Letters

To the Editor: The study by Robson and colleagues1 predictably produced alarming media headlines. However, we feel the study contains potential inaccuracies that seriously undermine the conclusion that the baby toll is lower in private hospitals. Did the authors ascertain the true impact of fetal abnormality or conditions requiring transfer to tertiary care? Women with threatened preterm labour often transfer to the public system and subsequently remain under the care of public high-risk teams. Babies with a known congenital anomaly detected antenatally by private obstetricians may be similarly transferred. This highest level of care is only available in public hospitals, usually those with dedicated subspecialist services in maternal–fetal medicine, neonatology or neonatal surgery. When babies who have been transferred from private hospitals to the public neonatal intensive care system subsequently die within 28 days of birth, this may be recorded as a public hospital death. The authors stated that the “major adverse outcome associated with social disadvantage is low birthweight” but that assessment of the effect of social disadvantage was beyond the scope of their study. Not accounting for the effect of low birthweight is unacceptable. The National Perinatal Data Collection records birthweight accurately and is easily used to remove the bias of low birthweight. An internationally accepted indicator of optimal perineal outcome is the rate of intact perineum. This is known to be lower in private hospitals but was curiously not reported. Further, a prospective data audit in a tertiary hospital revealed that where episiotomy extends to third- or fourth-degree tear, this is only reported as episiotomy.2 Thus, with twice the rate of episiotomy in private hospitals,3 and its association with severe perineal trauma, this outcome could be seriously under-reported in Robson et al’s study. The authors identify the significantly higher rate of caesarean sections in private hospitals as one of the measurable benefits of this model of care and dismiss any concerns as “orthodoxy”. They have overlooked the significant morbidity and mortality among term neonates associated with the rise in caesarean rates, both elective and following labour. Research using detailed Australian and New Zealand Neonatal Network data to ascertain admission rates of babies of ‘‘low-risk’’ women to neonatal intensive care units and special care nurseries not only found higher rates of admission across each gestational age group after elective caesarean section but also reported the death of two infants.4 Such morbidity and mortality is arguably the most important differential of care experienced by women who give birth in private or public hospitals.

Sally K Tracy · Alec W Welsh · Hannah G Dahlen · Mark B Tracy

To the Editor: More than 30 years ago, I was employed by the Commonwealth Government’s Maternal Mortality Committee to identify and evaluate factors contributing to maternal and infant mortality among Aboriginal Queenslanders. At that time, government policy was to transfer all pregnant Aboriginal women from their rural communities or missions to Cairns Base Hospital at 32 weeks’ gestation until 7–10 days after birth. Findings I presented in a report to Queensland Health in 19771 and at the Australian College of Paediatrics Annual Meeting in 19792 included: Babies of women who were compulsorily transferred at 32 weeks’ gestation to Cairns Hospital had lower neonatal death rates. It was assumed that — as Aboriginal women had unreliable gestational age assessments, antenatal care was irregular, and birthweights were lower than for other races3 — the risk of premature deliveries could be avoided by early transfer. There was a lower rate of breastfeeding among mothers transferred to Cairns, largely because if they opted to bottle-feed they could return home after 3–4 days (rather than waiting 7–10 days in Cairns to ensure breastfeeding was established).1 Growth failure was common in the month after weaning.4 Bottle-fed babies experienced slower weight gain and higher rates of illness.5 Suboptimal growth in bottle-fed babies during the early postnatal months predisposed babies to poor growth patterns during infancy6 and increased death rates.7 Women returning to their communities took with them infections acquired during the hospital stay in Cairns. Separation anxiety or maternal deprivation was common among children left in their communities. There was decreased family bonding (eg, between the father and the new child) and sibling resentment at maternal separation. There was unquantifiable resentment at having babies born away from the ancestral lands. Women delayed admitting their pregnancies in an attempt to remain in their communities, resulting in fewer antenatal visits. The antenatal transfer policy has been in effect for 30 years, despite conflict between those who were predominantly concerned with maternal and perinatal mortality (who favoured compulsory delivery in Cairns Base Hospital), and those (myself included) who were concerned about the effects on children’s growth and development. During this time, I have observed weakening family and community bonds, increasing alcohol and substance misuse and sexual abuse, low school attendance, poor employment records and domestic violence. My studies of Aboriginal and non-Aboriginal children born in Cunnamulla in western Queensland followed the same cohort of children from birth8 for 20 years.9 The presence of a father who was employed at the time of the child’s birth acted as a role model for the future and was more effective than all other social interventions with respect to the child’s successful education or employment 20 years later, irrespective of race or subsequent social support or interventions offered to the child during school years.9,10 Arnold and colleagues’ recent article in the Journal records a situation almost unchanged from 30 years ago, with the exception that antenatal ultrasounds have allowed transfer to occur at 36 weeks’ instead of 32 weeks’ gestation.11 Enormous resources of goodwill, effort and money have been spent in these communities over 30 years, yet the family disruption, unemployment and abuse statistics remain at variance with the Queensland norm. Perhaps the time has come to allow low-risk births to occur in selected towns, where the mother can be surrounded by her friends and relations, and be in closer contact with her ancestral land. I endorse the article by Arnold et al and hope they are more successful in implementing change than I have been.

John W Cox

To the Editor: A letter by Scrimgeour1 in the 18 May 2009 issue of the Journal praised an earlier study by Andreasyan and Hoy2 for adding to the evidence that Indigenous people living in small communities (known as “outstations” or “homelands”) in very remote areas are healthier than those living in larger settlements. I believe this was an error. The study found that mortality was lower in both outer regional areas and very remote areas than in remote areas.2 But, unfortunately, this tells us nothing about outstations. The Accessibility/Remoteness Index of Australia (ARIA) classification used by the authors groups larger remote communities and their outstations together — categorising both as “very remote”. Scrimgeour is not alone in overstating the evidence for better health among Indigenous people living in smaller communities. The possible health benefits associated with living in outstations has become a major argument against the Australian Government’s plan to create 26 remote “hubs” that will receive improved services (to the likely detriment of smaller communities)3 and the related plan of the Northern Territory Government to develop 20 larger remote communities.4 For example, Tom Calma, Aboriginal and Torres Strait Islander Commissioner at the Australian Human Rights Commission, was reported as saying that there was “a strong body of research showing that people on homelands lived longer, healthier lives”.5 Calma seriously overstated the evidence we have to date, which consists of two studies from Utopia in central Australia and one study of land management practices in an Arnhem Land community.6-8 Although this research is encouraging, we are a long way from determining the causes of better health among sections of those two communities, and whether residence at outstations/homelands or land management practices are associated with better health in other communities as well. There are many reasons for supporting the growth of outstations, and health may be one of them, but it is crucial that those in the field of Indigenous health get this right. Rather than jumping the gun, we need more research into the effects of outstation life on health and wellbeing.

Emma E Kowal

To the Editor: In an article about Indigenous mortality in the Northern Territory, Andreasyan and Hoy1 concluded that Indigenous residents in very remote areas (VRAs) had a better health status than those in remote areas (RAs) and outer regional areas (ORAs). This result is inconsistent with previous reports and prompted us to examine the authors’ outcome. A central problem with the authors’ analysis lies in the identification of “usual residence” for death registration. The usual residence is defined by the Australian Bureau of Statistics as the dwelling at which a person spends or intends to spend 6 months or more in the year in which the question is asked.2 This definition limits its usefulness for the authors’ purpose, but is further compromised by the common practice by certifying doctors of simply using the last known address as a proxy for usual residence. In either case, the address recorded at death registration may differ from the location where a person lived for the majority of his or her life. The latter is the location more closely associated with health risks, particularly for chronic disease. As a test, we investigated changes of residence for all NT public hospital inpatients who died within a 7-year period by linking multiple hospitalisations between 1 January 2001 and 31 December 2007. We found that 26% of inpatients classified as residents of RAs at the time of death were previously usual residents of VRAs. This “unhealthy migrant” effect can be readily recognised as people relocate due to illness from VRAs, which have limited health services, to RAs or ORAs to access secondary and tertiary health care. The likelihood of inconsistent classification of usual residence in mortality data can also be tested demographically. Assuming the authors’ mortality ratios were correct for a stable population, we estimate that the Indigenous life expectancy at birth in VRAs would be 72.3 years, or 23 years longer than the life expectancy at birth in RAs (49.1 years). Such a large discrepancy is implausible. The age structure of a stable population is determined by fertility and mortality,3 and reported NT Indigenous fertility rates show a lack of substantial variation across regions.4 If the life expectancy at birth in VRAs was significantly longer than the life expectancy in RAs, VRAs would have about five times more elderly people (aged over 75 years) than the current estimates.5 The mobility of residence shown by hospital data and the absence of a substantial elderly population in VRAs suggest that the reported differential mortality rates between VRAs and RAs are the result of misidentification of “usual residence”.

Yuejen Zhao · Steve Guthridge · Shu Q Li · Christine Connors

In reply: We thank Zhao and colleagues for their interest in our study.1 The main issue they raise is misclassification of deaths by remoteness of residence. In our article, we acknowledged the role of migration to larger urban centres to access health services and raised the issue of unhealthy lifestyle and its flow-on effects on mortality. If we were to assume that 26% of deaths in hospital in remote areas were deaths of people who previously lived in very remote areas (as Zhao and colleagues suggest), regional variation in the disparity between Indigenous and total Australian all-cause mortality would narrow but would still remain. A re-analysis of our data based on this assumption1 shows that standardised mortality ratios in remote areas would drop from 875% (95% CI, 799%–956%) to 646% (95% CI, 582%–717%) in 1998–2000 and from 731% (95% CI, 665%–801%) to 540% (95% CI, 484%–601%) in 2001–2003. In very remote areas, the standardised mortality ratios would increase from 214% (95% CI, 193%–236%) to 281% (95% CI, 258%–306%) in 1998–2000 and from 208% (95% CI, 189%–228%) to 264% (95% CI, 242%–286%) in 2001–2003. Furthermore, in our article we stressed the importance of examining Indigenous migration to enable more accurate interpretation of our findings and called for future studies to “clarify the reasons for these differences in mortality by remoteness, with a particular focus on migration”. Our studies of Queensland data and national data (unpublished) have shown a similar pattern of higher Indigenous mortality in remote areas than in very remote areas, but the difference in mortality rates between these areas is highest in the Northern Territory. Thus, the phenomenon of the “methodological” error applies to all of these studies. We propose to undertake a prospective study of Indigenous mortality that will overcome some study design problems inherent in a cross-sectional study.

Karen Andreasyan · Wendy E Hoy

To the Editor: We commend the Indigenous Health issue of the Journal (18 May 2009). Its editorial emphasis — that partnerships with and continued leadership by Aboriginal and Torres Strait Islander peoples will be key to closing the gap between Indigenous and non-Indigenous Australians1 — is supported by preliminary, unpublished findings from our research into improving mainstream general practice care of Indigenous patients. Couzos and Thiele emphasised that closing the gap in health and life expectancy between Indigenous and non-Indigenous Australians depends on Aboriginal community controlled health services (ACCHSs).2 By virtue of their governance structure and focus, these services deliver culturally safe and appropriate primary health care to Indigenous Australians, while addressing issues of Indigenous autonomy and other social determinants of health.2 However, although not all general practices see Indigenous patients, 0.9% of general practice encounters (range, 0.7%–1.6%) are with Indigenous patients, equating to about one million consultations a year. Indigenous Australians present to general practitioners with essentially the same range of clinical conditions as do non-Indigenous Australians, although consultation rates for diabetes and circulatory conditions are higher for Indigenous patients.3 Most Indigenous Australians (76%) live in urban and regional areas, and are widely spread through the general population. They are likely to need mainstream services including general practice and primary care services, at least some of the time and for the foreseeable future. There is clearly a need for initiatives beyond support for ACCHSs, Indigenous workers and communities, to improve mainstream services for Indigenous Australians in a culturally sensitive and appropriate manner.4 Better sociocultural education for health care providers, trainees and students is required to close the gap that exists in mainstream understanding and acceptance of Indigenous cultures and aspirations. The Inala Indigenous Health Service in Brisbane is an example of a mainstream practice successfully developing into an accessible service.5 Our research aims to improve mainstream general practice care for Indigenous Australians with diabetes who live in urban areas of Victoria. As the following statement made by Indigenous participants in focus groups for our research indicated, health services must be patient-centred: We need to hold the health system — and that includes GPs — accountable for delivering help to Aboriginal people. The best one to do that is the consumer. Focus group participants also highlighted the importance of working together: I think we need to take some responsibility ourselves as (Indigenous) workers in organisations and go to these mainstream services. Whether it’s diabetes or drugs and alcohol, we need to say look, you get funded to look after everybody, and we want to come here and tell you how to look after our people when they come to your centres.

Siaw-Teng Liaw

In reply: It is unclear how many mainstream general practice consultations involve Aboriginal and Torres Strait Islander patients. Of 485 300 patient encounters attributed to general practices in the BEACH study (2003–2008), 7292 were with Indigenous patients, but 2906 of these encounters took place in Aboriginal community controlled health services (ACCHSs) rather than in general practices.1 If funding to close the gap in Aboriginal disadvantage is being channelled to general practice and the Divisions of General Practice (through recent measures of the Council of Australian Governments), then the Indigenous health outcomes of mainstream services must be carefully attributed.2 The ACCHS sector agrees there is a need to make general practices culturally secure for Indigenous Australians. For example, the Aboriginal Health Council of Western Australia has developed modules for cultural safety training, which are accredited by the Royal Australian College of General Practitioners for GPs’ professional development.3 The National Health and Hospitals Reform Commission report recommends that health services be required, by accreditation processes, to meet specific standards of cultural safety for Indigenous patients to ensure high-quality care.4 Within a quality assurance framework, systematic reforms such as these can potentially improve mainstream services, while supporting Indigenous workers who face unrealistic pressures to improve their local services. As Liaw points out, these initiatives are in addition to the vital role of ACCHSs in closing the health disparity gap.

Sophie Couzos · Dea D Thiele

To the Editor: Finally! An article in the Journal about health reform that carries no spin and decries further futile experimentation in management. Lewis and Leeder are to be congratulated on their insightful review.1 As they point out, social and environmental determinants of health are far more important to a community’s well-being. This was recognised intuitively as long ago as 1986, with the promulgation of the Ottawa charter,2 and its recommendation that health services be reoriented primarily toward promoting heath rather than delivering acute care per se. As Lewis and Leeder articulate, the evidence in favour of this line of thinking is actually quite potent.1 However, as they point out, the thing that surprises students and practitioners in public health is why the “bigger picture” has been so difficult to comprehend and act on. Rather than a need for the “stars to be aligned”, perhaps this is the time for stronger advocacy? And while we’re at it, evaluation tools for health promotion programs need to become more sophisticated if we’re going to convince our colleagues, politicians and the community that we’re spending their money wisely. The transformation that Lewis and Leeder recommend is likely generational in its span, but what better time to make a start?

George Larcos

To the Editor: A previous report from our institutions identified a steady increase in the prevalence of paediatric treadmill friction burn injuries, from three in 2001 to 17 in 2006.1 We sought to determine whether there was any change in this trend during the past 2 years. Children younger than 16 years with treadmill-related injuries were identified from prospectively collected data from burns and trauma databases maintained by the trauma research nurses at two paediatric tertiary trauma centres in Sydney (the Children’s Hospital at Westmead and Sydney Children’s Hospital) between January 2007 and December 2008. Sixty-five children sustained treadmill-related injuries (17 in 2007 and 48 in 2008); 43 were boys. The mean age at the time of the injury was 3.7 years (range, 9 months to 14 years). Friction burns ranged from less than 1% to 7% of total body surface area, and most patients sustained a total body surface area burn of 1% or less (58 patients). The most common site of injury was fingers and/or hand (49), followed by forearm or upper arm (6), and torso (5). In most cases, a limb or part of a limb was trapped between the rear roller and the treadmill belt. Fourteen patients required surgery, including 13 who underwent a skin grafting procedure. Most injuries occurred while the treadmill was in use by others, with the children approaching unnoticed from behind (46). In nine cases, the injury happened when the patients themselves, at a mean age of 7.8 years (range, 2–12 years), were using the treadmill. The substantial increase in prevalence of treadmill injuries in children during the past 2 years may be related to increased sales of treadmills as the community becomes more conscious of obesity. The data also reflect other Australian studies that show that children younger than 5 years are at greatest risk, accounting for 90% of paediatric treadmill injuries during the period January 2004 to June 2007.2 Despite the risk of injury, particularly for children, there appears to be no current national regulations governing the supply of treadmills or advice that should be given to customers at the point of sale. The New South Wales Government introduced legislation in June 2009 mandating prominent permanent warning labels to be affixed to all new treadmills — the Fair Trading Amendment (Treadmills) Regulation 2008 (NSW). The NSW Office of Fair Trading, with assistance from the NSW Severe Burn Injury Service and Kidsafe NSW, has developed an alert poster (copies of which may be downloaded or ordered from their website) for display at childcare centres, playgroups and places where domestic treadmills are sold.3 The Australian Competition and Consumer Commission recently published a safety alert brochure on domestic treadmills, which contains a safety checklist.4 Although helpful, the brochure does not include previous recommendations such as caution with headset use (ie, decreased awareness of children near the treadmill), and the use of mirrors or alternative positioning to ensure children approaching the treadmill can be seen.1 As most injuries occur within the first 6 months of purchase of the treadmill,5 educating parents seems to be most important around the time of purchase. Design modifications could also reduce the risk of entrapment of a digit or hand.2 It is likely that, without better application of current injury prevention strategies, the prevalence of these injuries will continue to increase.

Lawrence H Kim · Deborah A Maze · Susan Adams · Sarah Guitonich · Siobhan Connolly · Anne Darton · Andrew J A Holland

To the Editor: Contact burns in children caused by hair-straightening devices are increasingly common. Although the dangers of hair dryers and other similar devices are well known,1 there is less awareness of the risks associated with hair straighteners. The relevant Australian Standard does not mention hair straighteners.2 Four recent studies from the United Kingdom have reported on this problem,3-6 but there is no readily identifiable published information from Australia. Hair straighteners consist of two opposing ceramic plates that are held apart when not in use. The plates are reported to reach average temperatures of 169.5°C within 4 minutes 20 seconds of being switched on. They can cause burns (temperature > 66°C) on short-term contact (10 seconds) for a period of up to 9 minutes 20 seconds after being switched off,3 and can take 30 minutes to cool to below 50°C, at which temperature they can cause superficial burns on prolonged contact. Using data collected by the Stuart Pegg Paediatric Burns Centre at the Royal Children’s Hospital, Brisbane, and the Queensland Injury Surveillance Unit, we identified 22 patients treated for hair-straightener injuries between January 2004 and June 2009. Sixteen of these were treated within the past 2 years. The median age of patients was 43.4 months (range, 9 months to 14 years). A mean of 1% of total body surface area was involved. Injuries were to the forearm and hands (16 patients) (Box), foot and lower leg (five patients), and the back (one patient). The burns were significant, with 19 partial-thickness burns, and three full-thickness burns requiring surgery. Nine of the 22 children (41%) required long-term scar management. We observed two typical patterns of injury. In toddlers (16 patients aged 9–48 months), the main mechanism of injury was grasping or pulling down a hair straightener that was either turned on or cooling, with inadequate supervision a common factor. An early-teen group (three patients) had self-inflicted burns from accidental contact or misuse, including one patient who misguidedly used the device in an attempt to remove leg hair, sustaining full-thickness burns requiring skin grafting. Increased awareness of the potential dangers of hair straighteners might help prevent burns. We suggest four precautions: Hair straighteners should be placed out of reach of children during use and storage; Children should be supervised while the device is warming or cooling; Manufacturers should label the device to warn of potential dangers; and Manufacturers should either redesign the device so that plates are not exposed, or provide a cool-touch cover. Burns to a toddler’s hand caused by contact with a hair-straightening device

Zoe M Poiner · Michael D Kerr · Belinda A Wallis · Roy M Kimble

To the Editor: The current pandemic influenza A (H1N1) strain first caused infections in Mexico in April 2009 and rapidly spread to over 160 countries. Confirmed laboratory infections now number over 160 000, with millions of people probably already infected and further spread inevitable.1 Fears have been expressed that the enormous death toll seen with the H1N1 “Spanish ’flu” pandemic of 1918–1919 might be repeated. Although many deaths during that pandemic were caused by the direct effects of the influenza virus, over 95% of deaths were due to secondary bacterial pneumonias.2,3 If, as in 1918, most people with the current H1N1 strain have a mild illness from which they fully recover, this raises the important question of how we can readily identify patients co-infected with bacterial pathogens who may need antibiotic treatment for pneumonia. This is important in a situation where large numbers of people may need to be assessed, and demand for both antiviral and antibacterial agents may be high. Certain clinical features, such as the presence of a biphasic illness or the late development of purulent sputum, may suggest bacterial infection, but we do not know how reliable these features will be.4 The Australian Community-Acquired Pneumonia (CAP) Study was the largest ever prospective aetiological study of CAP.5 All patients were assessed for both bacterial and viral pathogens, including seasonal influenza.5 Using data from that study, easily measured clinical markers in patients infected with both influenza virus and a bacterium were compared with markers in patients with only an influenza virus identified. Patients with both influenza virus and a bacterial pathogen tended to be younger and appeared to have poorer outcomes.6 The most notable clinical differences between the two groups at presentation were the higher mean respiratory rate and faster heart rate in those with influenza virus plus a bacterial pathogen (Box). However, there were major overlaps between the groups in these parameters, so they were not very discriminatory if used alone. Taken with other clinical features,4 our data suggest that the presence of a respiratory rate of ≥ 25 breaths/min and a heart rate of ≥ 100 beats/min may help identify people who are more likely to need prompt clinical assessment and a chest x-ray. While these features may help identify patients more likely to benefit from antibacterial therapy, we would also argue that the small proportion of patients whose influenza is serious enough for them to be admitted to hospital should probably be treated with empirical antibiotics as well as antivirals. Patients with influenza plus a bacterial pathogen compared with patients with influenza alone or a bacterial pathogen alone Influenza plus bacterial pathogen (n = 17) Influenza alone (n = 51) P Bacterial pathogen without influenza (n = 293) Mean age (years) (SD) 55.2 (26.1) 66.7 (20.0) 0.046 62.9 (20.9) Male sex (%) 52.9% 54.9% 0.89 62.1% Mean RR (breaths/min) (range) 28.6 (16–48) 24.6 (16–48) 0.14 24.9 (12–60) Age-adjusted tachypnoea (%)* 47.1% 25.5% 0.10 28.0% Mean systolic BP (mmHg) (range) 137.4 (107–215) 137.1 (65–196) 0.97 128.3 (60–215) Mean diastolic BP (mmHg) (range) 68.2 (45–116) 69.5 (25–103) 0.81 67.1 (29–116) Mean temperature (°C) (range) 37.8 (36.2–40.6) 37.8 (35.5–39.6) 1.0 37.9 (33.5–40.8) Mean pulse rate (beats/min) (range) 113.4 (75–145) 96.3 (56–152) 0.01 104.5 (43–175) Mean Spo2 (range) 92.4 (81–99) 91.8 (44–100) 0.79 92.6 (50–100) Mean SMART-COP score† (range) 2.7 (0–7) 2.2 (0–9) 0.49 2.5 (0–9) Need for intensive care (%)‡ 29.4% 13.7% 0.14 13.0% 30-day mortality (%) 5.9% 0 0.08 6.5% BP = blood pressure. RR = respiratory rate. Spo2 = oxygen saturation (as measured by pulse oximetry). * Age-adjusted tachypnoea was defined as RR ≥ 25 breaths/min in patients aged ≤ 50 years or RR ≥ 30 breaths/min in patients aged > 50 years.6 † A tool for determining severity of community-acquired pneumonia (for details, see Charles et al6). ‡ All patients required either mechanical ventilation or vasopressor support.

Patrick G P Charles · Paul D R Johnson · Peter J Collignon

To the Editor: A recent article in the Journal describes, again, barriers to implementation of cardiometabolic monitoring among patients prescribed antipsychotic drugs.1 The cardiac health of patients with psychosis is not routinely assessed at first presentation for mental health services, adverse side effects of antipsychotic drugs are not systematically monitored, and patients with treatable risk factors for heart disease are not identified.2 We propose a practical solution to the seemingly intractable problem of implementing guidelines for cardiometabolic monitoring — change the delivery system. We have employed a general nurse to conduct cardiometabolic monitoring in a pilot study at the Recovery And Prevention of Psychosis Service (RAPPS), a first-episode psychosis service in Melbourne. All 15 eligible patients had their height, weight, blood pressure, waist circumference, fasting total cholesterol, high- and low-density lipoprotein cholesterol, triglycerides and glucose assessed according to national guidelines3 within 1 month of entry to the service, in the hospital, as an outpatient, or in the patient’s home; 14/15 blood samples were taken while the patient was fasting. Very early monitoring (within 7 days of first exposure to antipsychotics) was not implemented for four patients because they were inpatients and judged by ward staff as too unwell to be approached by a general nurse. Future follow-ups will be conducted at 3, 6, 12 and 18 months. Abnormal findings are referred to the treating psychiatrist, who is responsible for ensuring the patient receives appropriate follow-up. A general nurse can implement clinical guidelines, but this initiative requires substantial planning and ongoing management. Systematic identification of all patients eligible for monitoring requires identification of all pathways into the relevant mental health service, so as to begin monitoring at, or very close to, the point of first exposure to antipsychotics; management tools to track patients over time; and a clinical pathway to track test results and ensure appropriate medical interventions occur when required. Failure to implement prescribed monitoring guidelines is important because individuals with schizophrenia have a 20% shorter life expectancy than individuals in the general community.4 Side effects of antipsychotic drugs may include dramatic weight gain and elevations in serum cholesterol and glucose levels, which exacerbate the risk for cardiovascular disease. Most early deaths among individuals with schizophrenia are due to cardiovascular disease.5 Failure to monitor cardiovascular health and the adverse side effects of antipsychotic drugs is an important, life-shortening, failure of care. A simple solution to a complex problem exists if an effective delivery system is used.

Debra L Foley · Katherine I Morley · Karyn E Carroll · John Moran · Patrick D McGorry · Brendan P Murphy

In reply: Foley and colleagues rightly point out that a way to improve the cardiometabolic health of patients with psychosis is to change the way that mental health services are delivered. Although barriers to monitoring exist at the level of the patient, the illness, and the service,1 by focusing too narrowly on the barriers presented by patients, a blaming culture can be perpetuated. If blame is to be attributed, it should be directed towards inflexible services with a medieval belief in separating mental and physical health care. A number of centres in Australia have started to innovate in service delivery, with structured physical health clinics running in parallel to, and integrated with, mental health clinical programs. Our own centre, the Concord Centre for Cardiometabolic Health in Psychosis (ccCHIP), has been developed to take the notion of integrated care a step further — to actually treat the cardiometabolic abnormalities present. Our model involves a multidisciplinary team comprising psychiatrists, endocrinologists, and dietitians. However, we believe the potential for broader multidisciplinary input exists, including nurses, pharmacists, psychologists, occupational therapists, social workers and the patient’s general practitioner. It is our philosophy that although detection is the first step to improving the parlous outcomes for our patients, without active intervention, these poor outcomes are unlikely to improve. Recently, we received funding from the New South Wales Department of Health to develop a more comprehensive plan for education and training, including the production of a manual, to help psychiatric services in NSW develop their own monitoring and intervention services, using ccCHIP as their resource base. This initiative points to the need for government involvement to support these initiatives. Finally, it is apposite that Foley and colleagues write from the perspective of an early psychosis service — we believe that early detection and intervention for psychosis should be for physical as well as mental health issues.2

Timothy J R Lambert

To the Editor: The growing shortage of medical educators, accentuated by the 20% increase in the number of medical students in Australia in 2007–2008, as described by Joyce and colleagues,1 is a concern that demands attention. It is understandable that clinicians might prioritise the provision of clinical services over the “less urgent” demands of teaching. However, if nothing is done about the shortage of medical trainers, the recent increase in medical student positions will be in vain. I propose that medical education should be taught as part of the medical school curriculum, with the aim of increasing the involvement of future doctors in teaching. Studies in other countries have shown that programs such as “Training Tomorrow’s Teachers Today” increase the competence and confidence of medical students as educators.2 Introducing peer education would also increase teaching opportunities for students; enlisting senior students to teach junior students has benefits for both, and has been shown to improve teaching technique over time.3 The inclusion of education in undergraduate curricula would provide students with valuable teaching skills and consolidate prior learning. By training students to be teachers, medical education would be promoted as the “norm”, which might help alleviate the growing shortage of clinicians involved in teaching. Although few doctors today take the Hippocratic Oath, perhaps we need to be reminded that it includes a promise to “teach the art” of medicine, as well as (arguably) the familiar precept “first, do no harm”.

Timothy C Mulherin

To the Editor: We wish to call for the removal of scheduling and legislative barriers in Australia that prevent easy access to naloxone for administration by peers to people suffering from a heroin overdose. Use of illicit opioids, typically heroin, remains the major cause of illicit drug-related mortality in this country, with at least one accidental opioid overdose currently occurring each day.1 Although population levels have not reached those seen during the peak in the late 1990s, geographically localised and transient increases in overdoses are evident.2 Death from heroin overdose typically occurs some time after use. In many cases, other people are present, and there is considerable scope for intervention to prevent death.3 Yet, in more than 70% of cases of fatal overdose, there is no intervention, and, where action is taken, calling an ambulance is seldom the first strategy, resulting in even greater risk of death.3 Opioid substitution treatment (with methadone or buprenorphine) is the mainstay of overdose prevention in Australia. Other interventions implemented here include outreach services and education for injecting drug users about the risks of overdose and how to respond to it.3 In 2000, Lenton and Hargreaves summarised the evidence for distributing the opioid antagonist naloxone for administration by peers to prevent deaths from heroin overdose. They concluded that an Australian research trial was needed.4 However, in 2001, the Australian heroin market was disrupted, heroin use and overdoses declined, and the trial did not proceed. Since then, emerging international evidence has demonstrated that injecting drug-using peers, family members and outreach workers can successfully administer naloxone to reverse heroin overdose — with few, if any, adverse effects.5 By December 2008 in the United States, 52 programs distributing naloxone for administration by peers were operating in 17 states, with over 1000 documented overdose reversals resulting from these programs.5 Most concerns about the intervention — such as the possibility of unsafe naloxone administration, reintoxication or more risky drug use — appear to have been unfounded, and naloxone administration by trained peers has been shown to be a remarkably safe intervention.5 In our view, the international evidence clearly indicates that increased naloxone availability will prevent many cases of fatal overdose, that conducting a trial in Australia is now unnecessary, and that naloxone should be made available without delay to be administered by peers in cases of opioid overdose. Careful monitoring and evaluation should be a part of this process. We call on all Australian states and territories to immediately enact Good Samaritan legislation to legally protect laypeople using naloxone in emergency situations. Naloxone should be reclassified from a Schedule 4 (S4) drug (available only on prescription) to S3 or S2 to make it available over the counter. As naloxone is no longer under patent,5 there may be little financial incentive for a drug company to pursue rescheduling. However, it could be rescheduled in Australia under provisions that allow state health authorities, professional associations or the National Drugs and Poisons Schedule Committee to initiate the process. Heroin overdose deaths are preventable. We need to take action now to enable peer-led intervention to reduce this serious outcome.

Simon R Lenton · Paul M Dietze · Louisa Degenhardt · Shane Darke · Tony G Butler

To the Editor: On behalf of the Australian Association of Musculoskeletal Medicine (AAMM) and the Australasian Faculty of Musculoskeletal Medicine (AFMM), we would thank Briggs and Buchbinder for raising the topic of back pain as a National Health Priority Area (NHPA) for debate.1 The AAMM and its teaching arm, the AFMM, have been arguing for years that higher priority should be given to the management of low back pain, and we certainly support the arguments for making back pain an NHPA. We agree that back pain is a major burden on society, and that coordinated action that includes all stakeholders is required. Currently, members of the AAMM and AFMM are involved in contributing to the Australian Core Competencies in Musculoskeletal Basic and Clinical Science project,2 which aims to standardise the undergraduate teaching in musculoskeletal medicine around Australia. As well as supporting undergraduate initiatives, the AAMM and AFMM have been the significant providers of postgraduate education for doctors around Australia over the past 20 years, being engaged in activities that include the development of evidence-based guidelines and university-based postgraduate diploma and masters programs.3 It is then very disappointing to read disparaging comments about doctors with a special interest in musculoskeletal medicine. Briggs and Buchbinder assert that one potential disadvantage of making back pain a health priority is that it may provide “justification for those with . . . vested interests to promote clinically ineffective interventions”. The example they cite is a questionnaire survey of Victorian doctors in 2004, highlighting that a self-reported interest in low back pain or musculoskeletal medicine, or both, was strongly associated with back pain management beliefs and practices that are contrary to the best available evidence.4 The title of the Spine journal article4 is mischievous and misleading. The article implies that members of the AAMM have poorer knowledge about low back pain than doctors with no special interest in back pain. The real facts are that no attempt was made by the authors to specify what training or continuing medical education doctors had received, or whether they were members of the AAMM.

Victor J Wilk · Michael J Yelland · Michael B Oei

To the Editor: In the recent article by Briggs and Buchbinder, the authors propose that one advantage of including back pain as a National Health Priority Area (NHPA) is that it will increasingly encourage the management of back pain in accordance with best-practice clinical care guidelines.1 The emergency department (ED) is a key point of presentation for people with back pain. Research suggests that optimal management of low back pain incorporates multidisciplinary input, ensuring the provision of adequate analgesia and rational use of further investigations.2,3 On the basis of such data, we devised an interactive decision-support tool for completion by ED staff. The “Low Back Pain Assessment and Treatment” (LBPAT) guidelines were designed by a multidisciplinary team of nursing, medical, physiotherapy and pharmacy staff, based on accepted evidence-based standards of practice already in use.4 Use of these guidelines enables rapid assessment of “red flags” that may require further investigation and treatment. The guidelines include a flow diagram divided into three pain-management options for patients with mild, moderate or severe pain, with prompts for referral to physiotherapy and neurosurgery. The LBPAT guidelines indicate which imaging studies and blood tests may be required and when, as directed by findings. Information to assist discharge planning, including follow-up by a physiotherapist and general practitioner, and provision of information, is included. Once they were developed, we evaluated the effectiveness of the LBPAT guidelines in improving ED clinical practice. A retrospective case-record study of patients presenting to the ED and identified from International classification of diseases, ninth revision (ICD-9) codes (for back pain or low back pain) was undertaken before implementation of the guidelines (October–December 2006; 87 patients), immediately after implementation (May–August 2007; 96 patients), and about a year later (July–August 2008; 28 patients). Uptake of the LBPAT guidelines was only 47%–50% in both postimplementation periods. However, adherence to pain management guidelines improved significantly after implementation (59% [57/96] compared with 46% [40/87]; P = 0.02), and was maintained a year later (57% [16/28]). Use of aperients in patients who were coprescribed opiates was much improved in the period immediately after implementation (40% [27/68] compared with 16% [11/69]; P < 0.001) and 1 year later (70% [14/20]). Postimplementation data suggested a reduction in unnecessary blood tests (ie, no red flags requiring further investigation). For patients discharged from the ED, discharge planning and continuity of care did not improve immediately after implementation, but were greatly improved 1 year later. If back pain is included as an NHPA, use of a clinical decision tool such as the LBPAT guidelines is one potential method of preserving resources and improving patient outcomes in cases of back pain in the ED setting.

Stephanie K Vaughan · Julie L Gawthorne · Andrew S Finckh · Susan A Welch

In reply: The two preceding letters highlight that there are many stakeholders with an interest in back pain in Australia. Many craft groups are involved with the provision of postgraduate education and have contributed to the musculoskeletal core competencies initiative. Vaughan and colleagues describe promising results from using a clinical decision tool for acute back pain in an emergency department. On a national scale, implementation of such a tool in all emergency departments may significantly improve outcomes and reduce costs. Making back pain a national priority would provide unique opportunities to rigorously study these types of approaches. A timely illustration of our concern about vested interests potentially promoting ineffective or unproven interventions1 is the lively debate centred on the recently published United Kingdom National Institute for Health and Clinical Excellence guidance for the early management of persistent non-specific low back pain.2 The recommendation that injections of therapeutic substances into the back for non-specific low back pain should not be offered, as the evidence of effectiveness is lacking, led to the forced resignation of the president of the British Pain Society because of his refusal to denounce the guidelines he helped to formulate.3 To suggest that the title of the Spine journal article is misleading and mischievous is patently incorrect, as the title, “Doctors with a special interest in back pain have poorer knowledge about how to treat back pain”, simply summarises the main study findings.4 Based on 3831 responses from general practitioners in New South Wales and Victoria, collected in 1997, 2000 and 2004, doctors who reported a special interest in low back pain were significantly more likely to believe that patients with acute low back pain should be prescribed complete bed rest until the pain goes away (relative risk [RR], 1.89; 95% CI, 1.53–2.33); that they should not return to work until they are almost pain-free (RR, 1.55; 95% CI, 1.31–1.83); and that lumbar spine x-rays are useful in their work-up (RR, 1.36; 95% CI, 1.21–1.52). We also reject the suggestion that we made disparaging remarks about doctors with a special interest in musculoskeletal medicine. After adjusting for the presence of special interests in back pain, there were no important differences in back pain beliefs between those with and without a special interest in musculoskeletal medicine. While we cited a survey that found that Australian Association of Musculoskeletal Medicine (AAMM) members see a high caseload of patients with back pain,5 we made no claims about the knowledge base of AAMM members.

Rachelle Buchbinder · Andrew M Briggs

To the Editor: We thank Page and colleagues for their important article considering the sociodemographic correlates of antidepressant utilisation in Australia.1 We note that fewer than 15% of the young people in the study were prescribed fluoxetine, and almost 40% were prescribed sertraline. These rates of antidepressant use contrast with the available evidence on treating young people diagnosed with depression. A recent review examining the effectiveness of selective serotonin reuptake inhibitors (SSRIs) for depression among children and adolescents demonstrated that fluoxetine is the only SSRI with at least some evidence for effectiveness.2 Current clinical guidelines recommend that a young person diagnosed with a major depressive disorder who is to be prescribed an antidepressant should be given fluoxetine in the first instance.3,4 Reasons for the apparent lack of concordance with the guidelines might include treatment of disorders other than depression, or prescriptions for those who have already had an unsuccessful trial of fluoxetine. It would thus be of great interest to learn what proportion of young people are prescribed an antidepressant other than fluoxetine, and what proportion of young people prescribed an antidepressant are concurrently undergoing guideline-concordant psychological treatments, such as cognitive behaviour therapy. The data presented by Page et al draw attention to challenges faced by doctors providing treatment for young people experiencing depression, which include a lack of good evidence about the effectiveness of newer antidepressants for this age group. However, there is an opportunity to support better use of evidence in decisions made about treatment options for young people. The provision of high-quality, evidence-based information for patients and their carers to enable informed decisions is essential, and shared decision making offers a way to enable this.5 By improving the knowledge transfer between doctor and patient, antidepressant prescription can be more judicious.

Magenta B Simmons · Michaela R Willet · Sarah E Hetrick

To the Editor: In their editorial commenting on our viewpoint article on Team Care Arrangements (TCAs), Harris and colleagues acknowledge a number of problems with TCAs and a lack of evidence for their effectiveness. However, they state that “the idea of team care still has merit” and that although there is “ample room for improvement in the current process”, they “cannot agree with Hartigan et al’s radical prescription to replace TCAs with patient summaries”.1 This seriously misrepresents our article. We did not suggest that TCAs should be replaced by patient summaries. We stated: “There are many elements in the solution to this problem, but one of primary importance is the development of a comprehensive patient summary”.2 Among these elements could be the one suggested by Harris et al — registration of patients in a shared database. However, a shared database would avail little if the patient information contained therein is not sufficiently comprehensive to permit development of well considered treatment objectives and priorities. The essential issue is whether TCAs are an improvement on the coordination that occurs under the traditional referral system. Coordination exists along a spectrum. At one end, it can be rigidly structured, with all participants having right of veto over action by any of the others. At the other end, it can be fluid and unstructured, with decisions over who should be involved in any particular management issue left to the good judgement of participants. By electing to impose on general practitioners a coordination model at the rigid end of this spectrum, the architects of TCAs appear to assume that GPs are usually unwilling, or cannot be bothered, to consult with other health professionals and must be made to do so. This is a massive assumption that demands very strong evidence. What is the evidence? And even if there were evidence, is coercion the answer? There is no doubt that coordination in care of patients with chronic illness is important and that it can be improved. However, improvements must be based on evidence and be practical. TCAs are neither.

Paul A Hartigan · Tuck Meng Soo · Marjan Kljakovic

In reply: We accept the clarification by Hartigan and colleagues of their article’s intent,1 and agree with the importance of good-quality information in electronic databases, especially if the information is to be exchanged. Unlike Hartigan et al, we do believe that Team Care Arrangements (TCAs) have some merit, as they provide structure for planned care and help coordinate access to multidisciplinary care provided over a period of time within the complexity of the current health system. However, they have not been very effective in developing trusting relationships and encouraging better communication between general practitioners and providers outside the practice because of a variety of systemic barriers. In our editorial, we suggested some alternative strategies.2 Although the paperwork and complexity of current requirements could be reduced, we cannot see that coercion is involved. Both patients and providers can choose to enter freely into TCAs and the requirements are for communication and consent, not for “right of veto over action”.

Mark F Harris · Bibiana C Chan · Sarah M Dennis

To the Editor: The Therapeutic Goods Association (TGA) has informed Australian eye banks that nucleic acid testing (NAT) of donor sera will be required in addition to routine serological tests for hepatitis B virus (HBV), hepatitis C virus (HCV) and HIV. NAT can detect viral genome in the window period between infection and the appearance of antigen or antibody, and is to commence in Australia as soon as possible. However, we believe the benefit of NAT would be small and the cost considerable. Importantly, it would be likely to have an adverse effect on the availability of donor corneas for transplantation. HBV, HCV or HIV have not been transmitted from seronegative donors by means of corneal transplantation. In the mid 1980s, HBV was almost certainly transmitted by corneal tissue to two recipients in the United States who developed acute hepatitis B; HBV surface antigen was detected subsequently in the donor sera.1 The transmission of HCV or HIV has not been reported despite corneal transplantation from infected donors.2,3 The American Red Cross instigated HCV/HIV (but not HBV) NAT for blood donors in 1999, and the US Food and Drug Administration mandated HCV/HIV NAT for eye donors in 2007. So, how many window-period eye donors could be identified by NAT? It has been estimated that 7.2 per 100 000 American tissue donors are in this period for HBV, HCV or HIV.4 With NAT, the number reduced to 1.8 per 100 000 (most assays, including NAT, have sensitivity limitations). Let us assume that the prevalences of HBV and HCV are the same in Australia and the US, the prevalence of HIV in Australia is half that of the US, and that prescreening by medical and social history is equally effective in both countries. At 700 corneal donors (1200 grafts) per year in Australia, one window-period donor would be expected every 23 years, falling to one every 93 years with NAT. Thus, one window-period donor would be detected by NAT every 30 years at an estimated cost of $9 million ($50 per donor averaged up for out-of-hours testing and kit wastage) plus any charges for specimen transport. If only HCV/HIV NAT is performed, such a donor would be detected every 52 years. What is the actual risk of infection? Assuming 25 000 corneal donors (43 000 grafts) per year in the US, the figures4 suggest that eyes were collected from one window-period donor every 7 months between 1990 (when a serological test for HCV became available) and 2007, but no infections have been reported. Perhaps there was no virus in the corneal tissue; even among donors who are seropositive for HBV, HCV or HIV, few have detectable viral genome in the cornea.5,6 It is not known whether these viruses can invade the cornea before the appearance of antibody, but the risk that infection will occur in recipients of corneas from HCV/HIV-seronegative, NAT-positive donors appears to be very small indeed. The TGA’s decision may have been based on factors other than this type of analysis. We believe the decision should be reconsidered.

Paul R Badenoch · Douglas J Coster

To the Editor: Compared with the Australian population, all-cause mortality rates among prisoners are four times greater among men, and eight times greater among women.1 Fatal drug overdoses and suicides comprise the overwhelming majority of deaths in post-release cohorts;2 hence, a large proportion of post-release deaths are preventable. However, the information necessary to determine appropriate public health responses to the issue of post-release mortality is largely absent, because there is no routine nationwide collection of numbers of prison separation episodes as there is, for example, for prison receptions.3 We aimed to estimate the number of inmate separation episodes occurring in Australia for the financial year 2007–08, using benchmark data obtained from public documents on the websites of each state and territory government department responsible for prisons. For each state and territory, we attempted to obtain the total number of inmates (ie, sentenced and on remand) released from prison for the financial year 2007–08. This figure was available for Victoria (5356), South Australia (4177) and the Northern Territory (2857). For New South Wales, data could only be obtained for separations of sentenced prisoners (7645). Hence, data on separations of either total or sentenced inmates were available for jurisdictions covering 62% of the national prison population. To adjust for separations of prisoners on remand in NSW, we assumed that separations of sentenced prisoners comprised 40% of all separations, reflecting the ratios of sentenced to total separations in other jurisdictions. Hence, the total number of separations in NSW was estimated to be 19 113. A multiplier of 1.6 (1/0.62) was applied to the sum of separations from Victoria, SA, the NT and the adjusted figure for NSW separations, in order to estimate separations for the national prison population. This produced an estimate of 50 405 prison separations nationally for 2007–08. Each of these separations is associated with a significant increase in risk of death. We have made an estimate of the number of prison separations in Australia for 2007–08, but believe that a more appropriate and precise method for obtaining this information would be to include separations for the total number of prisoners and for sentenced prisoners as items in routine data collection systems such as the Australia Bureau of Statistics Prisoners in Australia collection. This would enable better through-care service planning and provision of life-saving pre-release and post-release health services, such as prison-based opioid substitution treatment4 and overdose prevention and treatment interventions such as distribution of naloxone.5

Kristy A Martire · Sarah Larney

To the Editor: Pet abuse is considered another form of family abuse, and is a powerful marker for other forms of violence happening in the home. Women and children living with violence, isolated by their abuser, often form intense bonds with their pets. Perpetrators will use animal cruelty as a means to intimidate and control their partners and children.1,2 Women will put themselves at increased risk due to concern for their pets.1-5 RSPCA New South Wales acknowledges the link between domestic violence and cruelty to animals. In 2004, it initiated the Safe Beds for Pets program, which provides emergency accommodation for pets of women and children seeking refuge from domestic violence (http://www.rspcansw.org.au/programs/safe_beds_for_pets). The Central Coast, NSW, arm of this program conducted an anonymous, self-administered survey to identify the prevalence of perpetrator violence against the family pet and its effect on decision making by women. The survey was offered to local women who owned a pet, were aged 18 years or over, were experiencing domestic violence and had attended the Central Coast Domestic Violence Court Advocacy Service, and was conducted between August and November 2007. Of 187 women, 92 owned a pet, 85 of whom completed the survey. Almost 75% (63/85) owned at least one dog, 45% (38/85) owned at least one cat and 22% (19/85) owned at least one bird. We found that 27% (22/82) of partners had threatened to hurt or kill a pet, and 23% (18/80) had done so. These data are at the low end of findings of other studies, which found that 40%–53% of partners threatened to hurt or kill a pet, and 26%–57% had done so.3 Out of concern for their pets, 27% (22/83) of women in our study delayed leaving home (compared with 33% in another study3), and 27% (23/85) returned to their partner. Women also delayed calling the police (17%, 14/84) and refused to lay charges (11%, 9/83) out of concern for their pets. Women whose partners had threatened or killed a pet were more likely to delay leaving, return home, delay calling police, or refuse to lay charges than women whose partners had not (Box). The study’s main weakness was that it was restricted to current pet ownership, thus excluding women who may have recently rehomed or abandoned their pets, or whose pets may have already been killed.3 Clinicians do not necessarily ask clients about the role of pets in their health. If the issue is not raised, women often feel reluctant to speak up.4,5 We encourage clinicians to: ask patients experiencing domestic violence whether they have pets, whether their pets are at risk of abuse, and whether they need help to keep their pets safe; and talk with human and animal service providers about opportunities for cross-service programs, such as Safe Beds for Pets. We support the development of a nationwide strategy allowing the simultaneous relocation of a woman, her children, and pets from a violent home.3 Relative risk (95% CI) of women’s decisions if their abusive partners threatened or hurt their pets Partner’s action Delayed leaving Returned home Delayed calling police Refused to lay charges Threatened to hurt 3.00 (1.54–5.83) 3.78 (1.95–7.30) 34.29 (4.74–247.97) 9.00 (1.97–41.07) Threatened to kill 2.41 (1.23–4.71) 3.64 (1.91–6.94) 5.58 (2.21–14.09) 4.12 (1.28–13.32) Threatened to hurt or kill 2.81 (1.44–5.50) 3.55 (1.82–6.89) 32.18 (4.44–233.18) 8.43 (1.84–38.57) Actually hurt 1.93 (0.92–4.07) 2.12 (1.04–4.30) 7.63 (2.66–21.88) 6.78 (1.88–24.43) Actually killed 1.35 (0.26–7.01) 4.59 (3.01–6.99) 5.78 (2.57–12.97) 6.33 (1.82–22.00) Actually hurt or killed 1.93 (0.92–4.07) 2.12 (1.04–4.30) 7.63 (2.66–21.88) 6.78 (1.88–24.43)

Cheryl Travers · Andrew Dixon · Karen Thorne · Kaye Spicer

To the Editor: Training of adequate numbers of surgeons and other medical specialists has, until recently, been a sleeper issue.1‑3 Trainees are intimately involved in public hospital patient care, but it is generally assumed that private patients are off limits for training. To our knowledge, the perceptions of patients in a private hospital setting have not been formally explored before. We present the results of a survey of attitudes of private hospital patients towards trainee involvement in their care. At a major private hospital in Melbourne, 82 consecutive inpatients undergoing elective general surgery over a 6-week period in 2005 were invited by us to respond to a survey (given prior to surgery and collected prior to discharge). Ethics approval was obtained. Sixty-eight survey forms (83%) were returned by the 82 patients. The median age of respondents was 56 years (range, 18–78 years), and 35 (51%) were men. The most important reasons given for having private health insurance were choice of surgeon (68% of respondents), quality of medical facilities (28%) and shorter waiting-list time (4%). Availability of trainees in the hospital to (a) communicate with the surgeon, (b) manage emergencies, and (c) perform minor procedures were all ranked as very to highly important, as was the ongoing education of trainees. When asked how comfortable they felt with trainees being involved in their care, 88% of patients said they supported having a trainee assist the consultant surgeon with surgery. Providing the trainee was under direct supervision of the surgeon, patients supported allowing a trainee to perform part of the surgery (75% of respondents), most of the surgery (50%), or the entire operation (41%). No significant differences in attitudes attributable to patients’ age or sex were found (Mann–Whitney U test). Within the limits of our small survey structure and incomplete response rate, our results show that most patients support the concept of having a trainee involved in ward care. There was polarised opinion on how much of the operation could be performed by the trainee, and this may vary for different subspecialties. Currently, junior medical staff work in many private hospitals, gaining valuable experience in perioperative care and exposure to a range of surgery. These jobs would be ideal for junior Royal Australasian College of Surgeons (RACS) trainees at Surgical Education and Training Level 1 or 2. However, individual surgeons may be concerned about the potential for increased complications4,5 and longer operating times, with a reduced number of cases and earnings per list. Discussion of medicolegal issues, remuneration and accreditation of posts would need to take place between hospitals, government and the RACS. It is clear that increasing demand for surgical training necessitates the consideration of an accredited role for the advanced surgical trainee in the private health care system.

Lih-Ming Wong · Lisa P Wun · Julie Shaw · Dominic Vellar