Volume 186 Issue 3

Vaccination response has been impressive, but the hypermutable meningococcus is likely to continue to challenge us In the past 100 years there have many dramatic fluctuations in the incidence of meningococcal disease in Australia, as Patel describes in this issue of the Journal.1 Rising socioeconomic status with reduced household crowding has probably been the major factor in curbing the overall incidence of the disease in recent decades,1 but the introduction in early 2003 of a routine vaccination against serogroup C for all infants (and a catch-up campaign for those aged under 20 years) has also had an undeniably impressive impact. Worldwide, surges in the incidence of meningococcal disease due to serogroups A, B, C, W135 and Y have each been associated with the advent of a new strain to which the community largely lacks immunity. Furthermore, new strains are often associated with a higher case-fatality rate and a shift to the right in age incidence, with the disease especially affecting older teenagers. More commonly, these young people do not have the disease but are colonised by the organism and may become “superspreaders” of infection.2,3 The epidemiological signal of rising serogroup C meningococcal disease incidence was detected in the United Kingdom in 1994. A triumvirate of government, industry and academia combined to produce a “win, win, win” situation — an effective and safe vaccine being researched, developed, mass produced and introduced within a mere 5 years.4 As a consequence of vaccination, serogroup C disease is now controlled in a number of developed countries. However, the incidence of serogroup C meningococcal disease in teenagers had already peaked in half the regions of the UK and was declining before the vaccine was introduced in 1999. Similarly, in New South Wales, disease incidence was falling before the vaccine was introduced (perhaps due to naturally acquired herd immunity5), whereas, in Victoria, vaccination led to a rapid fall in incidence. Western Australia and South Australia had not seen the same surge of serogroup C disease as in the eastern states.6,7 However, it should be emphasised that, in the 3 years after the vaccine was introduced, there was a decrease in notification rates in all states and territories except WA, where notification rates remained low. As in the UK, a catch-up campaign was undertaken in Australia for those younger than 20 years of age. Unlike the UK, where routine vaccination was given as a three-dose course in the first 6 months of life, Australia was able to follow a simpler, more cost-effective approach and offer a single dose at 12 months of age. This was because almost all serogroup C disease was seen in older children, teenagers and young adults, and also because the vaccine is more immunogenic in older infants and children. Australia’s approach has proved successful. A similar outcome has been achieved in the Netherlands, where a single dose is administered routinely at 14 months of age.8 In 2005, only 50 laboratory-confirmed serogroup C cases were notified in Australia. This is a greater than 75% reduction from the 213 cases in 2002. There is also evidence that, in the absence of vaccination, disease incidence may fall as naturally acquired immunity rises5 but the impact is not as rapid; herd immunity may also be induced by vaccination but its longevity is uncertain.9 Given the extraordinary velocity of the vaccine’s development in the UK, inevitably some issues received attention later than desirable. One was cost-effectiveness — data on this were published well after vaccine introduction.10 There was interest in both the UK and Australia in the possible use of the cheaper polysaccharide vaccine, as it is effective in the age ranges mostly affected by serogroup C disease and may also be more cost-effective.11 This approach was tried in Spain, but waning immunity after several years led to a follow-up conjugate C vaccine campaign. Given that this disease can progress with frightening rapidity and has a high case-fatality rate, it receives, arguably, inordinate media attention. The political pressure and impetus to implement a vaccination program that this publicity generates is often lacking for diseases with a lesser public profile. When meningococcal C vaccine was introduced in Australia, there was debate about its merits in relation to, for example, the pneumococcal conjugate vaccine, which is now also part of the routine schedule. Australia was arguably the first country in the world to fund universal immunisation against Haemophilus influenzae type b, meningococcus C, and pneumococcus in early childhood. Long-term follow-up will be required to assess the ultimate success of the meningococcal C vaccine program and determine whether booster doses of vaccine might be required or whether serogroup replacement develops. The problem of serogroup B disease remains. New Zealand has recently completed a large and expensive public health intervention to vaccinate its young people against a particular serogroup B strain that, beginning 15 years ago, caused a major upsurge in incidence. This vaccine would cover only a small proportion of serogroup B cases in Australia,6,7 so a different product will be required here, and studies are underway in several Australian centres to assess candidate serogroup B vaccines. Without a vaccine, it is questionable whether behaviour modification would mitigate risk. Patel claims that “the societal determinants of the current hypersporadic disease pattern are unknown”.1 This is at odds with Australian and overseas data clearly showing that crowded environments, including university colleges and nightclubs, and risk-taking behaviour (eg, smoking and multiple deep-kissing contacts) are important risk factors12 and may be modifiable. Despite the recent unravelling of sequence data for the entire genomes of representative strains of meningococcal serogroups A, B and C, the hypermutable meningococcus will likely continue as an “accidental tourist” causing havoc in a small proportion of vulnerable hosts1 and go on challenging public health control measures throughout the 21st century.

Robert Booy MD, FRACP, FRCPCH · Jane Jelfs PhD · Haitham El Bashir FRCPCH, MRCP · Michael D Nissen FRACP, FRCPA

Medical education needs recognition and strengthening through research Modern medical practice relies heavily on research evidence to guide us towards predictable, safe outcomes of investigations and procedures that aim to improve the quantity and quality of life. As there are still substantial gaps in knowledge, researchers are supported to add to the evidence base. New technologies are proven to be (or not to be) safer or more effective. This continual reflection on practice and drive to expand options and reduce error margins is lauded as scientific progress. On the other hand, modern practice in medical education relies mostly on experience, opinion and rumour as it strives to produce graduates to meet the increasingly high standards of the community, partly in reaction to well publicised medical errors.1,2 Until recently, medical graduates received biomedical science training and apprenticeship-style clinical teaching without defined national exit standards. The current trend is towards broader (and possibly shallower) curricula guided by contemporary definitions of medical practitioner roles, such as CanMEDS from Canada, Good Medical Practice from the United Kingdom and local adaptations.3-5 National licensing examinations are either in place or under consideration, and international curriculum frameworks and accreditation are being developed.6 We are amidst an expansion in medical education driven by a workforce shortage. However, do we know how to achieve this expansion and maintain quality? There is a relatively small evidence base to support changes in medical education practice. The science of medical education has evolved much more slowly than the science of biomedicine. Lectures have been around for centuries, human dissection for about 500 years, and laboratory sessions for over 100 years. The only two real jolts to the conventions of medical education during the 20th century were the Flexner report, which standardised pre-clinical and clinical training phases because of concerns about a proliferation of small, lower quality medical schools in the United States,7 and problem-based learning, which was based on an attractive theory that case-based, integrated learning was more likely to be remembered when similar clinical presentations were encountered later. Neither change was based on strong evidence, and many years later there is still little education research evidence that any curriculum approach is better.8,9 A fundamental problem is that the academic basis of medical education receives little recognition amidst complex agendas. From a health care management perspective, the main role of the health care system is the provision within budget of quality clinical care, and it is difficult to direct funding towards academic development. From a university perspective, medical schools are sources of resources and prestige, but mostly through research achievement. As a result, medical schools are becoming “two-team” institutions. First, there are the researchers, who charge ahead, based on well planned and supported ventures, aiming to “score” grants and publications, win competitions and attract attention, usually based on advances in biomedical and clinical sciences. Then there are the teachers, who stay in the background and guide the students. This is the team that earns the core, stable funding of the school and is responsible for seeing that graduates meet the broader curriculum objectives, but it receives little attention except during medical school accreditation processes. The focus on research means that the real rewards of medical academic life generally go to members of the research team. Consider these questions: Why do some basic scientists who participate in education research decline to list that on their curricula vitae? Why are there apparently clear measures of research success, but not of teaching success? How hard is it for exemplary teachers to gain senior promotion without research success? How many national honours go to people who teach rather than research? How many prizes are available to reward great teachers? Why are there “league tables” for medical schools according to research success, reputation and difficulty of entry (based on very high academic performance), rather than for teaching quality? Just as clinical practice and research inform each other, so too do educational practice and research, and yet there are concerns about the variable quality and focus of recent research in medical education.10,11 There is a need to go beyond short term evaluation projects and to address current gaps in medical education research evidence, some of which are listed in Box 1. Most medical schools now have medical education units, but these are often seen as organisers of the teaching team, rather than contributors to a recognised academic agenda. These are missed opportunities, as medical schools manage large amounts of data about teaching and learning implementation, and could convert much of their teaching and learning activity to research without substantial cost. Medical education funding often underpins priority-driven research, and could more overtly support medical education research. The external funding environment could make medical education research more attractive to the research community. More medical academics could be trained in the methodology of education research, which differs from that of clinical and laboratory research, to increase research capacity. In summary, it is time for medical education research to enter mainstream research agendas and become a research priority for universities. Unless we strengthen academic medical education, teaching and learning will continue on a relatively uninformed basis, and our aim to produce safer, more efficient doctors will be under threat. This strengthening requires high level intervention to promote a culture change in medical schools, facilitated by strategies such as those listed in Box 2. Even if all these strategies were implemented, the impact on support for biomedical research would be small, yet the potential reward would be high. We would know much more about what we are doing in medical education, and medical schools could gain academic credibility not just through traditional indicators of success, but for doing what they were founded to do — to teach medical students well. After all, medical schools depend increasingly on their students’ fees. 1 Some current gaps in evidence in medical education Current basic medical education programs, including prior degree requirements for graduate entry, are 5, 6 or 7 years long. Which is better? Can selection procedures reduce the proportion of unhappy or unsuitable medical graduates? How much anatomy, chemistry, pharmacology, psychology and other basic and biomedical sciences do medical graduates really need to know? How can we measure the quality of teaching? What is the impact on learner performance of poor versus good teaching? Is it true that academic clinical groups provide better care? What is in the “black box” of small group learning that makes its cost worthwhile? What is good role modelling, and how can we facilitate it? To what extent can e-learning replace other modes of delivery? To what extent can skills laboratories replace real patient contact? How often should experienced doctors be reassessed for registration? Can poorly performing doctors be predicted or remediated? 2 Strategies to achieve better support for medical education Medical schools should be required to base curriculum development on medical education research wherever possible. Medical schools should be required to spend a certain proportion of their teaching income on education research that supports their program. Medical education units should be acknowledged as having research as well as service delivery roles. National research priorities should include issues beyond the scope of individual medical schools, such as identified gaps in evidence on medical education. Research funding bodies should establish panels for medical education research and allocate quarantined funds for proposals that address priority issues. Medical education research should be awarded its fair share of research fellowships at PhD and postdoctoral levels. Medical career structures should reward excellence in teaching and education research as much as other scholarly endeavours.

Richard B Hays PhD, MD, FRACGP, FACRRM

Education of our doctors needs to be a priority for the health services, not an add-on Training of doctors is expensive and takes a long time, and so we need to make sure all parts count. It had been recognised for some time that the PGY1 and PGY2 years, sitting between undergraduate and vocational training, were a lost opportunity, with no defined outcomes and marked variation in experience and supervision.1 Gaps in knowledge and skills were often likely after completion of this part of training. A significant number of junior medical officers (JMOs) feel unprepared to deal with, for example, medicolegal issues, emergencies or some procedures, which is unacceptable.2 Worryingly, they receive varying degrees of feedback and often feel inundated with administrative tasks. Many states had started working on defining outcomes for the PGY1 and PGY2 years, and, with leadership from the Confederation of Postgraduate Medical Education Councils and funding from the Medical Training Review Panel of the Australian Government Department of Health and Ageing, they collaborated to develop the recently launched national Australian Curriculum Framework for Junior Doctors (http://www.cpmec.org.au/curriculum). They drew on previously published frameworks from Canada3 and the United Kingdom,4 and the Australian National Patient Safety Education5 and Committee of Deans of Australian Medical Schools Indigenous Health Curriculum frameworks.6 The Framework documents key areas that we all know are important (clinical management, therapeutics), but also makes explicit areas that are usually minimally addressed, such as patient safety, communication and cultural safety. The consensus is a significant achievement in terms of agreement and collaboration across states. A curriculum can be viewed as a statement of the philosophy, content, learning methods and implementation of a course, which ideally would be linked with assessment. The Framework provides one part of a curriculum, namely the knowledge, skills and behaviours expected to be attained by PGY1 and PGY2 trainees, which they will carry forward into practice. However, will the Framework make any difference to the educational experience in those transitional years? This is a point of concern for JMOs, as discussed by Gleason et al in this issue.7 Two key aspects of training relate to how learning and assessment occur. Firstly, as regards learning, the curriculum framework working group recognised that learning will be progressive and opportunistic, and will occur in the work setting, and that effective supervision is important. Most learning still occurs in the work setting, and currently is unstructured and unlinked to any overall outcomes. The term “deliberate practice” has been used to define a model of learning characterised by good supervision and feedback, focusing on well defined tasks that can improve performance, and ensuring plenty of opportunity to practise these tasks.8,9 This requires a teaching program, guided by the Framework, which runs alongside clinical work, including simulation, tutorials, debriefing and online learning. However, this is not enough. A recent study on communication skills taught in a simulated setting showed that a program to transfer those skills back to the clinical setting (through close clinical supervision and feedback) is required for simulation training to have any effect on practice in real life.10 Assessment may be even harder. Although the JMOs are not keen for assessment to be used for registration,7 registration is a recognition of a level of competence and needs to involve some form of assessment. How can we be sure that doctors have reached a level of competence where they can be left to care for patients independently or are ready to move to the next stage of training? Currently, requirements vary from state to state, and may involve no more than completion of a form on overall impressions by supervising consultants in the PGY1 year. Although staff training has improved the amount of feedback JMOs receive, it is still lacking. The UK has moved down the pathway of having detailed determination of competence with multiple observed activities. Insufficient resources were provided for administration and training of staff, and this has led to both junior doctors and their supervisors developing innovative ways to subvert the system (Professor Janet Grant, Chair of the Curriculum Subcommittee, Postgraduate Medical Education and Training Board, UK. 11th National Prevocational Medical Education Forum, Adelaide 2006). Rather than moving straight to a complex system, at risk because of insufficient resources, we could continue to improve our current assessment of overall “global” competence of trainees, backed up by assessing essential areas in simulated settings (such as cardio-pulmonary resuscitation). When a trainee’s global competence appears inadequate, a second look, using assessments such as those used in the UK, could take place and more support be instigated. This may shift our assessment culture from punitive (blocking progress) to formative (helping progress), which presupposes that JMOs will all eventually be successful. JMOs may fear this aspect less. All of this is a big ask in an overburdened health system. Without resources, little is likely to change. We need to define and run simulations and tutorials. We need to train clinicians to facilitate learning and give feedback (professional development for clinical teachers). We need to train clinicians to recognise who is struggling. For both teachers and learners, we need to make sure there is enough time to do all this. The health services need to make education a priority, rather than an add-on.1,2 We agree that JMOs should be involved in developing and implementing an assessment plan. They also need to understand their responsibility in this partnership of learning. It is of interest that JMOs mostly request more tutorials and simulation to support training and think they learn little in the work setting. Learners need to be taught to recognise and make the most of learning opportunities, in part through reflection (teach the learners11). We cannot make all JMOs’ experiences uniform. It is unnecessary and impossible. However, the Framework allows administrators, clinicians and junior doctors in different states and clinical settings to consider the best ways to implement a better learning and assessment system for JMOs. Rather than lose the momentum, let us share resources and ideas nationally, and evaluate implementation through education research to make sure outcomes are achieved. We need to lobby as a group to make sure our clinical sites (including hospitals, community settings and new models of interdisciplinary care) are best structured and resourced to optimally train and assess our junior doctors.

Fiona R Lake MB BS, FRACP, MD · Lou Landau MB BS, FRACP, PhD

The current system of prevocational training does not meet the needs of junior doctors because of a high administrative workload, insufficient funding for education, and a lack of centralised guidance for trainees, teachers and hospitals. The Australian Curriculum Framework for Junior Doctors is designed to identify the training objectives for the prevocational years. The Framework has the potential to improve the quality of training of junior doctors, but this depends on how well it is implemented and resourced. It is imperative that any group responsible for implementing or assessing the Framework have a representative junior doctor, among others, on its decision-making committee. Stringent accreditation of training institutions is vital to the effective implementation of the Framework. The Framework should be used to promote teaching and learning, not as a barrier to vocational training or as a check-list to complete.

Andrew J Gleason MB BS(Hons), BSc · J Oliver Daly MB BS, BSc(Hons) · Ruth E Blackham MB BS

Why is so little known about what works in selecting medical students? In the past decade, some 15 000 students entered Australian medical schools, and in the United Kingdom, four times that number were admitted. Such a large number should imply that much is known about what to select on, how to select and whom to select. The sad reality is that surprisingly little is known. Instead, strongly held opinions are rife, inertia predominates, and change occurs more because of necessity, external pressure, political force or mere whim, than because of coherent evidence-based policy or theorising. Selection sometimes seems more to ensure the correct number of entrants on day one, than to identify those best suited to the course and profession. As if to illustrate the problem, the University of Adelaide recently reduced its emphasis on selection interviews, the University of Sydney extended its use of interviews, the University of Queensland may be ending interviews, and a meta-analysis in Medical Teacher suggested that selection interviews have only “modest” predictive validity and “little” or “limited” practical value.1 However, interviews differ in many ways, and although the meta-analysis found no moderating effect of factors such as interview method, structure, training, or scoring, some forms of interview may still be valid, as found outside of medicine,2 particularly for situational interviewing.3 An innovation in selection is written tests of “cognitive aptitude”, such as the Undergraduate Medicine and Health Sciences Admission Test and the Graduate Australian Medical School Admissions Test (GAMSAT) in Australia, and the BioMedical Admissions Test, GAMSAT and the UK Clinical Aptitude Test in the UK — tests whose enthusiastic adoption belies the absence of published evidence on predictive validity.4 In this issue of the Journal, an important article by Groves et al5 evaluates GAMSAT, used in nine Australian and four UK universities and which claims to measure “the mastery and use of concepts in basic science as well as the acquisition of more general skills in problem solving, critical thinking and writing”.6,7 The study by Groves et al is not large, but like another small study,8 addresses the predictive validity of GAMSAT, as well as further questioning the value of interviews. Cognitive aptitude tests4 measure either intellectual ability, general mental ability or intelligence (all effectively synonymous), as in GAMSAT’s first section, or academic achievement or attainment (typically with specific scientific content, as in GAMSAT’s third section, “Reasoning in Biological and Physical Sciences”). Although intellectual ability does predict outcome in primary and secondary schools, its prediction of university outcome is poor, tertiary-level performance depending more on motivation and specific knowledge. Consequently, achievement measures, such as A-levels in the UK, are more effective predictors of medical school outcome4,9,10 than are measures of intellectual ability, which predict little10 — a result known since the 1970s.11 For GAMSAT, Groves et al found a significant correlation with Year 2 examination results only for the biological and physical sciences section assessing achievement, and not for the sections assessing “more general skills in problem solving, critical thinking and writing”.5,6 It has to be acknowledged that even measures of achievement are not strong indicators of medical school outcome, with much variance still to be explained.9 A major problem for people developing and evaluating selection criteria is the lack of adequate outcome measures reflecting the broad and complex day-to-day behaviour of doctors. A rare and important exception is the study by Papadakis et al,12 who found professional misconduct was related both to unprofessional behaviour in medical school and to poor academic achievement before and during medical training. In medical school and beyond, most measures of competence assess knowledge, whereas being a competent, safe and effective doctor probably depends to an equal extent on behaviour, attitudes and approaches. Adequate knowledge is a necessary part of good medicine, but is far from sufficient for competent practice. Without robust, objective and valid measures of other professional skills to complement our robust measures of knowledge, any post-hoc assessment of selection criteria will be lopsided. The study by Groves et al is significant for including clinical reasoning skill as another outcome measure, but here there was a negative correlation with GAMSAT.5 Students from only two universities were included in the study by Groves et al, and the authors comment, “there is an urgent need for collaborative studies to explore these issues further”.5 Inadequate numbers of patients are the bane of clinical trials, so collaborations between hospitals are now standard when testing new drugs. Large-scale evaluations of medical education should be relatively easy, but few studies have been done. Randomised controlled trials (RCTs) should surely also have a place in assessing selection methods, but reasons are too easily found for them being impractical or unethical, despite being the ideal way to answer questions on the validity of novel selection procedures. If RCTs are ethical when assessing the effectiveness of drugs given to patients, they are surely also ethical for assessing the efficacy of tests used for selecting the doctors giving those drugs to patients. A more general issue raised by the very scarcity of studies like that of Groves et al is why, despite educationalists proliferating in medical schools, so little is known of what does or does not work in selection. The psychologist William James, more than a century ago, characterised his own nascent discipline as “A string of raw facts; a little gossip and wrangle about opinions; a little classification and generalization on the mere descriptive level [but] not a single proposition from which any consequence can causally be deduced”, concluding, “This is no science, it is only the hope of a science”.13 Recently, Schuwirth and van der Vleuten were equally critical of much medical education research, with its often “poorly performed or poorly reported” studies, problems that need solving “if the specialty wants to be taken seriously”.14 Selection encapsulates most of the important questions that medical education faces. The problems are not insoluble, and the intellectual tools mostly exist. The problem is in finding the collaborative will to solve them, and to confront negative answers when they arise.

Chris McManus · David Powis

Objective: To examine the relationship between medical school applicants’ performances in the Graduate Australian Medical School Admissions Test (GAMSAT) and structured interviews and their subsequent performance in medical school.Design: Students in Years 2–4 of two graduate-entry medical programs were invited to complete two previously validated tests of clinical reasoning. These results and their Year 2 examination results were compared with their previous performance in GAMSAT and at interview.Setting: The graduate-entry programs at the Universities of Queensland and Sydney.Participants: 189 student volunteers (13.6% response rate).Main outcome measures: Students’ test results on a set of Clinical Reasoning Problems (CRPs) and a Diagnostic Thinking Inventory (DTI) and their Year 2 examination results.Results: There was no association between performance in GAMSAT and performance in the CRPs; there was a weak negative correlation between performance in GAMSAT and the DTI (− 0.05 > r > − 0.31, P = 0.03). The correlation between GAMSAT and examination results was weak (r < 0.24, P = 0.02). The correlation between GAMSAT and interview scores for each school was weakly negative for University of Queensland (r = − 0.34, P < 0.01) and weakly positive for University of Sydney (r = 0.11), with a combined significance level P < 0.01.Conclusions: We did not find evidence that GAMSAT and structured interviews are good predictors of performance in medical school. Our study highlights a need for more rigorous evaluation of Australian medical school admissions tests.

Michele A Groves BSc, PhD · Jill Gordon MB BS, PhD · Greg Ryan RN, PhD

Objective: To identify whether the rate and average daily dose of stimulant prescribed for attention deficit hyperactivity disorder (ADHD) in Western Australia differed according to the geographical remoteness and socioeconomic status of the patient.Design and data sources: Secondary analysis of population-based administrative pharmacy data from 2004, stratified by the Accessibility/Remoteness Index of Australia (ARIA+) categories and the Index of Relative Socio-Economic Disadvantage (IRSD) quintiles for WA (2001 Census).Outcome measures: Rate ratios of stimulant prescription and mean average daily dose (in dex-equivalents) stratified by age (2–17, 18+ years), sex, ARIA+ category and IRSD quintile.Results: The rate of stimulant prescription was 2.3 to 5.3 times greater in major cities in WA compared with remote and very remote parts of the state. The association between socioeconomic disadvantage and the rate of stimulant prescription was highly variable. Adults with the least socioeconomic disadvantage were significantly more likely to receive stimulants compared with their most disadvantaged counterparts; however, the reverse association was seen with children. The average daily dose of stimulant prescribed did not vary greatly across remoteness or socioeconomic categories.Conclusion: Remoteness and socioeconomic disadvantage are significantly associated with rate of stimulant prescription for ADHD in WA, but not associated with average daily dose of stimulant prescribed. Further research is needed to understand why considerable variation exists in the use of prescribed stimulants for ADHD.

Janine Calver PhD · David Preen PhD · Max Bulsara MSc · Frank Sanfilippo PhD

Objective: To examine the frequency of dyslipidaemia and treatment with lipid-lowering drugs in patients with type 2 diabetes managed in Australian primary care.Design, setting and participants: The NEFRON study (National Evaluation of the Frequency of Renal Impairment cO-existing with NIDDM [Non-Insulin Dependent Diabetes Mellitus]) was an incident-driven, cluster-stratified survey of 3893 patients with type 2 diabetes from across Australian primary care between April and September 2005.Main outcome measures: The most recent fasting lipid levels were compared with therapeutic targets for lipid control and current prescribing guidelines.Results: 64% of patients with type 2 diabetes presenting in primary care received lipid-lowering medication. Despite the widespread use of statins (61%), 75% of patients had a total cholesterol level ≥ 4.0 mmol/L, and 47% had a low-density lipoprotein (LDL) cholesterol level ≥ 2.5 mmol/L. Few untreated patients met the Australian Pharmaceutical Benefits Scheme (PBS) criteria current at the time for subsidised primary prevention with lipid-lowering agents (4%). However, new PBS subsidy criteria will potentially include 93% of all diabetic patients seeing their general practitioner in primary care.Conclusion: Changes in the provision of subsidised therapy for high-risk diabetic patients are long overdue. However, more needs to be done to optimise management strategies, which still fail to achieve treatment targets in many treated patients.

Merlin C Thomas PhD, FRACP · Paul J Nestel MD, PhD

Objective: To estimate the incidence of diabetes and impaired fasting glucose (IFG), and increased risk associated with the metabolic syndrome, in a representative population-based sample of older Australians.Design, setting and participants: The Blue Mountains Eye Study examined 3654 residents aged 49 + years (82.4% response rate) during 1992–1994, and re-examined 2335 (75.1% of survivors) during 1997–1999 and 1952 (75.6% of survivors) during 2002–2004; 2123 participants with normal blood glucose levels at baseline were considered at risk of developing incident diabetes.Main outcome measures: Incident diabetes (or IFG) was defined in participants at risk who were newly diagnosed by a physician during the follow-up or found to have a fasting blood glucose level ≥ 7.0 mmol/L (or 5.6–6.9 mmol/L). Kaplan–Meier cumulative 10-year incidence was calculated.Results: The overall 10-year incidence of diabetes and IFG was 9.3% and 15.8%, respectively. Participants with metabolic syndrome at baseline had a higher risk of incident diabetes than those without metabolic syndrome (29.2% v 8.6%). Baseline factors associated with incident diabetes were elevated fasting glucose level (adjusted odds ratio [OR], 4.5; 95% CI, 3.4–6.1 per mmol/L), obesity (OR, 2.0; 95% CI, 1.3–2.8), diabetes family history (OR, 1.7; 95% CI, 1.2–2.5), current smoking (OR, 1.6; 95% CI, 1.0–2.7) and high density lipoprotein cholesterol level < 1.0 mmol/L (OR, 2.4; 95% CI, 1.5–3.8). Similar baseline factors were associated with incident IFG.Conclusion: This population-based study provides data on the incidence of diabetes and IFG in an older, predominantly white population, and confirms that metabolic and lifestyle factors are major risk factors for diabetes.

Sudha Cugati MB BS, MS · Jie Jin Wang MMed, PhD · Elena Rochtchina MApplStat · Paul Mitchell MD, PhD

Trends in meningococcal disease (MD) over the 20th century in Australia, as in other industrialised countries, have been characterised by epidemics during the two World Wars, a transient rise in incidence in the 1950s followed by endemic disease, and in the 1980s the emergence of a sustained hypersporadic phase. Epidemics occur at times of social upheaval and among marginalised populations, and resolve when living conditions improve. Periodic serogroup A epidemics have been replaced since the 1950s by endemic disease caused mainly by serogroups B and C meningococci. The current hypersporadic plateau in Australia, as in other industrialised countries, is associated with the intercontinental spread of hypervirulent clones of meningococci. The conjugate serogroup C vaccine has reduced the incidence of MD and carriage rates of serogroup C meningococci. However, the vaccine is expensive and its long-term impact on the emergence of non-vaccine strains and on nasopharyngeal microecology is unknown. A rising incidence of MD should not be viewed as the action of a virulent microbe exploiting a vulnerable population, but as the emergence of an “accidental pathogen” from an evolving host–microbial ecology. While it is essential to monitor the impact of vaccines on this ecology, we must find ways that can optimise our coexistence with microbes.

Mahomed S Patel MB BCh, FRACP, FAFPHM

Up to 45% of patients with debilitating and potentially lethal depressive illness do not achieve remission with initial drug treatment. Using combinations of antidepressants as an early option for treatment-resistant depression has become increasingly common. Before trying combination therapy, it is essential first to ensure diagnosis is correct, and then to optimise antidepressant monotherapy, using an effective dose for an adequate period. Subsequently, augmentation of antidepressants with lithium and triiodothyronine should be considered, as these strategies are strongly supported by numerous clinical trials. Electroconvulsive therapy is the most effective treatment for severe depression. There is little evidence to support use of antidepressant combinations. Risk of toxicity and drug interactions mandate that combinations be used as a last resort, and only in specialist settings.

Nicholas A Keks MB, PhD, FRANZCP · Graham D Burrows BSc, MD, FRANZCP · David L Copolov MB, PhD, FRANZCP · Richard Newton MB, MRCPsych, FRANZCP · Nick Paoletti MB, MPH, FRANZCP · Isaac Schweitzer MD, DPM, FRANZCP · John Tiller MD, BSc, FRANZCP

Over the past 10 years, the management of HIV infection has been transformed by an increased number of effective antiretrovirals (ARVs), with more convenient dosing and improved tolerability. Optimal management of HIV infection includes at least three effective ARVs; from at least two different drug classes. Current strategies and drugs can effectively control HIV and significantly reduce morbidity and mortality. However, no cure is yet possible. Appropriate use of ARVs leads to suppression of virological replication (to below the limit of detection using commercial assays to measure HIV in plasma) and an increase in CD4+ T cells with few adverse effects. Greater than 95% adherence to drug therapy is required for effective viral suppression and immunological improvement. Monotherapy, two-drug combinations, sequential ARVs, drug “cycling”, and treatment interruptions are ineffective management strategies and lead to earlier disease progression and emergence of drug resistance. Drug–drug interactions are common and caution is required when prescribing ARVs that inhibit or induce the cytochrome P450 pathway.

Luke F Chen MB BS(Hons) · Jennifer Hoy MB BS, FRACP · Sharon R Lewin MB BS(Hons), PhD, FRACP

From Australian coronial records, we identified five deaths involving implantable naltrexone between 2000 and 2004. One man died from acute narcotism with a naltrexone implant in place and a blood naltrexone level of 0.3 mg/L. A woman died of combined drug effect (including naltrexone) accompanied by severe pain from a naltrexone implant site. These cases indicate that patients can die from opioid overdose with a naltrexone implant and blood naltrexone levels higher than reported blockade levels. Clinical recordsTreatment with implanted naltrexone has been claimed to prevent relapse to opioid use1-3 and therefore fatal opioid overdose.4,5 We evaluated this claim by examining a series of deaths related to implanted naltrexone in the treatment of opioid dependence identified through Australia’s National Coroners Information System (NCIS). This research was approved by the University of New South Wales Human Research Ethics Committee and the Monash University National Centre for Coronial Information. The NCIS database provides access to all coronial cases in Australia, and is likely to capture most opioid-related deaths. Each database record is linked to available documents, including coronial finding, autopsy, toxicology and police reports. Naltrexone-related deaths were determined by searching for the keyword “naltrexone”. All closed cases were searched in December 2005 for deaths occurring between 2000 and 2004 inclusive. Ninety-one records were identified containing the keyword, of which five involved an implantable form of naltrexone (Box). The average age of the five people was 26 years (range, 22–32 years). Naltrexone was detected at autopsy in two and not tested for in the remaining three. All five deaths were related to the use of drugs. Two deaths occurred while undergoing naltrexone implant treatment, one shortly after naltrexone implant removal, and two about 6 months after insertion of an implant. In these last two cases, it is not known whether these people received an implant designed to last 6 months, or if shorter-acting implants were used. Both types are available in Australia.5 Blood concentrations of naltrexone were quantified in two cases. Male 1 died of acute narcotism with 0.3 mg/L total (0.2 mg/L free) naltrexone and 0.05 mg/L total (0.03 mg/L free) naltrexol detected. Heroin (as morphine and codeine) was the only other drug detected in toxic levels, and the autopsy report specifically highlights the danger of using excess heroin to overcome the blockade given by the active naltrexone implant. The only woman among the five people died of a combined drug effect, including naltrexone. Soon after the insertion of a naltrexone implant and 2 days before her death, she began to experience strong pain from the implant site (abdomen). She used methamphetamine the day before her death and was discovered by her carer to be delirious, in severe pain and clutching her abdomen on the morning of her death. Her blood levels of free naltrexone and naltrexol were 0.004 mg/L and 0.035 mg/L respectively, close to reported “therapeutic” levels.6 Naltrexone was viewed by the coroner as playing a causal role in the death. DiscussionNaltrexone is an opioid antagonist that acts to inhibit the effects of opioids,7 and is used for treating opioid dependence. Non-compliance with naltrexone is common, and is usually followed by relapse to heroin use,8 with a substantially increased risk of fatal overdose.9 Implanted forms of naltrexone have been advocated to reduce non-compliance.1,10 These devices are not currently registered in Australia, but can be accessed through the Therapeutic Goods Administration Special Access Scheme.2,4 In this series of deaths in people treated with naltrexone implants, we have identified a case of fatal opioid overdose with detectable post-mortem levels of naltrexone. The blood level of naltrexone (0.3 mg/L = 300 ng/mL) was several orders of magnitude higher than the 2.8 ng/mL previously reported to block 500 mg doses of pure diamorphine11 or the 2.4 ng/mL reported to completely antagonise 25 mg of heroin.12 Blood naltrexone levels of 1–2 ng/mL are commonly quoted as being sufficient to offer protection from fatal opioid overdose.6,13 The case of Male 2 highlights the risks of opioid overdose after chronic naltrexone treatment. The deaths of Males 3 and 4 by multiple drug toxicity 6 months after insertion of their naltrexone implants cannot be definitively linked to the naltrexone implant treatment, as it is not known if they were using long-acting (6-month) implants. In Male 4, the possible relationship between the naltrexone implant and the subject’s higher risk of opioid overdose death was noted by his doctor in the coronial finding. The likely under-detection of oral naltrexone-related death through the NCIS database has been previously reported,14 and naltrexone implant-related deaths might be even more under-reported. Naltrexone implants are highly likely to be overlooked at autopsy, unless the pathologist had prior knowledge of their presence (Associate Professor J Duflou, Chief Forensic Pathologist, New South Wales Department of Forensic Medicine, Sydney, personal communication). For this reason, this case study should not be used to estimate the naltrexone implant-related death rate in Australia. The clinical implications are clear: patients can die from an opioid overdose while undergoing naltrexone implant treatment with blood naltrexone levels higher than reported “blockade” levels. Although the risk of fatal overdose is probably reduced in naltrexone implant treatment, a risk still exists. Medical professionals have a duty to consider the risks of naltrexone implant treatment and to warn patients of the possibility of opioid overdose death during compliant naltrexone implant treatment as well as after treatment cessation. Details of deaths in people with naltrexone implants listed in the National Coroners Information System, 2000–2004 Case Cause of death* Substances detected post mortem† Comments Male 1 Acute narcotism Naltrexone 0.3 mg/L total (0.2 mg/L free), naltrexol 0.05 mg/L total (0.03 mg/L free), morphine 1.9 mg/L, codeine 0.2 mg/L, alprazolam < 0.1 mg/L, methamphetamine < 0.1 mg/L, benzoylecgonine 0.09 mg/L Autopsy reports the danger of using excess heroin to overcome naltrexone blockade. Had a current naltrexone implant (small scar and 10 naltrexone tablets in left groin). Female Combined drug effect (amphetamine, naltrexone, propranolol, doxepin, diazepam, paracetamol) Naltrexone 0.004 mg/L free, naltrexol 0.035 mg/L free, methamphetamine 1.2 mg/L, amphetamine 0.07 mg/L, propanolol 0.7 mg/L, doxepin 1.1 mg/L, desmethyldoxepin 0.7 mg/L Worsening stomach pain near recent naltrexone implant site for 2 days before death. Also had depression and had been prescribed numerous medications. Male 2 Multiple drug toxicity (heroin, diazepam) Naltrexone not tested, morphine 1.9 mg/L, codeine 0.2 mg/L, diazepam 0.3 mg/L, nordiazepam 0.5 mg/L, amisulpride 0.5 mg/L Naltrexone implant was removed 2 weeks before death for unknown reasons. Male 3 Mixed drug toxicity (heroin, codeine, diazepam) Naltrexone not tested, morphine > 15 mg/L (urine), codeine 11.2 mg/L (urine), diazepam trace, nordiazepam 0.1 mg/L Naltrexone implant 6 months previously.‡ Male 4 Combined drug toxicity (heroin, alcohol and diazepam) Naltrexone not tested, ethanol 0.04 g/100 mL, morphine 0.2 mg/L free, codeine 0.06 mg/L free, diazepam 0.4 mg/L, nordiazepam 0.5 mg/L Naltrexone implant approximately 6 months earlier with scar.‡ Doctor who inserted implant noted that after implant his opioid tolerance would have been low and the risk of opioid overdose high. * Causes of death were recorded directly from findings or autopsy documents. † Substances were detected in blood unless otherwise specified. ‡ Some naltrexone implants available in Australia are claimed to provide “effective” blood naltrexone levels up to 6 months after implanting.4 Without reported blood naltrexone levels we cannot be certain of the duration of action of these implants, as shorter-acting implants may have been used.

Amy E Gibson MPH, BSc(Hons), BA · Louisa J Degenhardt PhD, MPsychol(Clinical), BA(Hons) · Wayne D Hall PhD, BSc

To the Editor: The National Health and Medical Research Council’s National statement on ethical conduct in research involving humans1 recognises that prisoners can participate in research, but categorises them as “persons in dependent or unequal relationships”. They have limited capacity to provide informed consent. Responding to the high levels of transmission of bloodborne viruses in Australian prisons,2 the Australian Red Cross Blood Service excludes prisoners from donating blood and ex-prisoners are excluded for 12 months after they have been released from prison.3 The New South Wales Human Tissue Act 1983 is silent on whether prisoners can donate organs. We report here the case of a prisoner organ donor, highlighting the administrative, legal and operational hurdles that needed to be overcome. A 53-year-old male prisoner was a suitable living kidney donor for his first cousin. He provided consent willingly and without coercion. At initial assessment, the prisoner’s classification required that he be escorted to hospital and that constant surveillance by prison officers be maintained — at a cost of $1000 per day, for at least 7 days. These costs would have been borne by the family. Furthermore, as Australian prisoners are ineligible for Medicare under the Australian Constitution, the donor, as an uninsured patient, and his family would have been required to pay for all pre-, peri- and postoperative care. The donation was deferred for 14 months while these two administrative hurdles were overcome to permit the donation to proceed: 1. The Commissioner for the Department of Corrective Services gave approval for the prisoner to be reclassified to the lowest security classification, thus removing the need for surveillance while in hospital; and 2. A rarely used provision within the NSW Crimes (Administration of Sentences) Act 1999 was applied. Section 26(1) of the Act allows the Commissioner to issue a permit allowing an inmate to be absent from a correctional centre: (a) on such conditions and for such period as may be specified in the permit, and (b) for such purpose as the Commissioner considers appropriate. This allowed the prisoner to be temporarily reinstated to receive Medicare entitlements. The nephrectomy and transplantation were successfully performed. The donor returned to prison on the seventh postoperative day. The donor organ is functioning 4 months after the operation. Prisoners have a right to participate in organ donor programs; however, their precarious position to provide informed consent needs to be protected.

Elizabeth Magee · Michael H Levy

To the Editor: An important factor in the well documented shortfall of organs and eyes for transplantation is the apparent reluctance of people to agree to donate.1 One nearly universal strategy in attempting to raise donation rates has been to encourage individuals to register their wishes about donation. Although evidence that this strategy increases donation rates is lacking, there is some evidence that more individuals make and communicate a decision with appropriate education.2 Most families consent to donation when the deceased had indicated this was their wish, and virtually none override a stated wish not to donate.3 When wishes are unknown, half of families consent and half refuse.3 Encouraging declaration of intention aims to increase the rate of consent for families who would otherwise not know the deceased individual’s wishes. For this to be successful, most individuals newly recording their wishes must indicate a desire to donate. This assumption has underpinned Australian education campaigns, including “Talk about it”, “Share your life, share your decision”, and most recently the national “Sign on to save a life” campaign.4 A simple review of New South Wales Roads and Traffic Authority organ donation data over the period of these campaigns suggests this assumption may not hold. From 1997 to 2004, a significant proportion of drivers licence holders newly indicated a preference about donation; the proportion indicating some decision rose from 59.4% to 78.6%. Over the same period, the proportion indicating yes to donation of all organs rose from 35.6% to 41.9% (a 17.7% increase); however, the proportion indicating no to any donation rose from 19.9% to 31.4% (a 57.8% increase).5 These results raise the possibility that encouraging individuals to make a decision about donation may increase the number of families who refuse donation. Individuals who had previously not made a decision about donation, when encouraged to do so, displayed an unwillingness to become organ donors at twice the rate of those who indicated willingness. Although it is imperative to recognise and respect the decision of individuals to refuse organ donation, this unwillingness may reflect either formalisation of a considered desire not to donate, or a decision made without personal discussion of fears and concerns about donation. Generalised education campaigns are limited in that they encourage action without addressing fears and concerns. Further policy should recognise a possible danger in simply exhorting the public to make a decision, and research should investigate why individuals are refusing to become organ and eye donors.

Mitchell Lawlor · Frank A Billson

To the Editor: Opdam and Silvester concluded that the small pool of organ donors limits the potential for organ donation in Victoria.1 According to the definition of death in the Human Tissue Act 1983 (NSW): a person has died when there has occurred: (a) irreversible cessation of all function of the person’s brain, or (b) irreversible cessation of circulation of blood in the person’s body. Current cadaveric organ donation takes place predominantly after brain death, although it is also possible after cardiac death. This was previously described as “non-heart-beating organ donation”, but the name was changed to “donation after cardiac death” (DCD) to emphasise that organ donation occurs only after death. DCD can be classified by the Maastricht criteria (Box).2 Category III is relevant in select patients undergoing planned withdrawal of therapy in intensive care units (ICUs). Planned withdrawal of therapy is said to occur in about 60% of all ICU deaths.3 Permission for organ donation is required before the planned withdrawal of life support, so the donor organ retrieval teams can be available and ready to retrieve organs soon after a 5-minute “cooling off” period after circulation ceases. These 5 minutes also allow families to spend time with the deceased after death. A period of less than 60 minutes between withdrawal of therapy and cessation of circulation is recommended to minimise warm ischaemia time in the retrieved organs. Withdrawal of life support may occur in the ICU or in the operating room complex, and local hospital guidelines should address the issue of where this occurs. Between 1989 and 2004, 30 Australian organ donations involved non-heart-beating organ donors (data from the Australia New Zealand Organ Donor Registry). In 2005, eight New South Wales organ donations, accounting for 14% of all cadaveric renal transplantation, proceeded after cardiac death criteria were applied. In our unpublished retrospective audit, 7% of patients who had planned withdrawal of therapy in the ICU met the criteria for eligible organ donors after cardiac death. The organ donor pool for organ donation after brain death was 1.7% in Opdam and Silvester’s study.1 The long-term function of transplanted kidneys is not significantly different for organs retrieved after cardiac death compared with organs retrieved after brain death.4 Liver and lung transplantation are also possible from organs retrieved from non-heart-beating organ donors. Draft guidelines for DCD in NSW are soon to be released by NSW Health. Similar guidelines in other states and education of staff involved in organ donation would no doubt increase the organ donation pool and the potential for organ donation. Maastricht donation after cardiac death protocol categories2 I Dead on arrival (uncontrolled) II Failed resuscitation (uncontrolled) III Withdrawal of support (controlled) IV Arrest following brain death (uncontrolled)

Deepak Bhonagiri · Patricia Wills

To the Editor: In 2004, there were 580 cases of suicide in Queensland, and not 453, as reported by the Australian Bureau of Statistics (ABS) on 14 March 2006.1 These data alone reverse the declining trend for suicide mortality nationally in the most recent years. The Queensland Suicide Register, maintained by the Australian Institute for Suicide Research and Prevention (AISRAP), receives data directly from the Office of the State Coroner, and crosschecks them with other Queensland coroners, the John Tonge Centre (the Queensland Health Scientific Services mortuary), and the National Coroners Information System (NCIS). The ABS receives data from the state registries of births, deaths and marriages, and crosschecks them with the state coroners’ offices. The agreement between the two agencies has been decreasing in recent years, with AISRAP detecting 550 suicide cases in 2003 and 588 in 2002, compared with 466 and 537, respectively, detected by the ABS (Box). The ABS has acknowledged difficulties in getting reliable data for 2004 in a number of endnotes to its yearly report.1 Most of the problems were related to a very large backlog of cases still under investigation by coroners, a phenomenon that is reported as increasing in recent years. A confirmation of problems in official data comes from the NCIS, whose most recent report has evidenced, from 2000 on, declining percentages of completeness in mortality data in Queensland and elsewhere in Australia, with the most incomplete figures in 2004.2 It is important to note that cases that are under investigation and those that end with an open verdict would not enter official suicide mortality data, as these are never reconciled. Following the example of many European countries, it would be desirable to start a periodical publication (eg, every 3–5 years) to provide a more comprehensive picture of suicide mortality, including finalised investigations, reclassified (ex-accidental or ex-undetermined) causes of deaths, and deaths that occurred (especially in hospitals) with a delay from a self-injurious event. This would provide a more credible depiction of suicide mortality in the country, and permit better research.3 Meanwhile, efforts should be made to homogenise certification procedures (International classification of diseases, 10th revision terminology has yet to be extensively adopted) and streamline the bureaucratic procedures (we still suffer from a number of “lost in the system” data). In the registries of births, deaths and marriages, by law, the word “suicide” (or analogous term) does not appear. Frequently, the ABS, which collects data from the registries each month, has to reclassify the data obtained, and integrate the information received with further enquiries. Apart from being time-consuming, this routine does not provide foolproof results, and has potential for improvement. Some underreporting in suicide statistics is virtually ubiquitous,3,4 and has to be tolerated (eg, misclassification as accident, road accident, or disease-related, particularly in the elderly; cover-up because of stigma, sociocultural norms, or insurance reasons; or remoteness of location). However, federal and state governments in Australia are committed to suicide prevention plans that require credible baselines for evaluating their effects. All relevant parties need to work jointly on improving data quality. This is of crucial importance for scientists and policymakers, and for those personally affected by a suicide death. Number of suicides in Queensland according to the Australian Bureau of Statistics and the Queensland Suicide Register (QSR) QSR data were provided by the Office of the State Coroner, the National Coroners Information System and the John Tonge Centre. “Possible” suicides are not included. Data for 2005 are an estimate.

Diego De Leo

To the Editor: We read with interest the “Diagnostic Dilemma” by Peter et al.1 The case raises interesting management issues. The first is initiation of antibiotics. Despite 1 week of fever, rigors, haematuria and loin pain, we are informed that the patient was in no distress at initial assessment. In this situation there is, despite the anxieties of resident staff, no urgent need to administer antibiotics; hospitals are controlled, monitored environments in which observation, review and investigation can be undertaken, within reason, if a diagnosis is not immediately made. The second issue is antibiotic selection. The provisional diagnosis was a urinary tract infection, and ceftriaxone and gentamicin were administered. The justification for the use of two agents with a similar spectrum of antimicrobial activity is not given.2 Likewise, no justification is given for the use of a potent nephrotoxin in the presence of moderately severe acute renal failure. Flucloxacillin was added “to broaden the gram-positive antibiotic cover”. It is not apparent why staphylococcal cover was sought at this stage. All cultures (blood, urine and pleural fluid) remained negative. At Day 14, ceftriaxone and gentamicin were changed to ticarcillin/clavulanic acid and ciprofloxacin “because of persistent fever and rising [white cell count]”; this decision in the absence of positive cultures is not explained. The patient’s renal function deteriorated further and he became profoundly acidotic. In fact, the patient’s renal function had performed heroically, given administration of gentamicin for 2 weeks in the presence of acute renal failure at admission. In the intensive care unit, flucloxacillin was replaced with vancomycin; the rationale is not explained. This case illustrates important points regarding antibiotic use. Despite significant renal impairment at admission, the patient was administered a 2-week course of a nephrotoxic antibiotic, which contributed to renal collapse. This situation would have been terminal if not for supportive intensive care. The treating team appears to have managed the patient as if sepsis were a given, and yet all cultures remained negative. This illustrates a basic but crucial teaching point — fevers, chills, rigors, raised inflammatory markers and neutrophilia do not necessarily equate with sepsis. If this experience reflects routine practice elsewhere (and it is our experience that it does), is it any wonder that we have reached an era in which we now encounter organisms so resistant that they are essentially untreatable?3

Mark A Boyd · Stephen Hedger

In reply: Boyd and Hedger have raised concerns regarding the initiation and choice of antibiotics in our recent case report.1 Several aspects of this correspondence need to be addressed. The primary focus of the article was to highlight an important and probably underrecognised cause of unexplained metabolic acidosis, and discussion regarding antibiotic choice was not within the scope of the article. Further, the patient’s management before admission to the intensive care unit was by a different treating team. Subsequent case-note review did not reveal reasons for initiation or choice of antibiotics other than described in our article, although it was evident that gentamicin doses were adjusted based on drug levels. We agree that a less nephrotoxic agent could have been chosen and that the profligate use of antibiotics in the absence of strong evidence of infection could have been avoided. The excessive use of antibiotics in the current medical milieu may stem from a physician’s lack of confidence, or even legal ramifications of “watching and waiting” in the setting of “fevers, chills, rigors, raised inflammatory markers and neutrophilia”, as encountered in our patient.

John V Peter · Natasha Rogers · Sandra L Peake

To the Editor: Primary care skin cancer clinics continue to receive negative publicity. We have previously reported on the workload profile of one network of clinics.1 We report here the profile of clinical activity of the four MoleScan skin cancer clinics situated on the Sunshine Coast, Queensland. Between them, these clinics have been open for a total of 22 years, ranging from 2 years to nearly 9 years of operation. MoleScan is a service company with clinics across Australia. Doctors are employed as subcontractors and are provided with digital dermascopes. The clinics do not have dedicated day surgery facilities, and surgical procedures are conducted in the consulting rooms (http://www.molescan.com.au). Using Medicare Benefits Schedule item number billing data (as previously reported1), we calculated the number of consultations, biopsies, excised lesions (benign, non-melanoma skin cancers [NMSCs] and melanoma), surgical repairs, non-surgical treatment of skin cancers, and non-surgical treatment of other skin lesions. We also estimated the number needed to treat (NNT), defined as the number of benign lesions removed per melanoma. There were 98 276 consultations at the four clinics during the 22 years of operation (Box). In all, 14 982 skin cancers were treated: 395 melanomas and 7468 NMSCs by surgical excision, and 7119 NMSCs by non-surgical methods. The estimated NNT was 22.5. Of the 16 962 lesions excised, 11% (1812) were repaired by a skin flap, 68% (1226) of which were simple flaps. Our previous report, on a different network of clinics,1 showed a different pattern of surgical repairs: 33% (2651) of the 8055 lesions excised were repaired by a skin flap, 45% (1187) of which were simple flaps. Clearly, the clinical practices of these two clinic networks vary. Another area of apparent difference between the two clinic groups is the NNT — 22.5 reported here, compared with 28.6 from the other network.1 The lower NNT in these MoleScan clinics may result from the use of digital dermoscopy, but this requires further study. These early findings from our analyses of MoleScan data highlight the dangers of generalising about the activities of primary care skin cancer clinics from one dataset. Workload profiles of different clinical services may vary markedly, and the widely expressed concern about large numbers of inappropriate surgical repairs may not be warranted. Activities billed at four MoleScan skin cancer clinics on the Sunshine Coast, Queensland, over a total of 22 years of operation * Eyelid, nose, ear, lip, neck, hand, digit or genitals.

Deborah A Askew · David Wilkinson · Gordon L Patrick

To the Editor: We read with interest the article by Couper et al on the psychosocial impact of prostate cancer (PCA) on patients and their female partners.1 We agree that involvement of partners in the research pro-cess is pivotal to understanding the relational dimension of how PCA is both understood and approached by men and their partners.2 However, this could be extended to consider the unique experiences of gay men diagnosed with PCA and their partners. Heteronormative viewpoints are commonplace in PCA research. This bias is unfortunate, as there is a 28% possibility that one of the men in a gay relationship will develop PCA over the course of his lifetime.3 While it is probable that gay men and their partners have some of the same concerns regarding PCA as heterosexual couples, there are also unique concerns that are specific to gay men and their partners. Such considerations may include (but are not limited to): the prostate gland as a site of sexual pleasure and the associated implications of being able to engage in penetrative anal sex after prostate surgery;4 homophobia and/or disregard for sexuality within the health care system5 when being diagnosed with and treated for PCA; and the impact of polygamous (open) relationships and the ambiguous position of gay partners having to care for their mates.4 At this stage, the above concerns are purely speculative, as there is a paucity of literature on gay men and PCA.4,6 We believe future research on the psychosocial impacts of PCA should consider the experiences and special concerns of gay men with PCA, as well as those of their partners.

James A Smith · Shaun M Filiault · Murray J Drummond · Robert J Knapman

In reply: We thank Smith et al for their acknowledgement of the importance of the relational dimension to understanding how prostate cancer (PCA) is understood and approached by men and their partners. The psychosocial implications of prostate cancer for same-sex couples are important and need specific investigation. However, there are methodological difficulties in attempting to quantify the impact of PCA on same-sex couples and in comparing their experience with other couples. In our review of the literature,1 we discovered that where previous researchers had included same-sex partners in their studies, insufficient numbers were recruited for meaningful quantitative statistical comparisons. For example, Perez and colleagues,2 Neese and colleagues3 and Davison and colleagues4 each recruited only one same-sex couple into their studies of 134, 164 and 74 couples, respectively. We believe that a qualitative approach is needed, specifically seeking out and examining the experiences of a group of same-sex couples and comparing and contrasting their experiences with those of a group of male–female couples. This is an approach we are considering in future studies to help us develop and refine an effective but broadly applicable couple-focused psychosocial intervention for PCA.

Jeremy W Couper

Custom cancer treatment Creating a genetic profile of non-small-cell lung cancers may help predict prognosis and survival for individual patients, a Taiwanese study suggests.1 Researchers selected five genes as a signature of individual cancer specimens from 125 patients with this cancer, and found that this signature was an independent predictor of relapse-free and overall survival. The results of this study may lead to a refinement in the treatment of lung cancer, with only those patients having the high-risk gene signature being subjected to Cisplatin-based adjuvant chemotherapy. An accompanying editorial predicts that cancer genomics will expand into prospective trials designed to explore the response to standard and novel therapies of patients selected based on their molecular signature.2 1. N Engl J Med 2007; 356: 11-20 2. N Engl J Med 2007; 356: 76-78 Mad cows and Englishmen It may be possible to remove the agent causing transmissible spongiform encephalopathies from donated blood using an adsorptive resin, according to a combined UK and US study. Researchers inoculated leucoreduced, scrapie-infected blood into the brains of hamsters and determined the rate of infection before and after pre-treatment with affinity ligands fixed on a chromatographic resin. Leucoreduction is a process in which leucocyte numbers are significantly reduced prior to transfusion. The results of using the resins were compared with leucoreduction alone. Passage through the resins resulted in reducing blood-associated infectivity to below the limit of detection. The authors comment that, given the estimated number of individuals infected with variant Creutzfeldt–Jakob disease in the UK is around 4000, the method may prove invaluable in reducing the risk of transmission through blood transfusion. Lancet 2006; 368: 2226-2230 I don’t like Mondays Mondays are the most dangerous days to be at work, according to an Australian occupational health study. Researchers analysed injury data from the Australian Bureau of Statistics over a 20-year period and found that compensation claims were not evenly distributed throughout the working week. Most injuries occurred on Mondays, followed in decreasing order by the consecutive days of the week. More injuries were reported in the mornings than in the afternoons for every working day. The study supports the findings of international research showing similar trends in occupational injuries sustained over the course of the working week. Aust N Z J Public Health 2006; 30: 505-508 We are the champions Self-management can be more successful than standard care in the treatment of men with lower urinary tract symptoms, a British randomised controlled trial has revealed.1 Researchers placed 140 men with uncomplicated urinary tract symptoms into two groups, one of which received standard medical care based on initial watchful waiting, with escalation to medical treatment and surgery left to the discretion of the clinician and patient. In addition to standard care, the intervention group took part in sessions run by urology nurse specialists aimed at bringing about lifestyle modification and behavioural changes, such as fluid management and bladder retraining. Patients were followed up by clinicians at 3, 6 and 12 months. Treatment failure was defined as: a rise of three points or more on the international prostate symptom score; use of drugs to control lower urinary tract symptoms; occurrence of acute urinary retention; or surgical intervention. Failure was significantly more frequent in the group randomised to standard care alone. Those in the self-management arm had less severe symptoms at all stages of follow-up. Although an accompanying commentary suggested interpreting the results with caution until larger trials are completed, the results were considered promising in terms of improving patient quality of life and reducing the financial burden on health care systems.2 1. BMJ 2007; 334: 25 2. BMJ 2007; 334: 2 Straight to the heart Pexelizumab given to high-risk patients as an adjunct to coronary reperfusion does not improve mortality at 30 days, a large, multicentre, placebo-controlled study has shown. Pexelizumab, a humanised, monoclonal antibody that binds to the C5 component of complement, has previously been shown to favourably affect outcomes in myocardial infarction when given as an intravenous infusion. Over 5000 patients from 17 countries were enrolled in the trial. All patients had suspected acute ST-elevation myocardial infarction and were about to undergo percutaneous transluminal coronary intervention (PCI). One group were given pexelizumab prior to PCI, and the other a placebo. The primary end point was mortality at 30 days, and secondary end points included death through 90 days, cardiogenic shock and congestive cardiac failure. No difference in mortality or the incidence of heart failure and cardiogenic shock was observed between the two groups. JAMA 2007; 297: 43-51 Dr Tanya Grassi, MJA

Ann Gregory