Emergency medicine

Objective: To investigate species-specific envenoming rates and spectrum of severity of funnel-web spider bites, and the efficacy and adverse effects of funnel-web spider antivenom.Data sources: Cases were identified from a prospective study of spider bite presenting to four major hospitals and three state poisons information centres (1999–2003); museum records of spider specimens since 1926; NSW Poisons Information Centre database; MEDLINE and EMBASE search; clinical toxinology textbooks; the media; and the manufacturer’s reports of antivenom use.Data extraction: Patient age and sex, geographical location, month, expert identification of the spider, clinical effects and management; envenoming was classified as severe, mild–moderate or minor/local effects.Data synthesis: 198 potential funnel-web spider bites were identified: 138 were definite (spider expertly identified to species or genus), and 77 produced severe envenoming. All species-identified severe cases were attributed to one of six species restricted to NSW and southern Queensland. Rates of severe envenoming were: Hadronyche cerberea (75%), H. formidabilis (63%), Atrax robustus (17%), Hadronyche sp. 14 (17%), H. infensa (14%) and H. versuta (11%). Antivenom was used in 75 patients, including 22 children (median dose, 3 ampoules; range, 1–17), with a complete response in 97% of expertly identified cases. Three adverse reactions were reported, all in adults: two early allergic reactions (one mild and one with severe systemic effects requiring adrenaline), and one case of serum sickness.Conclusions: Severe funnel-web spider envenoming is confined to NSW and southern Queensland; tree-dwelling funnel webs (H. cerberea and H. formidabilis) have the highest envenoming rates. Funnel-web spider antivenom appears effective and safe; severe allergic reactions are uncommon.

Geoffrey K Isbister BSc, MD, FACEM · Michael R Gray MSc, PhD · Corrine R Balit BPharm, MB BS · Robert J Raven BSc, PhD · Barrie J Stokes BSc, MMath · Kate Porges FACEM · Alan S Tankel BSc, FACEM · Elizabeth Turner MAIBiol · Julian White MD, FACTM · Malcolm McD Fisher MD, FANZCA, FJFICM

Ariella Glaser,* Dwight Arakaki,† Gar Ming Chan,‡ Robert S Hoffman§ * Resident, Mount Sinai Medical Center, New York City, NY, USA; † Resident, Beth Israel Medical Center, New York City, NY, USA; ‡ Fellow (and corresponding author), § Director, New York City Poison Control Center, New York City, NY, USA. garchanATpol.net To the Editor: Two aspects of the recent article by Kelly et al comparing intranasal with intramuscular naloxone in suspected opioid overdose1 make their study difficult to interpret. The methods allowed for a great deal of bias. There was no attempt to blind evaluators to therapy, and knowing which therapy is to be used a priori may influence both therapy selection and perceived outcome. The second flaw we noted was the use of the Glascow Coma Scale (GCS) in a non-trauma patient.2 An improvement in GCS score may represent increased wakefulness or even withdrawal. The use of the GCS does not make it possible to determine what degree of improvement or worsening the therapy resulted in. In the opioid-intoxicated patient, the “alert/verbal/pain/unresponsive” (AVPU) scale is more appropriate. We agree that the use of needles in a high-risk patient is dangerous. However, if these patients do not respond to painful stimuli, there should be no danger at all.

Ariella Glaser · Dwight Arakaki · Gar Ming Chan · Robert S Hoffman

Anne-Maree Kelly,* Debra Kerr,† Paul Dietze‡ * Director, † Deputy Director, Joseph Epstein Centre for Emergency Medicine Research, Western Hospital, Private Bag, Footscray, VIC 3011. ‡ Research Fellow, Turning Point Alcohol and Drug Centre, Fitzroy, VIC. Anne-Maree. KellyATwh.org.au In reply: The prehospital setting for research poses challenges that require some flexibility in study design. While it would have been preferable to have used blinded naloxone and placebo solutions for both routes of administration in our study, financial and operational constraints made this impossible, so some bias in evaluations is possible. However, this is not necessarily in favour of the intranasal route, as before the study many paramedics were very sceptical about the intranasal naloxone preparation. Therapy selection was by random allocation in sealed envelopes as described in our article. The Glasgow Coma Scale score was chosen as an outcome measure because it was the parameter used operationally for treatment and disposition decisions in the ambulance service within which our study was conducted. We acknowledge its limitations in non-trauma patients. The potential for needlestick injury in this situation is real. Patients with opioid intoxication may be in cramped locations and may be irritable on waking, increasing the risks involved with handling a “sharp”. Given the prevalence of blood-borne viruses in the injecting drug user population, strategies to reduce the risk of needlestick injury are highly desirable. Additionally, a strategy that has been suggested for preventing opioid-overdose-related deaths is to make naloxone more widely available in the community. 1 The intranasal formulation of naloxone may be appropriate for this, as it has significant advantages including reducing risks of blood-borne virus transmission and minimising the requirement for training and the secure storage of syringes and needles. 2

Anne-Maree Kelly · Debra Kerr · Paul Dietze

1 Tsunami damage in Kandholhudoo on the Raa atoll Photograph: Andrew Robertson. Mention “the Maldives” and everyone immediately conjures up images of unspoiled coral islands, holiday resorts, spectacular diving sites and great surf. The Maldives (from the Sanskrit “mala-dvipa”, meaning “garland of islands”)1 is all that and more, from the bustling capital city of Malé to the 200 serene inhabited islands where the traditional occupations of fishing and boat building continue as they have for centuries. When the earthquake and subsequent tsunami struck Aceh on 26 December 2004, most Australians were contemplating the public holidays ahead of them. The tsunami, travelling at speeds of up to 800 kilometres per hour, struck countries around the Bay of Bengal and across the Indian Ocean. Tremors were felt in the Maldives at about 06:25 local time, and the tsunamis hit the Maldive atolls between 09:00 and 09:30. As the 1–4-metre waves struck the islands, 82 people died, 200 people were severely injured and a further 1100 required treatment. Twenty-six people remain missing. An estimated 2167 households (15 000 people or almost 5% of the population) were displaced from their homes,2 as over half the inhabited islands sustained damage (Box 1). The Australian responseLike many on Boxing Day, we had missed the early reports of the evolving disaster in Asia. However, we were soon thrust into its midst by the early morning news on 27 December 2004, and by an urgent teleconference of the Australian Health Disaster Management Policy Committee, as we considered what medical support might be needed. This Committee, chaired by the Commonwealth Department of Health and Ageing, and with State, Defence Force and Emergency Management Australia representation, played a key role in advising the Australian Government on what response could be mounted quickly. By early on 28 December, it became obvious that we needed to send civilian medical teams into the tsunami-affected areas. While the Australian Defence Force had prime responsibility for deploying medical teams into areas affected by both the 1998 Aitape (Papua New Guinea) tsunami and 2002 Bali bombing,3,4 Australia has not often deployed civilian medical teams into disaster areas. Most states and territories base their internal disaster relief medical teams around major hospitals; this is a practice which has been questioned since the 1997 Thredbo disaster.5 However, as the Western Australian State Health Coordinator in times of disaster, I knew we could put a medical team together at short notice. For the first teams, we relied on advice from Chief Health Officers and Directors of Medical Services within Australia as to who might be appropriate, available within hours and experienced in providing health care in developing countries. While effective, personal preparations were ad hoc, the initial choice of team members has since been debated, and issues such as in-country operating funds, team expenditure and telephone costs are still being resolved. 2 Destruction on Vilufushi in the Thaa atoll Photograph: Gavin Coppinger. Our key problems were time and distance, particularly as teams comprising members from different states were all leaving from Sydney (28 in two teams to Aceh and one team of 17 to the Maldives).2 For once, the “red-eye” overnight flight for those travelling from Perth to Sydney was to our advantage, enabling us to get the team to Sydney rapidly. The logistics of assembling a team, equipping it (for medical work and to live in the field), reassembling it when its role changed from surgical care to public health, and deploying it in it the 24 hours after arrival, was challenging. The team bound for the Maldives included a team leader (Andy Robertson), three general practitioners (Mark Adamski, Vince Duffy, and Grahaeme Hatfield), two public health physicians (Krishna Hort and Danny Csutoros), three emergency physicians (Colin Myers, Michael Novy and Peter Roberts), an infectious diseases physician (Dominic Dwyer), an anaesthetist (Gavin Coppinger), three nurses (Muriel Leclercq, Jeff Williams and William Kerr), a paramedic (Greg Gibson), an environmental health officer (Paul Miller) and a logistics officer (Chris Sykes). With great assistance from the NSW Ambulance Counter Disaster Unit, Westmead Hospital, Queensland Health, the NSW Fire Brigade and Emergency Management Australia, this team, along with tonnes of cargo, deployed on a loaned QANTAS 747 early on 30 December 2004, and arrived in Malé that evening. During this flight, it became clear that a doctor was required to accompany 70 injured Australians from Colombo back to Sydney, and emergency physician Peter Roberts readily volunteered. 3 Members of one of our small teams being conveyed in a small fishing boat (dhoni) Photograph: Colin Myers. In the MaldivesThe significant number of dead and injured had been well managed in the central Indira Gandhi Memorial Hospital in Malé and in regional hospitals and island medical centres.6 Having survived the initial onslaught, the Maldivians were now concerned about subsequent epidemics and other public health issues (including food and water supply), as well as primary care; our team, with its public health and infectious diseases physicians, environmental health officer and GPs had been structured with that in mind. The damage to the affected islands and the bravery of the people was noteworthy. Many reported the tsunami hitting from both sides of their island, leaving them with nowhere to run. There was a strong sense of community among the Maldivians, who banded together in this time of devastation. Maldivians pride themselves on cleanliness, and many went to neighbouring islands to help clean up. On islands such as Vilufushi and Madifushi, where near total destruction reigned and rubble lay everywhere (Box 2), “Where do you start?” was the question in everybody’s mind. The enthusiasm of the Maldivian people meant the teams were universally well received and the communities were keen to work with the teams to address local issues. Health care delivery across 200 islands was never going to be easy (Box 3). Moving personnel, equipment and resources and patients was a challenge, as virtually all transport meant traversing water. The teams used everything from small fishing boats (dhonis), Coastguard landing craft, hospital boats and ocean-going ships to seaplanes and Indian Airforce transports. 4 Reviewing patients at a clinic in Madifushi, Thaa atoll Photograph: Vince Duffy. It was critical to work closely with Maldivian Ministry of Health staff to “value add” to their efforts. This meant working in small teams with local staff throughout the Gaafu Alifu, Thaa and Raa atolls, south and north of Malé.6 Several islands had not seen medical staff since the tsunami and many were running short of pharmaceuticals — we were able to provide both (Box 4). There was a range of public health issues that needed addressing, from discouraging the use of chlorine on dead fish and animals, with resultant shortages of chlorine for the wells (Box 5), to monitoring the populations for outbreaks of dengue, scrub typhus and diarrhoeal diseases. Public health team members worked closely with the Ministry of Health’s Water and Sanitation division to implement strategies for accommodation, children’s health, water and sanitation, solid waste management and asbestos disposal. Strategies included acquiring bedding for islanders evacuated to other islands and arranging the supply of fruit and vegetables, especially for children, where local crops had been destroyed. There were also continuing problems with tsunami-related injuries. Many people on the worst affected islands had been swept out to sea, and presented with chest infections in resultant “near drowning” syndromes. Infected wounds, abrasions and crush injuries were also evident. Outbreaks of gastroenteritis and respiratory disease were fortunately uncommon, and exacerbation of locally endemic infectious diseases (including dengue and scrub typhus) had not occurred. Anxiety and depression, as the islanders struggled to come to terms with the destruction, were common. In giving health support, it was important not to become a burden on the local government. There were unfortunate cases of well-intentioned, but misguided, attempts by other international medical teams to take over the local health system or provide services that weren’t needed (eg, trauma surgery), and this placed further strain on Ministry of Health staff. 5 Damage to wells in Viligili, Gaafu Alifu atoll Photograph: Michael Novy. Courtesy of WA Health Department. ConclusionThe health response by the Maldives government was one of the few success stories after the tsunami. This rested on a well-organised, pre-existing infrastructure encompassing effective inter-island transport and island-based health care centres. Many issues remained, however, including profound anxiety about further waves; loss of the breadfruit, guava and other fruit trees following salt water contamination; contamination of drinking water; future withdrawal of foreign health care personnel; and concern that the Maldives may be forgotten in its recovery phase by both tourists and charities. Australia’s health response was rapid, effective and appropriate, but we did learn some lessons (Box 6). In the future, our response could be improved with the establishment of pre-selected state-based Disaster Medical Assistance Teams.7 Teams that later went to Aceh were state-based, and had the benefit of enough time to select, prepare and equip their personnel before deployment. The multi-jurisdictional nature of the earlier teams, however, captured the spirit of the Australian desire to assist all those affected by the tsunami. 6 Lessons learnt for team deployment Health intelligence Accurate health information needs to be provided to the teams before deployment. Team selection Military, developing country and/or rural and remote medical experience and disaster medicine training is useful. Team member flexibility is critical, especially being able to improvise and adapt to constantly changing circumstances. Interpreters, or team members who speak the local language, are highly desirable. Equipment National modular checklists of both self-sufficiency and medical stores need to be further developed, incorporating sections on primary care, paediatrics, chronic care and public health (including vaccines). There is a need for team-identifying clothing, principally vests and headwear. Communications A clear command and control structure is essential. Satellite phones with international coverage, and international roaming mobile phones are critical. Logistics Funding, insurance and indemnity issues should be resolved before deployment, including cash advances (US dollars were widely accepted) and credit cards. Guidelines on what will be funded on deployment (eg, mobile phone use, purchase of clothing) are necessary. Transport Agreements with commercial airline companies to rapidly deploy team members should be explored further.

Andrew G Robertson CSC, FAFPHM, FRACMA · Dominic E Dwyer MD, FRACP, FRCPA · Muriel G Leclercq BSc(Nursing)

Fiona M Wood,* Bess V Fowler,† Daniel McAullay,‡ Jocelyn R Jones§ * Plastic Surgeon and Director, † Epidemiologist, Burns Service of Western Australia, Royal Perth Hospital, GPO Box X2213, Perth, WA 6847; ‡ Senior Policy Officer, § Manager, Office of Aboriginal Health, Health Department of Western Australia, Perth, WA. FionawATmccomb.org.au To the Editor: People with major burn injuries (50% total body surface area or more) now have an improved likelihood of survival with the implementation of aggressive treatment regimens, including supportive therapy, nutrition, and advances in the control of sepsis. Technological developments and treatments, particularly expedient wound closure, early surgical debridement, covering of large burn wounds, early skin repair,1 use of cultured epithelial autograft2 and ventilation,3 have also contributed to improved outcomes for people with these injuries. In Australia, there are inequities in access to health services which may particularly affect Aboriginal people.4 We therefore undertook a retrospective, observational study to compare the incidence of major burn injuries, clinical and demographic characteristics of patients with burns, as well as treatment and outcomes between Aboriginal and non-Aboriginal children and adults in Western Australia between 1992 and 2002. Potential cases were identified using data linkage from the Western Australian Department of Health. Raw data came from clinical records. Of the 84 people identified with major burn injuries, nine were Aboriginal (11%) and 75 were non-Aboriginal (89%). The incidence of major burn injury among Aboriginal people is greater than expected, as data from 2001 show that 3.5% of the WA population are Aboriginal. Aboriginal people with major burn injuries were younger than non-Aboriginal people with those injuries (mean, 21 v 35 years). Eight of the nine Aboriginal people (89%) had flame-only burns, compared with 33 of 75 non-Aboriginal people (44%). No statistically significant difference was seen between the groups in the percentage of total body surface area affected, provision of treatment (including number of operative procedures, applications of cultured epithelial autografts, units of blood products used, nasogastric feeds, and antibiotic doses) or length of hospital stay. We found that, although a greater percentage of Aboriginal people sustained major burn injuries, after this group entered the hospital system they experienced comparable levels of service and outcomes to non-Aboriginal people. Further research into burn care is warranted, from culturally and environmentally appropriate prevention through to critical appraisal of outcomes.

Fiona M Wood · Bess V Fowler · Daniel McAullay · Jocelyn R Jones

Robust safety specifications and funding arrangements are needed A recent review of the safety of helicopter emergency medical services (HEMS) in the United States found that the risk of death for a HEMS crewmember (per hour engaged in the activity) was similar to that of rock climbers and skydivers.1 The study on the accident and fatality rate of HEMS by Holland and Cooksley (page 17) in this issue of the Journal2 is a timely reminder of the risks faced by HEMS crew in Australia. . . . at least one Australian state government is yet to conduct any independent audits of its contracted HEMS operators . . . Aviation safety, like patient safety, is a complex interaction of systems, human factors and technology. Many of the lessons learned in aviation in improving safety and management of risk, such as incident reporting, crew resource management and simulator training, have crossed over into medical practice. The underlying issues affecting safety are similar, and are frequently unrelated to operator error. Investigation of major incidents and accidents worldwide, in industries such as transport, mining, and indeed health, has revealed many common contributing factors identified as “latent conditions”, or failures of the system.3 These include the lack of a positive safety culture through poor governance, limited resources or misallocation of resources. Blame for accidents often lies with operator error, or active failures (slips, lapses and mistakes — errors at the level of the frontline operator), but it is the mitigation of latent failures that is likely to have the biggest impact on safety. HEMS in Australia operate in a risky environment for flight crew, medical crew and patients alike, for several reasons. First, HEMS in Australia are generally required to fulfil multiple roles, performing critical care interhospital transfer, land-on-scene response, hoist operations and search and rescue (SAR). In North America and Europe, there is generally a distinction between hoist and SAR operators and those who undertake interhospital transfers and land-on-scene response. Second, Australian HEMS operations are further complicated by the vast distances and the predominantly hot conditions, which challenge both aircraft and crew performance. All incidents with injuries or fatalities reported by Holland and Cooksley2 were flights conducted in helicopters without sufficient instrumentation for flight in cloud. Under the current regulatory requirements, flight in such aircraft over water or in rural areas at night is acceptable but is not viewed as best practice. Aircraft not equipped to fly in cloud have much lower acquisition costs than aircraft that are so equipped. Crew training and experience levels are also substantially less. Such aircraft continue to be used for HEMS in Australia, operating with minimal safety margins, as a result of inadequate funding arrangements. In Australia, the Civil Aviation Safety Authority (CASA) certifies aircraft operators to provide specified levels of service. However, CASA certification does not necessarily mean a safe operator, any more than accreditation by the Australian Council on Health Care Standards means a safe hospital. Furthermore, the supervision provided by CASA varies with the category of operation. HEMS is situated at the lower end of the oversight spectrum by virtue of the category of operations in which CASA has placed it, resulting in a level of scrutiny that, given the complexity and risk involved, is lower than perhaps required. Recategorisation of HEMS into a higher category requiring higher standards of compliance, and hence scrutiny by CASA, is probably appropriate. However, effecting regulatory change is a slow process. Given the low level of regulator scrutiny in some categories of aviation operations, the industry has recognised a need to enforce its own standards by commissioning aviation safety experts to conduct independent safety audits. For example, in high-risk areas, such as the off-shore oil industry, oil companies conduct independent safety audits of contracted helicopter operators as frequently as every couple of months. Although HEMS carry greater risk than off-shore oil work, at least one Australian state government is yet to conduct any independent audits of its contracted HEMS operators, despite this being a requirement of contract. Recent accidents in Australia2 have highlighted latent factors, such as equipment and crewing issues. However, to operate the equipment specified by either regulations or contracts, operators will only put in place systems they can afford. Maintaining the high standards mandated by this complex operating environment requires that health systems work in partnership with HEMS providers to ensure robust contract and auditing processes. This does not come without cost, and adequate funding of HEMS needs to be accepted and achieved. Cost cutting to ensure financial survival compromises the safety systems that HEMS operators endeavour to put in place. These are designed to mitigate error, and include hazard and incident reporting, training and education, audit programs, and safety officer appointment. Against this background, a group of community HEMS providers in NSW and Queensland have commissioned, at their own expense, the development of a safety and integrated risk-management framework for HEMS. This is being facilitated by a specialised aviation risk-management company, which has been responsible for the development of similar programs for the Royal Australian Air Force, commercial airlines, airports and other aviation organisations. This program is a collaborative effort by HEMS operators to exceed regulatory compliance and lead the way for best practice. The program has subsequently expanded to a trans-Tasman initiative, with a number of New Zealand operators joining the consortium. The framework will be formally launched in February 2005. Robust safety specifications and funding arrangements are essential to ensure that HEMS operations in Australia are performed at a more appropriate level.

Alan A Garner FACEM, MSc · Jeff Konemann CFS · Deanne M Keetelaar

Objectives: To determine the accident rate for Australian helicopter emergency medical services (HEMS) per 100 000 flying hours and to determine the patient mortality risk per mission from a HEMS accident.Method: Retrospective observational study of Australian HEMS flying hours and accidents from 1992–2002.Results: The calculated accident rate for Australian HEMS is 4.38 per 100 000 flying hours. One patient died as a direct result of helicopter accident in 50 164 missions. Overall, one accident occurred every 16 721 missions.Conclusions: The overall Australian HEMS accident rate is similar to that reported from other countries, with all accidents occurring in Queensland community HEMS. Helicopters flown at night under Visual Flight Rules (VFR) appear to represent a high-risk subgroup. HEMS flights do not appear to present significant mortality risk to patients being transported.

Jim Holland MB BS · David G Cooksley MB ChB, FACEM

Objective: To determine the effectiveness of intranasal (IN) naloxone compared with intramuscular (IM) naloxone for treatment of respiratory depression due to suspected opiate overdose in the prehospital setting.Design: Prospective, randomised, unblinded trial of either 2 mg naloxone injected intramuscularly or 2 mg naloxone delivered intranasally with a mucosal atomiser.Participants and setting: 155 patients (71 IM and 84 IN) requiring treatment for suspected opiate overdose and attended by paramedics of the Metropolitan Ambulance Service (MAS) and Rural Ambulance Victoria (RAV) in Victoria.Main outcome measures: Response time to regain a respiratory rate greater than 10 per minute. Secondary outcome measures were proportion of patients with respiratory rate greater than 10 per minute at 8 minutes and/or a GCS score over 11 at 8 minutes; proportion requiring rescue naloxone; rate of adverse events; proportion of the IN group for whom IN naloxone alone was sufficient treatment.Results: The IM group had more rapid response than the IN group, and were more likely to have more than 10 spontaneous respirations per minute within 8 minutes (82% v 63%; P = 0.0173). There was no statistically significant difference between the IM and IN groups for needing rescue naloxone (13% [IM group] v 26% [IN group]; P = 0.0558). There were no major adverse events. For patients treated with IN naloxone, this was sufficient to reverse opiate toxicity in 74%.Conclusion: IN naloxone is effective in treating opiate-induced respiratory depression, but is not as effective as IM naloxone. IN delivery of naxolone could reduce the risk of needlestick injury to ambulance officers and, being relatively safe to make more widely available, could increase access to life-saving treatment in the community.

Anne-Maree Kelly MD, MClinEd, FACEM · Debra Kerr RN, MBL · Zeff Koutsogiannis MB BS, FACEM · Paul Dietze PhD · Ian Patrick · Tony Walker

Daniel M Fatovich Specialist in Emergency Medicine, Royal Perth Hospital, GPO Box X2213, Perth, WA 6847 daniel.fatovichAThealth.wa.gov.au To the Editor: I have read with interest the debate in the MJA on the use of tPA in acute ischaemic stroke. Most recently, Levi et al published a position statement stating that it is a major advance.1 This was probably in response to Hoffman’s critical editorial.2 At the 10th International Conference on Emergency Medicine in June 2004, a session on the use of tPA in acute ischaemic stroke clearly portrayed thrombolysis as not standard care.3 I have attended other emergency medicine conferences where thrombolysis was seen as risking more harm than good. Conversely, I expect that stroke physicians attend stroke conferences that endorse thrombolysis. In my experience, when such divergent views exist, it usually means that we don’t have enough answers. I would like to outline here some other viewpoints that are not often considered. Number needed to harm (NNH): The best results to date were from the NINDS study that reported a number needed to treat (NNT) of 8.4 With their findings of an intracranial haemorrhage rate of 6.4%, the NNH is about 16. Hence, for every 16 patients treated with tPA, two may derive much benefit, but one much harm. These odds are worse than Russian roulette. The Cleveland study reported an intracranial haemorrhage rate of 22%.5 Hence, the worst possible NNH is about 5. Other authors have expressed similar ethical concerns.6 Risk tolerance is an individual judgement, but, when faced with the above issues, my practice is to ask what I would want for myself. Knowing that the earlier thrombolysis is given the better,7 my personal choice would be to only have thrombolysis if it is administered within 90 minutes of stroke onset (ie, maximal benefit and minimal risk). Unfortunately, it is rare for patients to present early enough for this to occur. Furthermore, many of my colleagues do not know what they would want for themselves, so how can we advise our patients? Pathophysiology: Heart muscle is relatively robust, whereas the brain is a softer structure. A haemorrhagic complication is very different in the two organs. Mode of thrombolysis: Giving thrombolysis by infusion is an outdated approach. Furthermore, thrombolysis is almost a forgotten therapy for acute myocardial infarction in tertiary centres because of the use of primary angioplasty. When thrombolysis is used, the agent is given as a bolus. Uptake of this mode of administration would be rapid if it were shown to be effective and safe for acute ischaemic stroke. Obviously, consensus among care providers on the use of tPA does not exist. This means that more research needs to be done to work out the answers to these difficult questions. I believe there is much support for this, as we need to define who should be receiving thrombolysis, and, perhaps more importantly, who should not. We all want something that works! However, we need greater knowledge to overcome the safety issues. The answer to Levi’s question “Why did it take so long?” is “Because it is a complex problem”.

Daniel M Fatovich

Ian R Rogers,* George A Jelinek,† Ian Jacobs* * Associate Professor, † Professor, Discipline of Emergency Medicine, Queen Elizabeth II Medical Centre, Nedlands, WA 6009. Ian. RogersAThealth.wa.gov.au To the Editor: We applaud the call of Levi and his co-contributors for collegiate communication and consensus regarding the use of tissue plasminogen activator in acute ischaemic stroke.1 Emergency care providers are acutely aware of their role at the centre of the acute healthcare system. Daily, we interact with colleagues from other disciplines in the course of seeking the best clinical care for our patients. However, the views expressed by Hoffman2 are shared by many emergency physicians and prehospital care providers. We remain unconvinced of the role of thrombolysis in acute ischaemic stroke outside the setting of properly constituted clinical trials. On review of the contributors list in Levi’s article, we are unable to identify a single specialist emergency medicine or prehospital care provider. Consensus is not likely to be achieved until position statements from expert groups include a strong representation from all the specialty disciplines involved in the care of stroke patients. We look forward to developments in this direction.

Ian R Rogers · George A Jelinek · Ian Jacobs

Christopher R Levi (on behalf of the Australasian Stroke Unit Network, the New South Wales Greater Metropolitan Clinical Taskforce Stroke Initiative, and the Towards A Safer Culture Stroke Expert Working Group) Director, Acute Stroke Services, John Hunter Hospital, Locked Bag No. 1, Hunter Region Mail Centre, NSW 2310. christopher.leviAThunter.health.nsw.gov.au In reply: We thank the authors for their comments on our recent position statement.1 We fully agree and accept the view of Rogers and colleagues that emergency physicians are central to the timely and safe delivery of emergency medical care in our health system. This is especially the case for a therapy such as intravenous tPA, given the narrow therapeutic window and coordination challenges. We view the development of linkages with our colleagues in emergency medicine as crucial in implementing not only tPA but also a number of acute stroke therapies showing great promise in the advanced stages of development.2 Our position statement is a starting point for broader discussion, and we are pleased that discussions between the key groups are under way. We agree that, when considering patient suitability for intravenous tPA, a number of uncertainties remain, and we fully support the rationale for the ongoing clinical trials of thrombolysis in acute ischaemic stroke (see www.astn.org.au/epithet/index.html and www.ist3.com/). The risk–benefit ratio will be improved in the 0–90-minute window, as indicated by Fatovich. However, it is likely that some patients at much later time points will also gain benefit. We would emphasise, however, that according to Australia’s independent arbiter of therapeutic safety and efficacy, the Therapeutic Goods Administration, intravenous tPA is an approved therapy if given within a 3-hour window, under appropriate clinical circumstances and within appropriate healthcare settings. Regarding the comments by Fatovich on number needed to harm, it is important to recognise that the most serious adverse outcome of intravenous tPA — fatal intracerebral haemorrhage — is already accounted for in the calculations of number needed to treat (for patients to survive free from dependency). Intra-arterial thrombolytic therapy in the form of prourokinase has been found to be effective in reducing dependency in acute ischaemic stroke, shown angiographically to be caused by middle cerebral artery occlusion.3 Feasibility issues, however, presently limit the application of the intra-arterial approach, and the relative risk of intracranial haemorrhage, even with this more targeted approach, is similar to that seen with intravenous therapy. The importance of cross-disciplinary teamwork in the effective application of current and future acute stroke therapies cannot be underestimated. Central to this is the need to develop an effective dialogue between the leaders of these teams — stroke physicians and emergency physicians. The Australasian Stroke Unit Network, the New South Wales Greater Metropolitan Clinical Taskforce Stroke Initiative, and the Towards A Safer Culture Stroke Expert Working Group are committed to the task of helping to build better links between stroke units and emergency departments.

Is there something wrong with the way CPR is presently practised? “That doctor — he should be sacked!” An elderly gentleman was talking about me, and he was doing it on the local television news! My crime was to make the observation in a letter to the MJA that “. . . regular involvement in cardiopulmonary resuscitation (CPR) makes me wish the technique had never been introduced.”1 I had waved a red rag in front of bulls. To disparage CPR creates fury in those who, professionally or otherwise, see it as the reason for their existence. My wife had warned me that I would be painted as the bad guy, and, when this happened, my daughter asked cheerfully, “Is Daddy going to be like Pauline Hanson?”. My first inkling of the coming storm was on the Monday the Journal appeared. A Perth radio personality wanted to interview me. I was not told that, immediately before me, he would be interviewing the head of St John’s Ambulance in Western Australia. “Doctors Give CPR Shock Treatment” was the headline in The Australian. A flurry of phone calls from journalists followed. The local paper picked up the story, and over the next few days a local general practitioner, the local ambulance chief, a surf-lifesaving identity and an editorial in the paper all attacked my purported position. I had several conversations with the medical superintendent of the hospital where I was working: I was free to express an opinion in the MJA, and I was not being reprimanded — as some in the media wished. But, he indicated that he was going to publicly distance the hospital from my remarks, and reaffirm hospital policy — to commence CPR in an emergency whenever a person has stopped breathing or has no pulse. (I am grateful to this medical superintendent, who expended time and effort publicly defending the hospital, and defending me for opinions which — like those expressed here — are mine alone, and not those of the hospital.) The local television station made the topic their main story, and repeatedly replayed the clip of the elderly gentleman who had survived two cardiac arrests and wanted me sacked. The story finished with the words, “Dr Mackay declined to be interviewed”. I did not know I had been invited. Someone had received and declined the invitation for me! An advanced healthcare directive Should I have my cardiac arrest while going about my duties in the emergency department — immediate defibrillation please! And maybe a whiff of oxygen. (If I don’t survive, I will be quite surprised.2) Should I arrest in the hospital dining room, forgo the mouth- to-mouth (I am squeamish about these things). I may (grudgingly) accept some chest compression, until the defibrillator arrives. But if you have not got me back after three shocks — call off the circus. Go back and finish your lunch. If I arrest in the street, you will do what you will. But I won’t be happy. I doubt you will be able to get a defibrillator to me quickly enough. If I arrest at home, I know it will be very difficult for you to do nothing. But it will be 15 minutes before the ambulance arrives. And to end up brain damaged on a ventilator is something I do not want. (But if you are clever enough to call the ambulance so that I arrest after it arrives, by all means use the defibrillator.) When I am in a bed in a hospital ward “old and grey and full of sleep”, do not use your hands to commit violence upon me — use them to comfort me. I became a member of a hospital cardiac arrest team in 1973. Formally or informally, I have been part of such teams ever since. Two things I have learnt from this: there are things I do to patients that I do not want done to me; and, an advanced healthcare directive can never be found when you want one — so I hope I will be forgiven for placing mine here (Box). I would have attended at least a couple of hundred cardiac arrests. My guess is that, sadly, only a dozen or so of these people survived to leave hospital. (A meta-analysis of 39 studies involving 33 124 out-of-hospital cardiac arrests has shown a survival rate of 6.4%.3) Some of the survivors I remember well. While one middle-aged man was telling me about his chest pain, I noticed the cardiac rhythm on the monitor change to ventricular fibrillation. I charged the defibrillator as he continued to talk. I waited until he lost consciousness and then shocked him. In seconds, he was asking me what happened. Such episodes are not unusual in emergency departments (or in the back of ambulances). Of the vast majority who have not survived, I have a clear memory of only a few. I was visiting a patient at his home. He was telling me about his “gallbladder pain” when he had a cardiac arrest. His wife phoned an ambulance while I began resuscitation. When the ambulance arrived, his wife and I had a pink patient with small pupils. At that time ambulances did not carry defibrillators. By the time we arrived at the hospital, which was only a hundred yards away, the patient was blue, his pupils were fixed and dilated, and he could not be revived. It is difficult to maintain effective and continuous external cardiac massage while loading and unloading an ambulance, and while the ambulance is in motion. An editorial in the MJA in 2003 bemoaned the fact that “cardiac arrest is more successfully treated in Chicago or Heathrow airport, on an American Airlines or Qantas jet, or in a Boston post office, than in the vestibules, corridors or general wards of Australia’s premier hospitals”.4 A review of 28 cardiac arrests occurring at the Melbourne Cricket Ground (MCG) revealed a quite extraordinary survival rate of 71%,5 compared with a 3% survival rate from out-of-hospital resuscitation reported, at about the same time, in metropolitan Melbourne as a whole.6 Each minute from the onset of ventricular fibrillation to the use of a defibrillator results in a 10% reduction in survival.7,8 Thus, there is every reason to encourage anything that can shorten the time between the onset of cardiac arrest and defibrillation.9 But there is a big difference between the population of the MCG and that of a general hospital ward. The former has been able to get to the MCG, whereas the latter may be unable to get to the bathroom. Patients in a general medical ward may have failing hearts, lungs, kidneys and brains; they may be dying of cancer, they may be failing to respond to treatment for severe infection, or they may be otherwise very unwell. Most MJA readers will be familiar with the following scene, which takes place regularly in hospital wards. The curtains are barely closed around the bed of an elderly woman; two people are taking it in turns to rhythmically compress her chest; three doctors are attacking her oedematous limbs with needles, unsuccessfully attempting to insert them into veins; a fourth is poking around her groin trying to cannulate her femoral vein; and a fifth has a laryngoscope in her throat. But on this occasion, I notice something rather unusual (though I have seen it before). As I ventilate her lungs through the endotracheal tube, her eyes are wide open, her pupils are small, she blinks, she seems to be looking straight at me. Someone remarks, “She has a dying heart”. Eventually, we allow the rest of her to follow. In the past, nurses used experience and common sense when deciding not to use CPR in most patients when they died. Now, they are expected to start CPR on anyone who collapses and does not have a “not-for-resuscitation” order. This order is supposed to be discussed with the patient. This exceptionally difficult task may fall to the most inexperienced doctor on the ward. The results of CPR in this ward population are likely to be poor, even if immediate defibrillation is available (which it is not10). If cardiac arrest in a general medical ward is to be treated, could management be limited to prompt defibrillation, oxygen by bag and mask, and little else? The management of out-of-hospital cardiac arrest seems particularly prone to controversy. During the 1990s, ambulances regularly arrived at emergency departments carrying patients on whom cardiac massage was being performed. Subsequent discussion with relatives revealed that many of these patients had not had a witnessed cardiac arrest. They had been found dead. It was treating these patients that made me wish that the technique of CPR had never been introduced. In the five years since I wrote those words (and while the notes for this piece were gathering dust), it has become accepted that “. . . survival for the victim of cardiac arrest not resuscitated by a determined trial of advanced cardiac life support at the scene is negligible and not improved by further emergency department efforts”.11 Ambulance officers now have authority to cease resuscitation at the scene when it has clearly failed. This has reduced the incidence of futile resuscitation being performed in ambulances, which then has to be continued for a respectable period of time in the emergency department. But it may not prevent futile resuscitation efforts being performed in patients’ living rooms. An 85-year-old woman may phone “triple 0” after finding her husband collapsed on the floor. She may be advised to commence CPR. Sometimes, mightn’t it be quite reasonable for her to disregard this advice, and, when the ambulance arrives, to ask the paramedics to let her husband remain undisturbed? One hundred and five paramedics, emergency nurses, and emergency physicians who regularly took part in CPR were asked at what point they would like CPR stopped if they were the patient.12 Ten per cent did not wish to have CPR started at all; and only 3% wished to complete a full CPR protocol based on standard American Heart Association guidelines. Does not this suggest that there might be something wrong with the way CPR is presently practised? I was explaining to an elderly woman that her dying brother would be unlikely to last the hour. I asked if she wished to be with him when he died. Her husband turned to her and said, “ No. You don’t want to be there when they put the paddles on.” The assumptions behind this remark startled me. Should cardiopulmonary resuscitation be a futile deathbed ritual — a secular last right? There are many ways to die. To die without fuss, here one minute gone the next — that is the best.

Michael J Mackay MB ChB, MHA, FACRRM

Envenoming is rare in Australia — multicentre studies are needed to improve the tenuous evidence base Whether it’s the live snake that escapes in an emergency department or the farmer, bitten by a brown snake, who drops into his wife’s work to say he will be in hospital, and then collapses and has a seizure on arriving in hospital — bites and stings are a fascinating topic and the occasional envenoming presenting to hospital makes the local news. Unfortunately, the rarity of envenoming in Australia has meant the evidence base in clinical toxinology is tenuous, with considerable reliance on case reports and anecdotes of successful treatment. Although case reports can be essential in providing information about rare effects, more importantly they help to develop hypotheses for further studies. Randomised controlled trials (RCTs) of antivenoms are difficult to undertake in clinical toxinology because of the rarity of envenoming, the rapid course of life-threatening effects, and the potential for complete reversal of effects with antivenom. Funnel-web spider envenoming and major box jellyfish envenoming by Chironex fleckeri are two contrasting examples of such situations, with very different outcomes over the past 30 years following the introduction of their respective antivenoms. Despite the absence of an RCT of funnel-web spider antivenom, most people would agree that its introduction has prevented death in some cases and changed the outcome in many more cases over the past 25 years. It is highly unlikely that any ethics committee would now sanction an RCT, except perhaps to conduct a short n-of-1 trial, randomising patients initially to antivenom or placebo, and providing rescue treatment after 1–2 hours.1 The initial prospective study of nine successfully treated patients,2 another study demonstrating a significant reduction in hospital length of stay,3 as well as the fact that there have been no fatal bites since the introduction of funnel-web spider antivenom, provide more than single-case or anecdotal evidence for its efficacy. In contrast, the introduction of C. fleckeri antivenom has been somewhat different, with reported deaths despite the administration of antivenom, and continuing controversy about its use intramuscularly in the prehospital setting and in treating non-life-threatening effects.4 Recent animal work suggests that pretreatment with antivenom is not completely effective in preventing cardiovascular collapse and adds to the concerns regarding the efficacy of this antivenom.5 Irukandji syndrome has come to the attention of most Australians over the past few years, with at least one confirmed death from Irukandji syndrome in far north Queensland,6 and reports of significant numbers of cases in northern Western Australia in this issue of the Journal ().7 This has attracted significant media attention, threatened tourism in Queensland, and prompted the rapid introduction of untested treatments.8 Unfortunately, this appears to have overshadowed the far more lethal C. fleckeri envenoming, which continues to claim lives, with recent deaths of young children in far north Queensland. Treatment for C. fleckeri envenoming remains controversial, with concerns about the efficacy of antivenom,4 disagreement over the role of pressure immobilisation bandaging9 and non-evidence-based ongoing support for the potentially dangerous adjunctive treatment with verapamil.10 Recent animal studies provide evidence that pressure bandaging in C. fleckeri envenoming may increase venom discharge,11 and a review of the literature found no evidence for the recommendation of pressure immobilisation in major jellyfish stings.9 A recent animal study investigating treatments for C. fleckeri envenoming demonstrated that pretreatment with antivenom only prevented cardiovascular collapse in 40% of rats.5 The addition of verapamil did not prevent any deaths, supporting previous studies showing that verapamil worsens outcome in C. fleckeri stings.12 Another finding was that the addition of intravenous magnesium sulfate to antivenom, as a pretreatment, prevented death in 100% of cases.5 Future studies will need to further evaluate antivenom and the possible benefits of magnesium. However, it must be emphasised that early resuscitation is likely to be the single most important measure in severe C. fleckeri envenoming. Back on land, Australia is extremely fortunate to have some of the safest and most efficacious snake antivenoms in the world and the only commercially available snake venom detection kits for patient management. Despite this, the management of snakebite continues to be dominated by anecdotes and case reports, with limited information on antivenom dosing and redose timing. In addition, many snakebites occur in rural or remote areas, necessitating use of retrieval services and telephone advice. In this issue of the Journal, Yeung et al () report a retrospective study of severe brown snake envenoming in Western Australia, suggesting that larger overall doses of antivenom are required.13 Although the authors have moved to using 10 ampoules as their initial dose, their study does not provide conclusive evidence to allow absolute recommendations for antivenom dosing, particularly in other parts of Australia. However, it reinforces the problems with severe brown snake envenoming in rural and remote areas, and the need for sufficient antivenom being available for a first dose (at least five ampoules) for patients with suspected snakebite being retrieved to larger centres. The study by Yeung et al13 also provides the impetus for prospective studies of snakebite to define the initial antivenom dose and the need for further doses. Such studies are only possible if there is serial estimation of venom concentrations in blood to determine the antivenom dose required to completely neutralise circulating venom.14 Because of the rarity of snake envenoming, a multicentre study is required. The study by Currie () demonstrates just how uncommon snakebite envenoming is in Australia.15 Despite enrolling patients at a hospital that has large numbers of snakebite presentations, the study period required was about 10 years. Multicentre studies are currently being conducted throughout Australia, with collaborative research between clinical toxinologists and emergency physicians in more than 30 hospitals. In addition to answering questions about antivenom dosing, these studies will prospectively evaluate the effectiveness of pressure bandaging with immobilisation. Many questions remain about the use of snake antivenom. The treatment of and premedication to prevent snake antivenom reactions is still of concern. There have been three RCTs,16,17 but because of problems with small numbers and methodology18 many questions remain. Such studies are difficult in Australia because of the infrequency of administration of antivenom in single centres. Again, we need either large multicentre studies or, alternatively, studies conducted in rural tropical countries such as Papua New Guinea or Sri Lanka, where snakebite envenoming is common and a major public health issue. Collaborative work between these countries and Australia will both improve the care of patients and contribute to our understanding of snake antivenoms.

Geoffrey K Isbister BSc FACEM MD

Objective: Fatal snakebites at Port Moresby General Hospital (PMGH), Papua New Guinea (PNG), were examined to identify interventions that may improve patient survival.Design: Retrospective case series.Subjects and setting: Inpatients at PMGH who presented with snakebite, had evidence of envenomation, and died as inpatients between 1 January 1992 and 31 December 2001.Outcome measures: Number and cause of fatalities; ventilation bed-days; antivenom timing, dose and price.Results: 87 deaths occurred among 722 snakebite admissions to the intensive care unit (ICU). Of these 722 patients, 82.5% were ventilated, representing 45% of all ventilated ICU patients and 60% (3430/5717) of all ICU ventilator bed-days. The median duration of ventilation in fatal snakebite cases was significantly less than in non-fatal cases for children (3.0 v. 4.5 days) and adults (3.0 v. 5.0 days). The case-fatality rate for children (14.6%) was significantly greater than that for adults (8.2%). Sixty fatalities were examined in detail: 75% received blood products; 53% received antivenom (mostly a single ampoule of polyvalent), but only 5% received antivenom ≤ 4 hours post-bite. Major causes of death included respiratory complications (50%), probable intracerebral haemorrhage (17%), and renal failure (10%). Antivenom unit costs increased significantly over the decade; in 2000 an ampoule of polyvalent antivenom was 40-fold more expensive in PNG than in Australia on a gross domestic product (A$) per capita basis.Conclusions: Management of severe snakebite is a major challenge for PMGH. Improved antivenom procurement and use policies (including increased use of appropriate monovalent antivenoms), combined with targeted snakebite education interventions (community- and hospital-based), are key interventions to reduce the ongoing toll from snakebite.

Forbes McGain MB BS, DipObs · Ken D Winkel MB BS, PhD, FACTM · Aaron Limbo MB BS, DipAnaesth · David J Williams BSc · Gertrude Didei MB BS, DipAnaesth

Objective: To describe the epidemiology of snakebite in the “Top End” of the Northern Territory, and the envenoming syndromes of individual snake species.Study design: Prospective collection of clinical data and snake identity.Setting: Royal Darwin Hospital (RDH), a 300-bed tertiary hospital servicing a population of 140 000 spread over 522 561 km2.Patients: All patients with bites by confirmed snake species between September 1989 and March 2003, and all suspected snakebite cases between September 1989 and March 1998.Outcome measures: Incidence rates of definite snakebite and envenoming. Clinical features of bites from defined snake species.Results: There were 348 suspected snakebites over 8.6 years, with 114 aerial evacuations to RDH, 216 patients (62%) definitely bitten (23.2/100 000 per year) and 79 (23%) envenomed (7.6/100 000 per year). There were 156 bites from confirmed species over 13.6 years: 31 (20%) from western brown snakes (Pseudonaja nuchalis), with early collapse in 14 (45%), consumptive coagulopathy in 26 (84%) and 25 (81%) given antivenom; 21 from death adders (Acanthophis spp.), with neurotoxicity in 8 (38%) and 6 (29%) given antivenom; and 20 from mulga snakes (Pseudechis australis), with local swelling in 19 (95%), myotoxicity in 12 (60%) and 15 (75%) given antivenom. In 34 bites from less venomous species, there was no life-threatening envenoming. There were no deaths.Conclusions: Snakebite still causes morbidity in tropical Australia, but, with access to hospital and antivenom, deaths are rare. This study has enabled further definition of the envenoming syndromes of three highly venomous Australasian elapids.

Bart J Currie FRACP, DTM+H

Objectives: (1) To assess the number and severity of episodes of Irukandji syndrome in Broome, Western Australia. (2) To correlate demographic, seasonal, geographic and climatic features of Irukandji stings. (3) To assess treatment of Irukandji syndrome at Broome Health Service. (4) To assess the public health impact.Design and setting: (1) A retrospective analysis of jellyfish data forms and charts of 111 patients, identified from Broome Health Service Emergency Department with a discharge diagnosis of marine sting between 1 January 2001 and 1 July 2003. (2) Correlation between climate and Irukandji envenomation data.Main outcome measures: Number of patients with Irukandji syndrome; their demographic and environmental features; the clinical syndrome; treatment requirements.Results: 111 patients were prospectively identified with marine stings; 88 were identified with Irukandji syndrome. Non-Irukandji syndrome data were excluded for analysis. The “jellyfish season” extends from January to May, although stings occur all year round. Only 38% of patients had vinegar applied to the sting site before hospital presentation. Signs and symptoms were variable between individuals, with 20% having no signs of sting at all and welts found in 16%. Fifty per cent of patients were hypertensive at presentation. Distress was found in the majority of patients, with 90% requiring opioid analgesia (morphine equivalent: mean, 20 mg; median, 13 mg) and 17% requiring admission. There was one evacuation to Perth with cardiotoxic marine envenomation resulting in pulmonary oedema, which necessitated 4 days in intensive care. Stings were significantly more common when the ambient median temperature was greater than 28.3°C, after midday, on an incoming high tide and on windy days.Conclusion: The rate of envenomation in northern WA is likely to be the highest currently documented in Australia. There is syndromic variability when compared with the north Queensland experience. This implies different causative jellyfish species that are not yet identified. Stings in Broome can be severe and life threatening; there are significant commercial and public health implications as a result. Management at Broome Hospital is contemporary and effective.

Conrad J Macrokanis MB BS, FRACGP, BSc(Hons) · Nicole L Hall · Jacki K Mein MB BS, FACCHP, MAE

Objective: To investigate the doses of antivenom administered to adult patients with severe brown snake envenoming.Design and setting: Review of charts from Western Australian adult teaching hospitals, December 1991 to December 2001.Patients: 35 patients with severe brown snake envenoming, defined prospectively as afibrinogenaemia (< 0.3 g/L) after a bite by a brown snake (genus Pseudonaja).Main outcome measure: The dose of antivenom required to neutralise venom, defined prospectively as the dose of antivenom given before the return of detectable fibrinogen levels.Results: Of 88 patients with brown snake envenoming admitted over the 10 years, at least 35 had severe envenoming. Afibrinogenaemia persisted for 10 hours (range, 1.4–68 hours) after the first dose of antivenom; in four patients afibrinogenaemia lasted more than 24 hours. The dose of antivenom given before venom neutralisation ranged from one to 23 ampoules. In two-thirds of cases, venom was neutralised with five ampoules, and 89% had venom neutralised with 10 ampoules. Two patients died, and another had serious bleeding complications. Another patient died during the study period from intracerebral haemorrhage, but did not have fibrinogen levels measured.Conclusions: Patients received initial doses of antivenom too small to neutralise circulating venom, and remained afibrinogenaemic for prolonged periods, with serious consequences. The authors now use 10 ampoules as an initial dose in severe brown snake envenoming.

Justin M Yeung MB BS, FACEM · Frank F S Daly MB BS, FACEM · Mark Little FACEM, MPHTM, DTMH · Lindsay M Murray MB BS, FACEM · George A Jelinek MD, FACEM, DipDHM

A novel tagging technique has uncovered some surprising information about jellyfish behaviour If you spend any time at all in tropical Australia, especially in the water, you will know about box jellyfish. You will also know that they have a major effect on the way people use the water, that they are capable of killing humans within minutes, and that vinegar is the first aid treatment of choice.1 But did you know that they “sleep”? We certainly didn’t! About 12 months ago, we came up with a novel idea to try to track box jellyfish (Chironex fleckeri) using small ultrasonic transmitters (about 4 cm long and 12 mm in diameter). Using these for tracking marine animals is not new, but tracking jellyfish with them — that’s certainly never been done before. Normally, when tracking marine organisms (such as fish), you open the body cavity, insert the transmitter, suture the wound and let the fish go. With jellyfish, it’s not that simple. Firstly, jellyfish don’t have a body cavity (they only have two cell layers, an ectoderm and an endoderm, with a non-cellular layer, the mesoglea, between these). Secondly, suturing jellyfish is not easy. In fact, it’s impossible! After many failed attempts to attach transmitters, we finally struck upon a simple but effective method. We glued them on using histoacryl, a superglue used by surgeons. All you need to do is catch a box jellyfish without getting stung (an art in itself!), glue a transmitter to it (Box 1), release it, and follow it with an underwater directional microphone. You can then work out where they go and how active they are. In the last jellyfish season, we managed to track several tagged box jellyfish (Box 2), and came up with some staggering results. It seems that these jellyfish show marked diurnal behaviour. During daylight hours (from about 0600 to 1500), they moved in straight-line distances of about 212 m an hour. However, from about 1500 to 0600, they moved an average of less than 10 m an hour.2 During these periods of “inactivity”, the jellyfish lie motionless on the sea floor, with no bell pulsation occurring and with tentacles completely relaxed and in contact with the sea floor (Box 3). Shining lights on the jellyfish while they are inactive on the sea floor, or causing vibrations close by on the seabed, causes the animals to rise from the sea floor, swim around for a short period, and then fall back into an inactive state on the sand. If you have any interest in animal biology, this type of action in a lower invertebrate should immediately raise the question of “Why?”. We believe it is related to the way the jellyfish collect food. The box jellyfish is an active visual hunter of vertebrates. It has four sets of six eyes, some of which are image-forming with lenses and retinas, lying around the four facets of the bell (body) of the animal.3 Box jellyfish are also extremely active, with metabolic rates at least an order of magnitude greater than those of any other jellyfish we know of.4 So, at night — when vision is limited and you cannot see your prey or your predators (turtles for box jellyfish) — rather than burn a lot of energy swimming around, it makes a lot of sense to become inactive, decrease your energy used in locomotion and divert it to growth (these animals can grow at 2–3 mm across the bell per day). A really simple strategy, but one we had not thought box jellyfish used. All we need now is for someone to design a tag small enough to put on Irukandji box jellyfish (Carukia barnesi) — which are about as big as your thumbnail — and then we will really make some progress! 1 Attaching a tag to the most venomous marine creature in the world — it’s easy when you have the right glue! 2 A large box jellyfish with its tag attached, off in search of food 3 A tagged box jellyfish lying motionless on the sea floor — the first time this sort of behaviour has been recorded in jellyfish

Jamie E Seymour BSc(Hons), PhD · Teresa J Carrette BSc, MSc · Paul A Sutherland MSc

Hypothermia is now standard care for some types of cardiac arrest Out-of-hospital cardiac arrest is a leading cause of unexpected death in the developed world, occurring in about 1 in 1500 adults each year.1 Successful recovery from out-of-hospital cardiac arrest depends on the rapid activation of the “chain of survival”: an immediate call to the ambulance service, bystander delivery of external cardiac massage and expired-air breathing, defibrillation and the provision of advanced life support by paramedics.2 Unfortunately, survival with good neurological outcome at hospital discharge is rare after out-of-hospital cardiac arrest. Studies in Perth and Melbourne show that less than 5% of these patients survive to hospital discharge.3,4 As the average response time of ambulances in most Australian cities is between 7 and 12 minutes, considerable neurological injury occurs during this prolonged period of cardiac arrest, even with bystander cardiopulmonary resuscitation. While paramedics may restore spontaneous circulation and transport some patients alive to an emergency department, most remain comatose because of the severe anoxic brain injury.4 To improve outcome, considerable emphasis has been placed on shortening the time between cardiac arrest and defibrillation. As decreasing ambulance response times towards 5 minutes would be prohibitively expensive, alternative approaches to earlier defibrillation have been proposed. These include fire-fighters co-responding with ambulance services to patients with suspected cardiac arrest,5 or installation of automatic defibrillators in public places.6 On the other hand, recent data from Canada have cast doubt on the effectiveness of paramedic advanced life-support programs, which did not improve survival rates when introduced.7 What therapies are available after arrival at the hospital? In most cases, no immediate cardiology intervention is required, and treatment has therefore been largely supportive until the neurological outcome could be determined. Common intensive care practice has been to defer neurological assessment for at least 3 days, to allow more accurate clinical assessment.8 Recently, an “old” therapy for anoxic brain injury — therapeutic hypothermia — has been re-introduced into clinical practice. In this issue of the Journal, Williamson and colleagues (page 500) describe the use of this therapy in a patient who was comatose after near-drowning.9 The use of mild therapeutic hypothermia after cardiac arrest was first described in the 1950s, but later abandoned without being formally tested in clinical trials.10 Interest in hypothermia was revived in the early 1990s when animal studies and preliminary clinical studies suggested benefit. Subsequently, two prospective, randomised, controlled clinical trials have been conducted.11,12 In a recent Australian trial, patients who remained comatose after resuscitation from out-of-hospital cardiac arrest were treated with either 12 hours of therapeutic hypothermia (33°C) or standard care.11 At hospital discharge, 49% of those treated with hypothermia were discharged home or to rehabilitation, compared with 24% of those treated with standard care. In a European study, 55% of patients treated with hypothermia (33°C for 24 hours) had a favourable outcome at 6 months, compared with 39% of those treated with standard care.12 Subsequently, the International Liaison Committee on Resuscitation (which includes the Australian Resuscitation Council) endorsed the use of therapeutic hypothermia for patients with anoxic brain injury after out-of-hospital cardiac arrest, particularly when the initial cardiac rhythm is ventricular fibrillation (Box).13 Therefore, this treatment should now be regarded as a standard of care for this condition. However, a number of issues require further consideration if therapeutic hypothermia is to be applied more widely. Firstly, uncertainty remains about the effectiveness of this therapy in patients with out-of-hospital cardiac arrest due to causes other than ventricular fibrillation. Patients with asystolic out-of-hospital cardiac arrest have a dismal prognosis,14 as do those with coma after near-drowning, hanging, or other causes of asphyxia. Clinical data on the effects of therapeutic hypothermia in these groups are lacking. The role that therapeutic hypothermia played in the recovery of the patient reported by Williamson and colleagues is uncertain. Secondly, a protocol needs to be established in the emergency department for the rapid induction of hypothermia in patients who are unconscious after out-of-hospital cardiac arrest. In most hospitals, this will require consensus to be reached between emergency physicians, intensive care physicians and cardiologists on the indications for the provision of this treatment. Finally, there are technical issues to be considered in the rapid induction of hypothermia. In previous studies, hypothermia was induced through surface cooling with ice packs and/or refrigerated air blankets.11,12 This approach is slow and logistically difficult in busy emergency departments. Other technologies for the rapid induction of hypothermia are therefore under investigation.10 Currently, we are exploring the use of a rapid intravenous infusion of large-volume (30 mL/kg), ice-cold crystalloid fluid to induce hypothermia. Preliminary data suggest that this is relatively simple, effective, inexpensive and not associated with pulmonary complications.15 As there is often a delay between resuscitation and emergency department initiation of hypothermia, cooling in the ambulance would be ideal. In a study supported by the National Heart Foundation, paramedics in Melbourne are now infusing large-volume (2000 mL), ice-cold crystalloid fluid, together with a muscle relaxant, immediately after out-of-hospital cardiac arrest to induce hypothermia as soon as possible after resuscitation. If this is confirmed as feasible, further studies are planned which will examine the use of therapeutic hypothermia after asystolic and asphyxial cardiac arrest. Recommendations on therapeutic hypothermia from the International Liaison Committee on Resuscitation In October 2002, the Advanced Life Support Task Force of the International Liaison Committee on Resuscitation recommended: Unconscious adult patients with spontaneous circulation after out-of-hospital cardiac arrest should be cooled to 32°C–34°C for 12–24 hours when the initial rhythm was ventricular fibrillation. Such cooling may also be beneficial for other rhythms or in-hospital cardiac arrest. Preliminary data from clinical trials of perinatal asphyxia indicate that induced hypothermia is feasible and safe, but data on long-term neurological morbidity are not yet available. Until additional paediatric data become available, clinicians should tailor therapy for individual patients based on their assessment of the risks and benefits of hypothermia.

Stephen A Bernard MD, FACEM, FJFICM

Clinical record A 46-year-old English tourist became distressed when caught in a strong rip while swimming at a Sydney beach. An off-duty lifesaver dragged him to shore, where he was found to be apnoeic but had a weak pulse. After a minute of mouth-to-mouth ventilation, his pulse was lost, and chest compressions were begun. Ten minutes after cardiac arrest an ambulance arrived, and a wide complex bradycardia with no pulse was documented. During resuscitation, a total of 3 mg adrenaline, 70 mmol sodium bicarbonate and 1.5 mg atropine was administered. At 26 minutes after the cardiac arrest, the patient regained a pulse and was noted to be in atrial fibrillation. By this point, he was intubated and ventilated. On arrival in the emergency department, his systolic blood pressure was 90 mmHg, pulse rate was 136 bpm and irregular, and core temperature was 34.7°C. Auscultation of the chest revealed coarse bilateral crackles. His pupils were fixed and dilated. Initial arterial blood gases showed a severe mixed respiratory and metabolic acidosis, with a temperature-corrected pH of 6.47 (reference range [RR], 7.35 to 7.45), Pco2 of 98.6 mmHg (RR, 32.0–48.0 mmHg), Po2 of 104 mmHg (RR, 83.0–108 mmHg), arterial lactate level of 29 mmol/L (RR, 0.5–1.6 mmol/L), base excess of –28.1 mmol/L (RR, –2.0 to 2.0 mmol/L), and HCO3 level of 7.1 mmol/L (RR, 22–32 mmol/L). Chest radiography showed changes consistent with aspiration pneumonitis. The ventilator was adjusted to hyperventilate and reverse the respiratory component of the acidosis. Computed tomography of the head and neck revealed no abnormalities. It was decided to use controlled hypothermia to limit further hypoxic brain injury. The patient was packed in ice, and cooled intravenous fluids were administered, aiming for a core temperature of 33°C for a period of 12 hours (Box 1). The patient’s condition improved rapidly. The heart reverted to sinus rhythm spontaneously within a few hours, with no subsequent electrocardiograph or biochemical evidence of myocardial injury. The metabolic and respiratory acidosis normalised over the 10 hours after injury (Box 2). After 12 hours of hypothermia, cooling was ceased, and his core temperature was allowed to rise to 37°C. He was extubated on Day 3 after admission and discharged from hospital on Day 7. Neuropsychological assessment performed 2 weeks after the injury (including Wechsler Memory Scale Three and Delis–Kaplan Executive Function System) revealed relatively intact cognitive function, although there was moderate impairment in new learning ability and capacity for visuospatial information, and slowed information processing. This may be consistent with hypoxic brain injury, but, alternatively, might have been present before the accident. Repeat assessment 5 months after the accident (in the United Kingdom) revealed normal psychometric performance, with persisting impairment of visuospatial processing and organisational abilities. The patient and his family reported he was functioning just as he was before the injury, performing domestic duties and caring for his elderly mother. The World Health Organization reported 409 272 drowning deaths worldwide in the year 2000, making this the second leading cause of unintentional injury death globally, after road traffic accidents.1 The burden of near-drowning morbidity has not been defined but is likely to be high. Acidosis, especially pH < 7.0, usually correlates with a poor outcome. However, in drowning, the prognosis is often less grim, and patient survival has been reported after pH levels as low as 6.29.2 In our patient, the acidosis had both respiratory and metabolic components. The respiratory component, caused by apnoea, was the most readily correctable with gentle hyperventilation to “blow off” excess CO2. This had a rapid effect on the pH. The metabolic component was most probably a result of both physiological and pathological processes — the extreme muscle exertion while struggling in the surf and end-organ hypoxia as his respiration and circulation failed. In intensive care, blood lactate levels > 10 mmol/L are associated with 95% mortality.3 Our patient’s arterial lactate level was 29 mmol/L. We believe one reason for his good outcome was that the acidosis had a different aetiology to that in most intensive-care patients, in whom sepsis, hypoxia and shock predominate. This case therefore highlights the importance of continuing resuscitation efforts despite severe acidosis and fixed dilated pupils in drowning victims, as ultimate neurological recovery is still possible. Hypothermia is not a novel therapy for near-drowning. For over four decades, favourable outcomes have been reported for victims of near-drowning in ice-cold water.4 The effects of hypothermia are thought to be mediated by a reduction in brain metabolic requirements, limiting hypoxic cellular damage. A 1986 trial in Canada used hypothermia in near-drowned children to reduce intracranial pressure and limit brain injury.5 However, the death rate in the hypothermic group was higher than in the normothermic group, with most deaths attributed to neutropenic sepsis. This complication is now thought to occur only in children.6 The study also used lower temperatures and a longer period of hypothermia than is now recommended. This may have contributed to the poorer outcome and subsequent lack of interest in hypothermia for near-drowning. Interestingly, survivors in the study were more likely to be neurologically intact if they had been assigned to the hypothermia group. This suggests that the benefit of hypothermia may be not in preventing death but in improving neurological outcome in survivors. Recent evidence for controlled hypothermia as a neuroprotective therapy has been published in a study assessing its benefits in treating out-of-hospital cardiac arrest.7 This study used moderate (33°C), short-term (12 hours) hypothermia. Based on the favourable outcome of this and similar studies, the 2002 World Congress on Drowning, in Amsterdam, recommended this therapy for near-drowning victims who have restoration of adequate spontaneous circulation and remain comatose.8 A recent ILCOR (International Liaison Committee on Resuscitation) advisory statement also suggested that hypothermia may be beneficial for other rhythms causing cardiac arrest in settings such as drowning.9 We suggest that, in near-drowning cases where the patient remains comatose yet has adequate circulation, controlled hypothermia is an important tool to maximise neurological recovery. Further, severe acidosis should not be seen as a poor prognostic marker and should not preclude ongoing resuscitation efforts. Lessons from practice Severe acidosis or fixed dilated pupils are not useful markers for prognosis in the near-drowned patient. Controlled hypothermia at 33°C for 12 hours should be used in near-drowned patients who have spontaneous circulation but remain comatose. Controlled hypothermia cannot be recommended at present for children because of the risk of neutropenic sepsis. 1 Patient’s temperature over time after near-drowning 2 Arterial pH, lactate level and Pco2 over time after near-drowning* * Shading indicates reference ranges.

Jonathan P Williamson MB BS, BSc(Med) · Stan Braude MD, FRACP · Rowland Illing MRCS · Paul Gertler BA(Hons), MPsych, MAPS

A 74-year-old woman presented to hospital with dyspnoea of sudden onset. She was in cardiogenic shock, with blood pressure of 90/50 mmHg, and pulse rate of 115 bpm. The electrocardiogram (ECG) was difficult to interpret because of severe, persistent and inexplicable artefacts (Box 1A). Chest x-ray showed acute pulmonary oedema thought secondary to left ventricular failure. Echocardiography showed a left ventricular ejection fraction of 40% and anterior akinesia. Coronary angiography confirmed a proximal occlusion of the left anterior descending coronary artery, which was successfully treated by direct angioplasty and stenting. Review of the patient’s previous medical records revealed, on a thoracoabdominal image derived from computed tomography data, a foreign body in the central abdominal region (Box 2). Questioning of the patient after the angioplasty revealed that this was an electrostimulator (Itrel II, Medtronic, Minneapolis, USA) implanted a few years earlier as part of a dynamic graciloplasty to treat faecal incontinence. An ECG performed 12 hours after the angioplasty with the device deactivated (by the patient’s control programmer) produced an ECG tracing free of artefacts (Box 1B). Persistent and inexplicable ECG artefacts should raise the suspicion of interference from a device generating high frequency electrical impulses. Such devices may include minute ventilation rate-responsive pacemakers1 and central or transcutaneous neurostimulators (eg, deep brain neurostimulators used in Parkinson’s disease,2 and spinal neurostimulators used in chronic back pain3). Occasionally, when the vector of impulses is perpendicular to an ECG lead, the tracing from that lead may be free of artefacts (eg,Box 1A, lead I). Intrathoracic devices are usually discovered on physical examination or standard chest x-ray, but devices in other parts of the body may not be obvious. In dynamic graciloplasty to treat anal incontinence, the gracilis muscle is transposed around the anal canal and electronically stimulated by a device placed in a subcutaneous pocket in the abdominal wall.4 In our case, this device led to ECG artefacts potentially interfering with the ability to diagnose the acute myocardial infarction. Implanted electrostimulation devices are being used increasingly in medicine and should be considered as a possible source of ECG artefacts, especially by frontline clinicians confronted daily with such life-threatening conditions as acute coronary syndromes. Electrocardiograms in a patient with an electrostimulation device A. On presentation, showing artefacts. B. After temporary deactivation of the device. 2 Thoracoabdominal image calculated from computed tomography data, showing the device (arrow)

Robert F Bonvini MD · Edoardo Camenzind MD

Clinical record A 36-year-old man was found semiconscious in a park. He was carrying a syringe which was later found to contain high-purity amphetamine. He became combative, and was sedated and intubated on arrival at a hospital emergency department. Preliminary examination revealed proptosis of the left eye, which was surrounded by swelling and erythema, and a small conjunctival laceration on the nasal aspect. Pupillary light reflexes were normal. Examination revealed track marks in both cubital fossae suggesting intravenous drug use, but no other abnormalities. Non-contrast computed tomography (CT) of the brain revealed two minute areas of high density in the left frontal lobe, interpreted as evidence of early infection. CT of the orbits revealed left-sided swelling of the subcutaneous soft tissue, interpreted as preseptal cellulitis. A fracture of the left orbital roof was also noted adjacent to a superior orbital phlegmon (Box 1A). A lumbar puncture was performed, and, with a provisional diagnosis of preseptal cellulitis and meningitis, the patient was empirically treated with broad-spectrum intravenous antibiotics (vancomycin, ceftriaxone, gentamicin and metronidazole), while remaining intubated in the intensive care unit. However, as the cerebrospinal fluid appeared normal on microscopy and biochemical examination and showed no growth on culture, the initial diagnosis was questioned. Magnetic resonance imaging of the brain and orbits was performed the day after presentation. This showed a 5 cm tract extending obliquely from the roof of the left orbit into the white matter of the left frontal lobe, associated with a fracture of the orbital plate of the frontal bone (Box 1B). Inflammatory changes, presumed infective, were noted superiorly within the left orbit, resulting in proptosis. It was concluded that the patient had incurred a penetrating injury of the left orbit entering through the nasal conjunctiva and extending superiorly to the globe, through the orbital plate of the left frontal bone, and into the frontal lobe. Treatment with vancomycin, ceftriaxone and metronidazole was continued, and the patient was successfully extubated after 5 days. He was then able to recount jumping over a fence and landing on bushes, a branch of which penetrated his orbit. Re-examination revealed a marked decrease in the soft-tissue swelling around the left globe, but significant limitation of movement of the left eye in all directions (Box 2). These findings, along with the known intracranial penetration and the possibility of retained organic fragments, necessitated surgical exploration. A superior orbitotomy revealed an abscess containing tiny fragments of organic matter. This was washed out. The abscess was sterile on culture. The orbital tract was explored, and further small fragments of organic matter were retrieved. A left frontal craniotomy and repair of the roof of the left orbit were performed concurrently. Over the ensuing 10 days, the patient’s ocular movements improved to near full range, but he had intermittent temperature spikes. Multiple blood cultures showed no growth. Transthoracic echocardiography revealed a vegetation on the tricuspid valve. Intravenous antibiotics were continued for a further 6 weeks to treat presumed infective endocarditis. Although the patient developed no further symptoms, magnetic resonance imaging of the brain 11 days after the initial surgery revealed an elongated abscess in the frontal lobe, its inferior end abutting a small fragment of intracranial bone. The left frontal craniotomy was reopened, and the abscess drained. Culture of the abscess again showed no growth, and the patient required no further surgical intervention. Transorbital intracranial penetration by a wooden foreign body is unusual.1 The resilience of the sclera and ability of the globe to be displaced usually protect the eye from perforation.2 Metallic objects and glass fragments are the foreign bodies most often encountered in the orbit.3 Although computed tomography is excellent for identifying these high-density objects, it is much less sensitive for low-density organic objects.4 Magnetic resonance imaging is more sensitive for delineating the extent of orbital injury and is safe when non-magnetic foreign bodies, such as wood, are suspected.5 Because of its porous organic nature and frequent proximity to soil, wood is an ideal reservoir for bacteria and fungi and is likely to provoke inflammation. A narrow deep tract, such as occurred in our patient, is conducive to the proliferation of anaerobic bacteria.6 The most usual complications of an orbital foreign body are proptosis of the eye, development of a chronic fistula, orbital abscess or cellulitis, and damage to the extraocular muscles or optic nerve.7 Intracranial extension of the foreign body is associated with a 48% incidence of brain abscess and a 25% mortality rate.1,6 This case illustrates the possibility that a seemingly trivial lesion, such as a conjunctival laceration, can be associated with severe lesions in the orbital region. This case demonstrates the need to suspect intracranial penetration in orbital injuries, as intra- and extracranial complications often lead to prolonged hospital admission and carry a significant risk of mortality . Lessons from practice The diagnosis of transorbital intracranial penetration of a foreign body requires a high index of suspicion, as it can present with trivial findings on examination. Intracranial penetration carries significant morbidity and often leads to local and systemic complications. Early magnetic resonance imaging is recommended when there is a possibility of transorbital intracranial penetration. 1 Imaging in a patient with an orbital fracture A. Computed tomography showed a fracture of the roof of the left orbit (F), adjacent to a superior orbital phlegmon (P). B. Subsequent magnetic resonance imaging showed a 5 cm oblique tract extending from the fracture of the orbital plate of the left frontal bone into the white matter of the left frontal lobe. 2 Limitation of movement of the left eye The patient is attempting to look to his right.

Dana Robaei MB BS(Hon), MPH · Glen T Fernando MB BS(Hon) · Charmaine MacDonald MB BS · Michael G Branley FRACO, FRACS

High-quality primary evidence from an Australian and New Zealand study provides a definitive answer The 1998 report of a meta-analysis by the Cochrane Injuries Group Albumin Reviewers presented an important public health issue and questioned the practices of many doctors in Australian intensive care units (ICUs).1 Using data from 24 studies involving 1419 patients, the meta-analysis reported that the administration of albumin-containing fluids to critically ill patients increased the absolute risk of death by 6%, suggesting one extra death for every 17 patients given albumin. The authors recommended that albumin should not be administered to critically ill patients outside the context of rigorously conducted, randomised trials. A subsequent meta-analysis did not resolve the issue of albumin’s safety in the critically ill.2,3 Due to its ready availability, albumin has been widely used in Australian ICUs. Even after the publication of the Cochrane review, half of all patients in surveyed ICUs in Australia received albumin during their ICU stay (R Bellomo, S Finfer, unpublished data). Extrapolating these results would mean that 50 000 patients received albumin in Australian ICUs each year and, if the Cochrane meta-analysis was correct, this would result in an additional 3000 deaths annually. Thus, the issue of albumin’s safety was of particular importance in Australia. The recent publication of the SAFE (Saline versus Albumin Fluid Evaluation) study in the New England Journal of Medicine4 not only brings certainty to the issue of albumin’s safety in a heterogeneous population of adult ICU patients, it also marks the coming of age of clinical research in Australian and New Zealand ICUs. The SAFE Study, a collaboration of the Australian and New Zealand Intensive Care Society Clinical Trials Group, the Australian Red Cross Blood Service, and The George Institute for International Health, was a double-blind, randomised controlled trial of albumin versus saline for fluid resuscitation involving 6997 patients. Conducted in 14 ICUs in Australia and two in New Zealand, it was funded by the National Health and Medical Research Council and the Health Research Council of New Zealand, and by direct grants from the Australian federal, state and territory governments, the two New Zealand hospitals and CSL Ltd. Internationally, the SAFE Study is the largest randomised controlled trial conducted in intensive care or transfusion medicine to date. An accompanying editorial acknowledged that the SAFE Study heralded a new era in critical care marked by the large, simple randomised trial.5 The design of the study reflected the SAFE collaborators’ desire to conduct a definitive trial that would answer an important clinical question and provide results that would be widely applicable in ICUs around the world. As a result, the trial sought to include as many adult patients admitted to participating ICUs as possible by using broad, simple inclusion criteria and few exclusion criteria. The main inclusion criterion was that the treating clinician believed that fluid administration was indicated for the treatment of intravascular volume depletion. The primary outcome measure was all-cause mortality 28 days after randomisation. The only broad patient groups excluded were those admitted to the ICU after liver transplantation, burns or cardiac surgery. The study enrolled 6997 patients in 69 weeks, an average recruitment rate of 101 patients per week. The rapid recruitment rate was made possible by the commitment of the clinical staff in the participating ICUs, the provision for delayed consent, and web-based randomisation and fluid distribution. The blinded study design was possible as both study fluids were manufactured by CSL Ltd and packaged in specially designed blinding materials6 before distribution by the Australian Red Cross Blood Service, New Zealand Blood Service and participating centres’ blood banks. What did the study find? The primary outcome was available for 99.1% of the 6997 patients randomised; of these, 726 assigned albumin (20.9%) and 729 assigned saline (21.1%) had died by Day 28. The relative risk of death for patients assigned albumin compared with patients assigned saline was 0.99 (95% CI, 0.91–1.09; P = 0.87). There were no differences in secondary outcomes, with patients assigned albumin and saline having similar incidences of new organ failure, similar durations of mechanical ventilation and renal replacement therapies, and similar ICU and hospital lengths of stay. Key findings of the study are shown in the Box. The study identified six predefined subgroups: patients with and without trauma, with and without severe sepsis, and with and without acute respiratory distress syndrome (ARDS). As a previous meta-analysis had suggested that trauma patients resuscitated with colloid solutions had higher mortality than those resuscitated with crystalloid solutions,7 trauma was a stratification variable in the study. Patients with severe sepsis and ARDS were identified at baseline to determine whether the increased capillary permeability to albumin seen in those conditions8,9 resulted in a treatment effect that differed from that seen in the study patients without those conditions. Within the predefined subgroups there was limited evidence of a treatment effect favouring saline in patients with trauma, and favouring albumin in patients with severe sepsis. The possibly detrimental effect of albumin in patients with trauma was limited to patients with evidence of traumatic brain injury, namely those patients admitted to the ICU as a result of trauma who had a documented unsedated Glasgow Coma Scale score less than 14 and evidence of brain injury on cerebral computed tomography. The investigators cautioned readers that such subgroup differences frequently occur by chance, and the accompanying editorial advised cautious interpretation of the subgroup findings.5 Thus, the study demonstrated that, in this heterogeneous population of adult ICU patients, albumin can be considered safe, without demonstrating any clear efficacy advantage over saline. The SAFE Study achieved its goal of providing a definitive answer to an important clinical question. The result is widely applicable in those ICUs around the world where purified albumin is available. In addition, the study has demonstrated that investigators in Australian and New Zealand ICUs are capable of conducting large-scale collaborative trials on modest budgets and in a realistic timeframe. The SAFE Study has been described as a landmark study that heralds a new era in critical care.5 We hope that it will be only the first of many such studies. Key findings of the SAFE Study4 6997 patients were randomised to receive either albumin (3497) or saline (3500). The primary outcome (alive or dead at 28 days) was available for 6933 patients (99.1%). No significant difference was seen between the albumin and saline groups in: 28-day all-cause mortality (20.9% v 21.1%; P = 0.87) days in the intensive care unit (6.5 [SD, 6.6] v 6.2 [SD, 6.2]; P = 0.44) days in hospital (15.3 [SD, 9.6] v 15.6 [SD, 9.6]; P = 0.30) days of mechanical ventilation (4.5 [SD, 6.1] v 4.3 [SD, 5.7]; P = 0.74) days of renal replacement therapy (0.5 [SD, 2.3] v 0.4 [SD, 2.0; P = 0.41)

Simon R Finfer FRCA, FRCP, FJFICM · Neil W Boyce FRACP, PhD · Robyn N Norton PhD, MPH

Michael J Sinnott Emergency Physician, Princess Alexandra Hospital, Ipswich Road, Woolloongabba, QLD 4102 michael_sinnottAThealth.qld.gov.au To the Editor: In physiology, the Frank–Starling curve demonstrates that cardiac muscle initially responds to an increased workload with an increased force of contraction.1 However, after the point of maximum efficiency is reached, further workload produces a decrease in both the force of contraction and the ejection fraction, leading to cardiac failure. It now appears that the same curve could describe the current situation in many Australian emergency departments. We used to operate at point A on the curve (Box). If there was a mini-disaster or a moderately large number of victims of a road accident, the department was able to increase output to cope with the situation. The “adrenalin stimulation” experienced by all members of the team meant that the department coped, and that staff were left with a sense of satisfaction. Now our department finds itself at point B on the curve. Extra workload can result in a decrease in performance and output. The patients obviously suffer, but so do the staff. The once-challenging and enjoyable parts of the job now generate frustration and exacerbate the background dysfunction. On a recent weekend, the emergency department experienced an influx of sick elderly patients as a result of a local heatwave, with temperatures reaching 42°C. The problem was identified as a mini-disaster only in retrospect. At the time it was thought to only exemplify another bad day. This is an example of the syndrome of “learned helplessness”2 that staff are experiencing. Politicians and health administrators need to understand that our public hospital emergency departments are struggling with their daily workloads and are no longer equipped to deal with medium- to large-scale emergencies. Access block as a medical model At point A, an increase in workload leads to increased performance to cope (moving to point A1). At point B, an increase in workload leads to a decrease in performance (to point B1).

Michael J Sinnott

On 18 January 2003, Canberra experienced major bushfires. Over 6 hours, The Canberra Hospital Emergency Department treated 139 patients, 105 with fire-related problems (mostly ophthalmological and respiratory), representing an additional workload of one patient every 4 minutes above average. Only 15% required hospital admission. We believe this is the largest single emergency department response to a disaster since Cyclone Tracy devastated Darwin in 1974, although the total severity of injury was relatively low. Major issues were communication difficulties and transport, with most patients (including the two most critically ill) arriving by private vehicle. Overall, medical outcomes were excellent, and the hospital system coped well. On Saturday, 18 January 2003, bushfires hit Canberra, the national capital of Australia. Driven by hot westerly winds, wildfire swept through most of the urban bushland in the southern suburbs, crossed the bush–urban interface, and penetrated deeply into the westernmost urban area, destroying 501 homes and causing four deaths (Box 1). We describe the response at The Canberra Hospital, which is the regional major trauma service and the community hospital for the southern suburbs of the city. 1 Canberra, 18 January 2003 View south-west from Red Hill, showing fires burning on Mount Taylor, within the southern suburbs of Canberra. Woden Town Centre is in the centre of the skyline, with The Canberra Hospital on the left. (Photo courtesy of Brian Gunning, published at http://canberrafires.xsnet.org.) Canberra, in the Australian Capital Territory, is a planned city with a population of about 300 000. It is roughly divided into northern and southern sections by Lake Burley Griffin and features large tracts of bushland within urban areas. The Canberra Hospital is one of two public hospitals in the city. It is a modern, tertiary hospital of 483 beds, of which 362 were nominally available, and an Emergency Department (ED) which treats 52 000 adult and paediatric patients annually. The hospital has a regularly reviewed external disaster plan, which had been tested in a tabletop exercise 2 months before the bushfires. That exercise concerned a bombing causing multiple burns, but, like previous exercises, involved multiple casualties from one site rather than the effects of a geographically extensive disaster. The fires caused more property damage than any single recorded Australian fire except that at Mount Macedon in Victoria on “Ash Wednesday” (16 February 1983), when 628 homes and seven lives were lost.1 However, there were multiple fires at multiple sites on Ash Wednesday, as well as in Tasmania on 7 February 1967, which caused much more damage overall than the Canberra fires. ACT police took an aggressive approach to evacuation, which contributed to an extremely low ratio of deaths to property damage (Box 2). The Canberra Hospital response On the basis of available information, the ED staff specialist activated the “Standby Code Brown” (external disaster) procedure at 14:30 on 18 January. The Chief Minister of the ACT declared a state of emergency at 14:45.2 The first patient with a major fire-related presentation (exacerbation of chronic lung disease) arrived at 15:35, and the first patient with burns at 16:30. At 17:00, “Code Brown” was activated in response to three events in rapid succession: unexpected radio communication from a helicopter about to land with three injured fire fighters, loss of all computing facilities due to network outage, and loss of the power supply to the medical imaging department. Disaster mode was maintained until 09:00 the next morning. PatientsThe patients treated at The Canberra Hospital represented all the cases of serious injury and most cases of significant injury from the fires. Minor injuries were treated by general practitioners in private surgeries and evacuation centres, and almost no fire-related cases were seen at Calvary Hospital, the public hospital on the north side of Canberra, during the first day. Our description of patients is based on data from the Emergency Department Information System (EDIS), medical records, and the debriefing process. As part of the response, clerical staff identified disaster-related cases on arrival; these were later audited by the authors and other ED staff. Usually, the relationship to the fires was obvious, but, in doubtful cases, the definition was a clinical decision that the patient would not have presented had the fires not occurred. Over the 6 hours from 16:00 to 22:00, the ED treated 139 patients, 105 of whom had fire-related problems (Box 3). This represents 90 more patients than average, or one additional patient every 4 minutes. As in similar Australian fires, respiratory and ophthalmological problems were more prevalent than burns3,4 (Box 4). Only two patients required transport to interstate burns centres, and there were no deaths in hospital. Unlike many disasters, a significant feature was the continuation of “background” ED workload (Box 5). Even during the critical 6 hours, there were 34 non-fire-related presentations, and the expected fall in other workload did not occur until the following day. There was little opportunity to divert patients elsewhere, as The Canberra Hospital is the major tertiary hospital in the region and offers the only paediatric, acute orthopaedic, neurosurgical and major trauma services. Twenty-one people presented requesting supplies of their usual medications, as their homes had been destroyed or were inaccessible. These requests were concentrated in the first 16 hours. Medications commonly required were insulin, antipsychotics, antihypertensives and home oxygen. This was also reported as a major problem by a general practitioner (GP) working in one of the evacuation centres. The hospital’s GP liaison was contacted about this problem on the morning after the fires, after which GPs visited the evacuation centres and provided prescription and other services. StaffingThe fires occurred during school holidays and the hospital’s medical changeover period, when junior staff move to new positions, often interstate. Many regular staff were out of the city, and many of the previous year’s staff had already left. During the disaster, the ED medical workforce was boosted by staff working overtime and by 14 staff not rostered to the ED that day, including four from outside the hospital. Some of the staff contacted were unable to attend because of fire in or near their homes, and some responded without returning home, working while unsure whether their properties had survived. A major contribution was also made by a small number of relevant hospital specialists (plastic surgery, general surgery, respiratory medicine) and outsiders, including an emergency physician who travelled from Sydney. Issues and lessonsThe major issues that became apparent from the disaster were transport and communication difficulties, both of which are well recognised. Use of private transport: Sixty percent of disaster patients and 30% of those who required hospital admission arrived by private vehicle (including the two most critically ill). Although this made little difference to the hospital response on this occasion, it might be a major problem when multiple hospitals are involved, and pre-hospital personnel have the task of balancing workload between institutions. The experience was similar in the Granville rail disaster (1977)5 and Cyclone Tracy,6 and had been specifically predicted as a problem in the most recent Canberra Hospital disaster exercise. The high use of private transport during this disaster is an important lesson for the national capital, where exposure to chemical, biological or radiological weapons is considered a real threat. As there appears to be no realistic prospect of containing an exposed population, services must plan around patients presenting by private transport and requiring decontamination at hospitals.7 Communication difficulties: Communication was hampered by disruption of the mobile phone network, and also by the lack of a television, radio or adequate mobile phone coverage in the disaster control room, which was therefore moved to a more appropriate location. It should perhaps be noted that mobile phones were not a significant means of communication when the disaster plan was last activated (the landslide at the Thredbo ski resort, 19978). Communication lines with the ACT Emergency Control Centre remained open, but little or no information relevant to the hospital response was available, as the Emergency Services Bureau experienced significant command and control problems.2 In addition, it was found that the key hospital staff were unfamiliar with the two-way radios provided as part of the disaster plan, and that there was no STD telephone in the room designated for relatives, both of which increased the load on the hospital switchboard. Electricity supply: Fluctuations in the electricity supply caused by the fires led to frequent power outages. In retrospect, the hospital should have continued to use its emergency generators until the disaster was over. Operating theatre standby for “major cases”: All non-urgent surgery was cancelled, and operating theatres set up to receive “major cases”. This proved unnecessary, as no such cases presented, and also significantly delayed surgery, particularly acute orthopaedic surgery, increasing hospital bed use. Emergency department triage officer: Within the ED, it was found that experienced triage nurses provided a more appropriate triage service than the ED registrars designated in the disaster plan. While senior medical input may be required (for example, when a decision is needed about withholding care because of “expectant death”), triage and other tasks should generally be allocated to the most experienced staff. External liaison: A need for prescription and other GP services at evacuation centres became apparent. GP after-hours services should be opened early, and GPs should be asked to attend evacuation centres early in the course of a disaster. Photographic documentation: A less important failure was the lack of any photographic documentation of the disaster response for historical and teaching purposes. Although volunteers, such as medical students, were available, none were allocated to “camera duty”. Hospital evacuation: Fire came to within 2 km of the hospital campus (Box 1), but little thought was given to the possibility that it might reach the hospital. Although unlikely, this is a potentially catastrophic event which should be included in future planning. In light of The Canberra Hospital experience, detailed plans were drawn up for the evacuation of Calvary Hospital when fire threatened the northern suburbs of Canberra a few days later. ConclusionsBy definition, a disaster is an event that overwhelms available resources. Many internal and external hospital disasters have been described, with different workloads falling on a range of hospital services. In this case, the major burden clearly fell on the ED. The overall ED workload on 18 January was 252 presentations, almost 7 standard deviations above the daily mean of 137 for the period 1998–2002. However, the number of admissions (52) was less than 3 standard deviations above the mean for this period (34), and, measured by admissions, this was only the 14th-busiest day in the past 5 years. To our knowledge, this is the highest number of patient presentations from a single event treated in one ED since Cyclone Tracy devastated Darwin in 1974.3,5,6,9 It represents one of the largest responses to a disaster ever by a single Australian hospital, although injuries were less severe overall than in many other disasters. Overall, The Canberra Hospital coped well, despite significant disruption of staffing and infrastructure. Disaster conditions exposed some weaknesses in previous planning, particularly patient transport and the possibility of hospital evacuation, but the medical outcomes of the response were excellent. Most of the problems which occurred involved interdepartmental or interagency issues. 2 Significant single urban or rural fires in Australia in the past 40 years1 Date Location Deaths Homes lost Ratio 16 Feb 1983 Mt Osmond, SA 9 100 1:11 16 Feb 1983 Beaconsfield, VIC 21 238 1:11 7 Feb 1967 Hobart, TAS 20 310 1:16 14 Jan 1964 Dandenong, VIC 8 454 1:57 16 Feb 1983 Mt Macedon, VIC 7 628 1:90 8 Jan 1994 Como, NSW 1 101 1:101 18 Jan 2003 Canberra, ACT 4 501 1:133 3 Hourly workload during the disaster * Average weekend workload during the fourth quarter of 2002. 4 Major fire-related presentations at The Canberra Hospital, 18–19 January 2003 Condition Presentations Admissions Breathing problems/smoke inhalation 65 10 Eye problems (irritation, ulcer, foreign body) 43 0 Trauma (falls and motor vehicle) 45 6 Burns 24 10 Medication issues* 21 0 Accommodation and chronic disease† 5 5 Other 30 5 Total 233 36 *Supply of usual medications required by people unable to return home. † Person with chronic disease requiring emergency accommodation. 5 Daily workload for the fortnight in January 2003 that included the bushfires

Drew B Richardson FACEM · Sashi Kumar FACEM