the event of hypertension,1 but this was not necessary. After the slow administration of 5 units of HBOC-201, the patient’s Hb level increased from 35 g/L to 62 g/L (Box). Echocardiography performed before and after HBOC-201 treatment showed a reduction in cardiac output from 6.8 L/min to 5.0 L/min. Electrocardiography findings and troponin I levels returned to normal and no further arrhythmias were noted. Somatosensory evoked potentials revealed intact lower-limb neurological pathways. On Day 7, closed reduction was performed and a body cast was applied to treat the T12/L1 fracture dislocation. Imaging showed improved alignment, and an inferior vena cava filter was placed. From Day 6, the patient’s temperature began spiking, secondary to femoral pin site infections and pneumonia. Despite treatment with antibiotics, temperature spiking continued. Cooling was commenced to minimise metabolic demand. A transoesophageal echocardiogram on Day 11 showed right ventricular regional wall motion abnormality but no evidence of endocarditis. A CT scan showed persistent bilateral pneumothoraces, a left pleural effusion and an epidural haematoma at L1 level. Bilateral tube thoracostomies were re-performed and an antifungal was added to the anti-infective regimen. A percutaneous tracheostomy was performed on Day 12. Low-dose heparin therapy for thromboprophylaxis was deferred until Day 17. By Day 21, the infections had resolved and the tracheostomy tube was removed. The femoral and humeral fractures were internally fixed on Day 20 with minimal blood loss. On Day 30, the patient’s Hb level was 107 g/L and operative reduction and internal fixation of her thoracolumbar spine was performed. She was well when discharged to a rehabilitation facility on Day 43 — her cognition was formally assessed as normal, lower-limb neurological pathways were intact and Hb level was 101 g/L. DiscussionWe have described compassionate use of HBOC-201 in a severely injured Jehovah’s Witness patient. To our knowledge, this is the first report to describe reversal of documented cardiac hypoxia secondary to anaemia following trauma. Haemorrhagic shock is responsible for one-third of deaths following high-energy trauma.2 Integrated trauma care systems coordinate rapid haemorrhage control, shock recognition and surgical interventions to minimise blood loss and coagulopathy.3 Healthy volunteers can tolerate Hb levels of 50 g/L without evidence of end-organ hypoxia.4 However, it is estimated that the median Hb concentration associated with mortality is about 25 g/L.5 During the phase of increased metabolic demand in our patient, there was evidence of cardiac hypoxia when her Hb level reached 29 g/L. This prevented further operative interventions and placed her at high risk of cardiac dysrhythmias and death. HBOC-201 is a modified lactated Ringer’s solution containing 130 g/L of polymerised Hb of bovine origin. It is compatible with all blood types, stable for 3 years when stored at 2–30°C and stable for 2 years when stored at 40°C. When fully saturated, HBOC-201 has the same oxygen-carrying capacity as whole blood with the same Hb concentration. The partial pressure of oxygen at which HBOC-201 is 50% saturated (40 mmHg) is higher than that for cellular Hb (27 mmHg), which facilitates oxygen delivery to tissues. The half-life of HBOC-201 is approximately 20 hours.6 Polymerisation of the Hb reduces its glomerular diffusion and nephrotoxicity. A potential complication of HBOC-201 administration is hypertension and increased left ventricular afterload. Infusing each unit slowly (over 8 hours) in our patient may have diminished any vasoactive side effects. Two case reports of using HBOC-201 to treat severe anaemia following blunt trauma have been published. The first described improved cerebral oxygen delivery, but not survival, in a patient with head injuries.7 The second described successful reversal of haemorrhagic shock in a patient whose Hb level dropped to 45 g/L before HBOC-201 administration.8 However, the lack of clear HBOC-201 transfusion indications and end points, as well as the lack of data to support widespread use of HBOCs, has been criticised.9 A meta-analysis of data from HBOC trials has demonstrated an increased incidence of myocardial infarction and death in anaemic patients without life-threatening haemorrhagic shock.10 However, the analysis did not address the issue of “risk versus benefit” for use of these agents, including HBOC-201, in cases where blood transfusion for severely anaemic patients is not possible. A subsequent series of 54 consenting non-trauma patients with a median Hb level of 40 g/L demonstrated improved chances of survival with no serious adverse events following HBOC-201 administration.11 When blood transfusion is not possible, HBOCs can sustain oxygen delivery to hypoxic tissues.12 Such treatment may represent a life-saving intervention for patients with acute anaemia.13 Interest in safe and effective red blood cell substitutes for oxygen transport is increasing. Agents such as HBOC-201 show particular promise and could make a large difference to survival of trauma patients when blood is not accessible, available or acceptable. Haemoglobin and troponin I levels of a woman who was treated with a haemoglobin-based oxygen carrier (HBOC) following severe trauma * Troponin I levels were measured using the Architect i2000 immunoassay analyser (Abbott Diagnostics, Abbott Park, Ill, USA).

Mark C Fitzgerald MB BS, FACEM · Julie Y Chan MB BS(Hons), BMedSci · Andrew W Ross MB BS, FANZCA · Susan M Liew MB BS(Hons), FRACS(Orth) · Warwick W Butt MB BS, FCICM, FRACP · David Baguley MB ChB, BSc(Hons) · Hatem H Salem MB BS, FRACP · Matthias K Russ Orthopaedic and Trauma Surgeon (Germany) · Conor Deasy FACEM, FCEM, MB BCh BAO · Katherine E Martin MB BS, BMedSci, FRACS · Joseph K Mathew MB BS, MS · Jeffrey V Rosenfeld FRACS, FRCS(Edin), FACS

Clinical record A 33-year-old woman was admitted to our service for investigation and management of acute hepatitis. She reported symptoms of nausea, fatigue, pruritus and right upper abdominal discomfort. The symptoms had first occurred 3 weeks earlier, resolving over 5 days, then recurred 2–3 days before presentation. The symptoms were temporally related to varicose vein sclerosing procedures, of which she had had three in total. The first procedure occurred 4 weeks before admission, with no subsequent side effects. It was after the second procedure, 1 week later, that symptoms first developed. The final procedure occurred a week before admission. During each procedure, the patient was given methoxyflurane as an inhaled analgesic administered from a 3 mL disposable cartridge. Other medications administered during the procedures were the sclerosing agent sodium tetradecyl sulfate and fexofenadine. There was no history of exposure to alcohol or to other prescription or over-the-counter medications. The patient had no risk factors for viral hepatitis, and there was no history of hepatitis or liver disease in her family. Clinical examination revealed jaundice and mild tender hepatomegaly only. Initial investigations showed hepatic enzymosis, with an elevated alanine transaminase level (2710 U/L [reference range, < 34 U/L]) and hyperbilirubinaemia (bilirubin 92 μmol/L [reference range, < 20 μmol/L]). Markers of liver synthetic function (albumin and prothrombin time) were within normal limits, as were full blood counts, electrolyte levels and renal function. Abdominal ultrasound demonstrated a normal-sized spleen and mild hepatomegaly, with an increased liver echotexture. Doppler sonography of the portal vein was normal. Serological tests for hepatitis A, B and C, Epstein–Barr virus, cytomegalovirus and HIV were negative. Iron and copper studies and levels of α-1-antitrypsin, antinuclear antibodies, antimitochondrial antibodies, anti-liver/kidney microsomal antibodies, anti-smooth-muscle antibodies and antinuclear cytoplasmic antibodies were all normal. Paracetamol was undetectable. Bilirubin levels continued to rise over the following 8 days (peaking at 202 μmol/L), although liver synthetic function remained normal throughout this time. A liver biopsy revealed evidence of resolving acute hepatitis with confluent perivenular hepatocyte dropout and bridging necrosis. There was no evidence of cholestasis or underlying fibrosis. The pathological diagnosis was of an idiosyncratic drug reaction, with the implicated drug being methoxyflurane. The patient’s condition continued to improve, with resolution of symptoms over 4 weeks and associated normalisation of liver enzyme and bilirubin levels. She has since remained well, and has been advised to avoid future exposure to methoxyflurane. Methoxyflurane, a short-chain halogenated ether, is a volatile anaesthetic agent. It was used for inhalational anaesthesia in the 1960s, but was withdrawn from use for this purpose when newer anaesthetic agents with more acceptable side effects became available.1,2 Methoxyflurane also has significant analgesic properties at subanaesthetic concentrations,3 and is thought to have minimal side effects in analgesic doses.4 These properties led to its adoption for use as an analgesic in a variety of settings for over 40 years.4 It is widely used by paramedic services in Australia,5 and has recently been studied for use in procedural analgesia in children and adults.4-6 It is provided in single-dose, pre-filled delivery devices (Penthrox, Medical Developments International, Melbourne, Vic), allowing accurate dosing and convenient delivery. In 2010, Penthrox was added to the Australian Schedule of Pharmaceutical Benefits as an item available free of charge for doctors’ bags. Although subanaesthetic doses of methoxyflurane (in the form of Penthrox) are used widely in Australia by ambulance services for prehospital analgesia, there is a paucity of data on its efficacy and safety. A recent observational case series and a review article found no significant side effects associated with its use for this purpose.4,5 Over three million inhalers have been dispensed in Australia since 1970,4 with the majority of doses administered for single-episode analgesia. Lessons from practice Taking a history of all medication exposures is important in assessing acute hepatitis. Repeated exposure to methoxyflurane may increase the risk of acute hepatitis. Reporting of suspected adverse drug reactions, such as this case, are important to raise awareness of possible rare side effects of commonly used medications. Hepatotoxicity resulting from the use of methoxyflurane as an inhalation agent in general anaesthesia is well described.1-3 However, hepatotoxicity associated with low doses of methoxyflurane for analgesic purposes appears to be rare. Three cases of hepatitis complicating methoxyflurane use (at subanaesthetic doses) during labour have been reported.7,8 In another case report, repeated exposure in the form of misuse of methoxyflurane was found to be associated with hepatotoxicity.9 The mechanisms of methoxyflurane-induced hepatotoxicity are unclear and may be multiple. Adverse effects of halogenated ethers are thought to be related to immune-mediated, direct toxic effects of metabolites and/or host idiosyncrasy.1 Reactive intermediates formed during metabolism of methoxyflurane can lead to tissue acetylation, with proteins modified by acetylation forming neoantigens that may trigger an immune response.1 Drug re-exposure has also been implicated as a factor contributing to methoxyflurane-induced hepatitis.2 Although unproven, it is possible that our patient’s repeated exposure to the drug may have contributed to the development of hepatitis through dose-dependent toxicity. In the prehospital setting, where methoxyflurane is being widely used, the side effect profile is minimal.5 It seems the exposure to methoxyflurane in our patient was the likely cause of acute hepatitis, and it may be that repeated exposure was a contributing factor. This observation has implications for the way methoxyflurane is prescribed, including its use for procedural analgesia in cases in which several procedures (and hence, repeated dosing) are required.

Kacey M O’Rourke BAppSc, MB BS · Stuart McMaster MB ChB, FRACGP · Karin M C Lust MB BS, FRACP

CorrectionIncorrect subheadings in box: In “Increased mortality associated with after-hours and weekend admission to the intensive care unit: a retrospective analysis” in the 21 March 2011 issue of the Journal (Med J Aust 2011; 194: 287-292), two subheadings in Box 8 (B and C) were transposed. The correct subheading for Figure B is “Elective surgical patients — day of the week SMR” and the correct subheading for Figure C is “Emergency medical and surgical patients — hourly SMR”. The correct version can be viewed online at http://www.mja.com.au/public/issues/194_06_210311/bho10921_fm.html.

Deepak Bhonagiri · David V Pilcher · Michael J Bailey

Objective: To study variation in mortality associated with time and day of admission to the intensive care unit (ICU).Design: Retrospective cohort analysis using the Australian and New Zealand Intensive Care Society Adult Patient Database.Setting and participants: 245 057 admissions to 41 Australian ICUs from January 2000 to December 2008.Main outcome measures: Observed mortality and standardised mortality ratio (SMR) based on Acute Physiology and Chronic Health Evaluation III, 10th iteration (APACHE III-j) scores. Subgroup analysis was performed on the basis of elective surgical or emergency admission to ICU.Results: 48% of patients were admitted after hours (18:00–05:59) and 20% of patients were admitted on weekends (Saturday and Sunday). Patients admitted after hours had a 17% hospital mortality rate compared with 14% of patients admitted in hours (P < 0.001); and SMRs of 0.92 (95% CI, 0.91–0.93) and 0.83 (95% CI, 0.83–0.84), respectively. Weekend admissions had a 20% hospital mortality rate compared with 14% on weekdays (P < 0.001), with SMRs of 0.95 (95% CI, 0.94–0.97) and 0.92 (95% CI, 0.92–0.93), respectively. Variation in outcome with time of admission to ICU was accounted for predominantly by elective surgical patients.Conclusions: Patients admitted to ICUs in Australia after hours and on weekends have a higher observed and risk-adjusted mortality than patients admitted at other times. Further research is required to determine the causes and relationship to resource availability and staffing.

Deepak Bhonagiri MB BS, MD, FCICM · David V Pilcher MRCP, FRACP, FCICM · Michael J Bailey PhD, MSc, BSc(Hons)

To the Editor: While clinicians often communicate with patients and families, documentation of these conversations is inconsistent. Documentation is critical for continuity of patient care, medicolegal reasons and research,1 and is particularly important in the intensive care unit (ICU), where discussions regarding prognosis and withdrawal of care occur frequently. There are scant published data on documentation of conversations with patients in ICUs and their families. We conducted a retrospective audit of patients admitted to the ICU of Wesley Hospital (a 500-bed private teaching hospital in Brisbane) between 1 January and 31 August 2009 to: determine levels of documentation of communication with patients and their families by the ICU medical staff; and compare this with documentation of communication by the primary physician before and after admission to the ICU. After obtaining Wesley Hospital ethics committee approval, all patients who were cared for in the general ICU for more than 5 days were studied. During the study period, there were 862 ICU admissions, 100 of which met our inclusion criteria. The charts of only 82 patients could be successfully retrieved and these were used for final analysis. These patients had a mean age of 64 years (SD, 14 years), and a mean Acute Physiology and Chronic Health Evaluation II (APACHE II) score of 17 (SD, 7). Sixty patients (73.1%) survived to hospital discharge. In the ICU, there were 39 family conferences at any stage between an ICU clinician and the families documented for 24 patients (29% of patients; 1.6 conferences per patient). Eleven of these conferences took place in the first 48 hours, eight between 2 and 4 days, and 20 beyond 96 hours of ICU admission. The Box shows that there was a greater proportion of documented family conferences for patients who died compared with those who survived at both 2–4 days (27% v 3%; P < 0.01) and beyond 96 hours (86% v 1.5%; P < 0.001). Of note, there was no documented communication between the hospital’s admitting physician and patients or families before ICU admission or after discharge from the ICU. Intensivists use any prior discussions to make decisions about continuing or withdrawing care.2 The absence of documentation before ICU admission is clinically relevant in this context. Potential reasons for inadequate documentation include (i) discussions occurring informally at the bedside or outside of the ICU (corridor conversations) and (ii) nurses providing updates in the clinician’s absence. Better documentation in the charts of dying patients may reflect their longer ICU stays, which provide more opportunity for communication. Moreover, discussions on treatment withdrawal are more likely to be documented as they are often a shared decision-making process. Globally, levels of documentation vary (10%–90%).3,4 Potential strategies to improve documentation include: bedside reminders (such as “have you documented family conferences?”); availability of a communications officer; an internal appointments system for formal discussions; and the use of communication kits.5 Despite being a retrospective study, our findings suggest a need to improve levels of documentation, and may prompt clinicians to examine their documentation practices and develop protocols to improve record keeping. Comparison of the documented communication rates for intensive care unit (ICU) patients who died and those who survived

Riad L Silcock · Bala Venkatesh · Ranald L Pascoe · Dianne K Fisher

To the Editor: In their recent letter, Colquhoun and Jenkins1 correctly note that the external jugular venous pressure is as reliable as the internal jugular venous pressure in estimating right atrial pressure. Furthermore, the external jugular vein is more readily visible and accessible for cannulation should accurate measurement be required. While routine clinical observation is important, direct invasive measurement of right atrial pressure may be required in patients who are critically ill or experiencing rapid fluid shifts, as when undergoing major surgery. Potential serious complications of central venous cannulation are a significant barrier to direct measurement, but it has been suggested that cannulation of a central vein may not be required to assess right ventricular filling pressures.2 Indeed, the early observations of direct measures of venous pressure by Berger3 and others in the 1930s were performed in peripheral veins. A more recent study by Amar et al2 showed a reliable correlation between peripheral venous pressure (PVP) and central venous pressure (CVP) in 150 patients without cardiac disease undergoing major non-cardiac surgery. PVP, measured in either the hand or forearm, was found to be 2–3 mmHg higher, on average, than CVP, with similar changes when fluid boluses were administered. Further studies have shown similar utility of PVP measurement in cardiac surgery,4 neurosurgery and paediatrics.5 The insertion of central venous catheters for pressure monitoring alone may not be warranted if connecting a pressure transducer to a simple cannula in a peripheral vein can attain the same information. Using peripheral venous pressure measurement, the risks of arterial puncture, pneumothorax and central venous catheter-related bloodstream infections can easily be avoided.

Stuart D Marshall

The importance to patient safety of clinician-led mortality auditing needs system-wide recognition In April 2010, the Council of Australian Governments (COAG) agreed on health and hospitals reform, with the establishment of the National Health and Hospitals Network. The aims of the network include “helping patients receive more seamless care across sectors of the health system” and “improving the quality of care” with “high-performance standards”.1 As a key component of the reforms and a funding condition, health facilities will be required to regularly report performance data to the federal government. Data will be based on national performance indicators that are already agreed to by COAG and address “access to services, quality of service delivery, financial responsibility, patient outcomes and/or patient experience”.2 Disappointingly, the COAG reforms appear to neglect clinical patient outcome data that are reported for the purpose of monitoring and improving patient safety, not least of which are mortality data. In Australia, pooled data on anaesthesia-related and surgery-related perioperative mortality are routinely analysed by the Australian and New Zealand College of Anaesthetists Mortality Working Group and the Royal Australasian College of Surgeons’ Australia and New Zealand Audit of Surgical Mortality, respectively.3 The practice of anaesthesia is highly regarded for its patient-safety record, and mortality reporting is considered an important tool in monitoring safety by informing standards of care with respect to equipment, techniques and classification of patients’ fitness.4 This year will see the first national public reports on surgical mortality in Australia, with early reports from Western Australia supporting the argument that clinician participation and leadership in mortality audits produce changes and improvements in patient care. Data from WA show that the proportion of deaths associated with deficiencies of care has fallen, and 73% of participating surgeons have changed their practice in at least one way.5 Clinician-led mortality reporting can contribute meaningfully to health reform but, sadly, there are few other instances of peer review of treatment-associated mortality and centralised public reporting in Australian health systems. Closer examination reveals that even the national anaesthesia dataset is incomplete because several states do not participate in mortality audits.3 Cooper and Gaba, in an appraisal of international anaesthesia-related mortality reporting, explain the limitations of, and possible reluctance to participate in, mortality reporting, which they claim is “plagued by confounding variation in definitions, relatively small sample sizes from selected institutions, and the lack of large population studies”.4 The Special Committee Investigating Deaths Under Anaesthesia (SCIDUA) in New South Wales — the longest-serving committee of its sort in Australia — is an excellent model of clinician-led mortality reporting.6 Appointed by the NSW Minister for Health and administered by the Clinical Excellence Commission (CEC), SCIDUA reviews all deaths occurring within 24 hours of anaesthesia or sedation. Data analyses by SCIDUA, which this year is celebrating its 50th anniversary, have substantially contributed to mortality reporting internationally.3,7,8 The Australia and New Zealand Audit of Surgical Mortality developed from the SCIDUA model via the NSW Special Committee Investigating Deaths Associated With Surgery (established in 1993, the latter is now the Collaborating Hospitals Audit of Surgical Mortality, and it too is administered by the CEC). SCIDUA’s terms of reference provide solutions to several of the limitations described by Cooper and Gaba4 and are useful starting points for other groups establishing mortality registers.6 Important starting points include a clearly defined preoperative period and phrasing of degrees of contribution to death. SCIDUA’s registry includes both expected and unexpected deaths, enabling identification of unanticipated emerging threats to safety, such as those associated with new drugs and procedures. Patients are classified according to their risk of death from comorbidities. This classification allows data about expected and unexpected deaths to be analysed separately, an essential requirement for trend analysis when concomitant increases in patient and surgical complexity could confound mortality rates. All sources of data obtained by SCIDUA are protected by qualified privilege under section 23 of the Health Administration Act 1982 (NSW). Of historical interest, SCIDUA was responsible for this section of the Act, which ensures qualified privilege to peer-review committees across all clinical disciplines within NSW. Recent changes to the NSW Public Health Act 1991 and Public Health (General) Regulation 2002 have led to a modified procedure for reporting deaths occurring within 24 hours of anaesthesia or sedation (now classified as a Category 1 scheduled medical condition). However, the new procedure retains not only mandatory notification of perioperative death but protection by privilege of anaesthetists who voluntarily submit information and analyses. These provisions encourage frank and comprehensive reporting, evidenced by the breadth of information available for analysis.3 Lessons in maintaining patient safety generated through SCIDUA are communicated widely. SCIDUA sends a confidential report outlining its conclusions about the circumstances contributing to death to the notifying anaesthetist. Pooled de-identified data are incorporated into the Australian and New Zealand College of Anaesthetists’ national triennial mortality report.3 The health care community is alerted to perceived safety risks through an annual report provided to the NSW Minister for Health and through periodic reports published in national and international journals.9,10 All deaths in all health facilities should be subject to clinical scrutiny. De-identified and pooled data should be systematically analysed for the purpose of continually monitoring patient safety as therapies change. The models provided by SCIDUA and, more recently, the Australia and New Zealand Audit of Surgical Mortality are there to be applied by all health care providers. Food for thought!

Leonie M Watterson MB BS, FANZCA, MClinED · Ross B Holland MB BS, FANZCA, FHKCA · Jan M Davies MSc, MD, FRCPC · Clifford F Hughes AO, MB BS, FRACS