Volume 185 Issue 5

We can learn from the US Breakthrough Collaborative, which increased organ donation rates by 20% in 3 years Organ donation saves lives. This trite but true slogan applies to all the solid organs that are currently transplanted. Even in patients with kidney failure, where dialysis provides long-term treatment, the mortality rate is reduced significantly in those who receive transplants compared with those remaining on the waiting list.1 Improved survival, improved quality of life and reduced economic costs of care provide an unusual coincidence of benefits from this therapy. For example, with respect to cost, the recently published Economic impact of end-stage kidney disease in Australia2 contains a careful and conservative assessment of the direct costs of kidney transplantation (excluding costs from comorbidities, transport, loss of income and other incidentals). It reports that the cost of the first year after both live and deceased donor transplants is $62 375, with additional costs related to the donor of $8178 and $3000, respectively. The subsequent annual cost is $10 749 for both types of recipients. These costs compare with an annual dialysis cost of $48 631 for satellite-based patients, $56 828 for peritoneal dialysis, and $82 764 for hospital-based patients (excluding costs from comorbidities, transport, loss of income and other incidentals). Consequently, the cost advantage of transplantation is clear within 1–2 years. However, organ donation and transplantation continue to provoke ethical, legal and clinical debate with respect to procurement of organs from both living and deceased donors. The shortage of organs for transplantation in Australia is a key problem that has been well documented in recent times; it was the subject of an editorial in this Journal just over a year ago,3 responding to a report in the same issue of the adverse impact on Australians arranging kidney transplantation overseas.4 The variability of donor procurement rates between major hospitals was highlighted as demonstrating the potential for improvement. Since that time, the national Australian organ donation rate has sadly remained static at 10 donors per million population. Now, in this issue of the Journal, Opdam and Silvester address the process of organ procurement in hospitals in Victoria and underline the potential for increasing organ donation through change in hospital practice (page 250).5 They describe a carefully performed audit of Victorian hospitals, which assessed the potential for increasing the number of organ donors and confirmed the interhospital variation in donor performance in the Melbourne area. The main issues identified by Opdam and Silvester as opportunities for improving the donation rate were an increase in the consent rate and an increase in the identification of potential donors. Their study found that “there is limited but real potential to increase the number of organ donors in Victoria”. They state that “a maximum practically achievable organ donation rate for Victoria was estimated to be 15 to 17 donors per million population”. This “practically achievable rate” compares to a mean actual donation rate for Victoria of 9.2 donors per million population over the past 5 years. The economic impact alone of an increase of 50% in the donor procurement rate applied nationwide, as conservatively modelled in the report Economic impact of end-stage kidney disease in Australia,2 would be to save $26 million over the next 10 years (apart from an arguably even more important significant improvement in quality of life). There is no reason to doubt that the findings of Opdam and Silvester would apply in all states of Australia — they have been replicated (albeit in unpublished studies) in other states. How then does Australia break the present pattern and lift its donation rate to an internationally accepted benchmark of 15–20 donors per million population? Over the past few years, we have looked to the Spanish experience, but we should perhaps also pay attention to recent activities in the United States, where the organ donation rate has increased significantly. Over about the past 30 months, there has been a sustained rise in numbers, from a national mean of 508 donors per month (before October 2003) to a current mean of 614 donors per month — an outstanding increase of 20% on top of an already respectable rate of 24 donors per million population. Even allowing for the difference in rates of traumatic death between Australia and the US that might affect the number of potential donors, this performance is impressive. The explanation for the improved performance in US donor procurement appears to lie in a sustained strategy, starting in 1997 and culminating in the Organ Donation Breakthrough Collaborative, announced as a national program in April 2003 and commenced in October 2003.6,7 The details of the Collaborative approach have been recently described and can be summarised as an intensive effort to facilitate “breakthrough transformations” in the performance of organisations working together to achieve common goals, based on strategies shown to work in highly successful organisations. A key element is to identify best practices associated with higher donation rates in lead organisations, with a particular focus on early identification of potential donors and the consenting process — factors already identified by Opdam and Silvester as pertinent locally. The methodology of the Collaborative begins with selecting a planning committee of expert teams from hospitals interested in participating, and developing strategies to achieve the target (a “change package”). These strategies are then tested, modified if necessary and implemented. The strategies adopted are guided by what has been shown to work elsewhere and are always multiple in their focus (see Box). The announcement by the federal Minister for Health and Ageing, Tony Abbott, in February this year, that up to 20 of Australia’s major hospitals would be invited to participate in a similar Collaborative, with the aim of increasing organ and tissue donation rates, is welcomed.8 This project, facilitated by Australians Donate (the peak national body for organ and tissue donation for transplantation in Australia), commenced in May 2006 with the aim of building on the US experience. The potential for increase in Australia’s organ donation rate is substantial, as evidenced in the article by Opdam and Silvester. The expected outcome of the Collaborative project is a sustainable increase in the donation rate emanating from more effective early detection of potential donors and a higher conversion rate of potential to actual donors, thus realising the wishes of most Australians to be donors after death.9 There seems every reason to believe that many of the critical “success” factors in the US experience — engagement and commitment of the whole institution, a strong emphasis on support and teamwork of those in the frontline of the process, and a particular focus on an effective consenting process — will also be successful in Australia. The US Organ Donation Breakthrough Collaborative6 Overarching principles Unrelenting focus on change, improvement and results Rapid, early referral and linkage Management of an integrated donation process Aggressive pursuit of every donation opportunity First things/first changes Create a hospital presence or in-house coordinator for the organ procurement organisation Analyse and apply current hospital-specific data Identify a physician or clinician “champion” Conduct monthly reviews of death records Establish clinical triggers for referrals Hold “huddles” for the donation team Identify effective donation “requesters” Conduct after-action reviews Strategies established as successful (“high-leverage” changes) Advocate organ donation as the mission Involve senior leadership to get results Deploy a self-organising team of staff from the organ donor agency and hospital Practice early referral and rapid response Master effective requesting Implement donation after cardiac death

Timothy H Mathew MRACP, FRACP · Jeremy R Chapman MD, FRACP, FRCP

The new National Health and Medical Research Council toolkit is designed to help Effective communication between health care consumers and professionals is fundamental for evidence-based clinical practice.1 Clinicians and consumers recognise how difficult it is to exchange even apparently simple information with each other, while difficulties with more complex information are usually obvious to either or both. With increasing expectations of health care consumers that they will share more in decision making, and with greater access to information of variable quality outside of consultations, including the media and Internet, effective communication during consultations becomes even more necessary than before. One consequence of poor communication between consumers and clinicians is limited understanding by consumers of the benefits and risks of tests and treatments. This can result in decisions that might have been different with better understanding. For example, fewer men are interested in prostate-specific antigen (PSA) testing, and fewer women in adjuvant therapy for breast cancer after receiving more detailed or applicable information about these interventions.2,3 Other strategies that enhance communication have been shown to lead to improvements in understanding. A systematic review found that communication tools in most formats (eg, verbal, written, illustrative diagrams, cartoons and graphs, video, provider-delivered, computer-based) can increase patients’ understanding, especially if they are structured, tailored and/or interactive.4 The style in which data are presented also influences comprehension: “framing” numerical data about benefits and risks of interventions in either negative or positive and relative or absolute ways influences how data are perceived, not only by consumers but by clinicians as well. Addressing these issues has been shown to increase both consumer satisfaction and involvement in decision making, to result in more realistic expectations of outcomes, and to reduce decisional conflicts.4 Better communication contributes to consumer satisfaction in other ways, and with better outcomes. For example, agreement between physicians and patients about diagnosis and management of back pain is not only associated with higher patient satisfaction, but with better health status outcomes as well,5 suggesting that communication which promotes clinician–patient agreement leads to better outcomes. However, the extent to which poor communication contributes to clinician dissatisfaction, and also to other unfavourable outcomes, such as inappropriate underuse and overuse of investigations and treatments, and to health care costs, is unclear at present. There are many barriers to effective communication between health care consumers and professionals. They include physical impairments such as hearing and vision loss; learning disabilities; differences in languages spoken, and in cultural, educational, religious and socioeconomic backgrounds of the clinician and consumer; anxiety; and problems associated with illness and poor literacy. As consumers become more able to access information from an increasing array of sources, the skills of reading, writing and numeracy assume even greater importance. Patients with lower literacy skills are more likely to have problems communicating with health care professionals and to have less understanding about their medical conditions and choices.6 Not all clinicians have the skills to communicate well with a broad range of consumers with differing needs, or to involve consumers in making decisions when consumers want this. In addition, many consumers are not accustomed to having their views heard, or sharing in decisions about their care.7 Implementing better communication also requires negotiation and agreement about the particular arrangements for decision making within specific consumer–professional partnerships.8 The National Health and Medical Research Council has extended and applied the recommendations from two earlier publications to develop a toolkit, based on the best available evidence, to help consumers and clinicians learn the principles for better communication.9 The toolkit highlights common communication problems and provides practical strategies, from both consumer and provider perspectives, for recognising, exploring and solving these problems. The toolkit aims to help both health care consumers and professionals expand their communication skills and use them more effectively in a variety of clinical encounters. The toolkit recognises five principles for effective communication (Box). Each principle is presented in an identical and structured format: background; what the principle means for both health care consumers and professionals; and examples, with tools to put the principles into practice. To apply these five principles, illustrative scenarios, with questions for health care consumers and professionals, and sources of relevant additional information are included. Communicating benefits and risks of interventions between health care consumers and professionals is complex, to say the least. While there is a growing body of research evaluating ways to improve communication between health care consumers and professionals,10,11 we have only just begun to integrate this information into routine clinical practice. Improved communication tools and more effective techniques to overcome the barriers to better communication are likely to result in greater mutual understanding, better decisions and better health outcomes. However, as there is often a long period between communication during consultations and tangible health outcomes, and as the effects on health outcomes may also reflect other interventions and changes, it is easier to establish the effectiveness of communication strategies on shorter-term processes like knowledge, understanding and satisfaction. More research is needed, not only to prove whether communication strategies lead to better health outcomes, but also whether such strategies can be effectively taught and learned. Five principles for effective communication Good communication between health care consumers and health care professionals has many benefits. Health care consumers vary in how much participation in decision making they desire. Good communication depends on recognising and meeting the needs of health care consumers. Perceptions of risks and benefits are complex, and health care consumers and health care professionals may have different priorities. Information on risks and benefits needs to be comprehensive and accessible.

Peter B Greenberg MD, PhD, FRACP · Christine Walker PhD · Rachelle Buchbinder MB BS, MSc, FRACP

Many disparate groups in Australia now concur about the need for continuous food and nutrition monitoring Poor nutrition contributes to Australia’s current health problems in several ways. Heart disease and cancer, both strongly related to nutrition, remain the leading causes of death. At the same time, the prevalence of obesity and diabetes is alarmingly high, and deficiencies of vitamin D, iodine, and calcium are re-emerging. As a consequence, policymakers, food regulators and health professionals need up-to-date and specific information about what people are eating and how much they are eating. They need to know the health and nutritional status of the Australian population. They also need to know how supplies of food and food consumption patterns are changing over time, and what food products contain. In turn, consumer education policies and tools, such as population dietary guidelines and food selection guides, need to be built on a solid foundation of knowledge about the national nutrition profile to reduce the risk of serious nutrition-related diseases and conditions. Food safety regulators also need this information to estimate current exposure to bioactive compounds that may be of concern, such as food additives and contaminants, to inform food fortification policies, and to ensure that nutrition information on food labels is relevant to current consumption patterns. Yet, Australia is unusual among its peers for not having continuous nutrition intelligence. Health-related data about Australians are compiled biannually, but current information about food and nutrient consumption, and trends in these, is conspicuous by its absence.1 The United States, the United Kingdom, and many European nations have had ongoing, systematic programs for monitoring the diet and nutritional status of their populations for many years.2-4 These programs are not confined to large countries with big budgets. For example, in 2001, New Zealand embarked on a 10-year strategic plan for a coordinated national population survey program that includes nutrition surveys in adults and another in children every 10 years, with the next surveys of adults and of children due in 2007–08, and 2012, respectively.5 In contrast, Australia has conducted only three national surveys of diet in the past 50 years: a national dietary survey of adults in 1983, and of children in 1985, and the National Nutrition Survey in 1995, which included both adults and children.6-8 Each cross-sectional survey was conducted by a different agency, using different sampling and collection methods and food composition data. These differences limit our ability to describe trends in food and nutrient consumption.9 Several state and territory governments in Australia have established monitoring systems that survey health behaviours, including food habits.10 These systems provide important information for tracking change, but food production, retailing and consumption are not limited by state boundaries, and information about selected food habits is not a sufficient base on which to build nutrition and food regulatory policy. Those with commercial as well as health interests in nutrition surveillance now favour a new national effort — one that goes beyond the brief encounters of cross-sectional surveys — to provide continuous detailed information on trends in food and nutrient consumption, the food supply and the nutritional status of Australians. Continuous nutritional surveillance must form part of a comprehensive policy to combat nutritional disorders. Where such surveillance exists, such as in the US, the data have been used to evaluate dietary guidelines, revise food selection guides,11 develop and evaluate fortification programs, set “real-life” serving sizes for nutrition information panels on food labels, make decisions about specific food processing regulations, and model the impacts of bioterrorism threats from food contamination.2 The centrepiece of the US system is the continuing National Health and Nutrition Examination Survey (NHANES), which is supplemented by many other sources of data.12 International experience suggests that there are two important actions for Australia to take in developing a food and nutrition monitoring system: 1. Establish a small, affordable, but statistically robust ongoing nutrition survey program, with data collected from a sample each year and reported cumulatively over a number of years. This program should be based within a federal agency that has health information responsibilities, uses consistent methods, can document and maintain databases, and reports on a predictable and timely basis. 2. Create a small nutrition monitoring unit to compile, disseminate and promote the use of all appropriate information about the food and nutrition situation in Australia for various policy, program, and regulatory purposes. The Australian Government Department of Health and Ageing recently commissioned the preparation of a business case for a no-frills national nutrition surveillance system in Australia, and consulted widely with stakeholders on its importance and suitability.13 Many disparate groups in Australia — in food production, marketing, regulation, and consumer health — now concur about the need for continuous food and nutrition monitoring, as well as the imperative to find workable solutions to long-term funding needs. While a national cross-sectional nutrition survey of children is currently being planned, its value would be greater if it were the start of a continuing surveillance program. In this regard, the recent announcement by the Minister for Health and Ageing of $3 million initially for a survey of diet, physical activity, and the weight status of Australian children, and $1 million annually thereafter for the collection of similar data on all population groups in Australia, is welcome, especially if it evokes a matching response from other key data users, including the states and territories.14 This may be the politically propitious moment for a long-term commitment to a system of continuous monitoring of food and nutrition in Australia.

Karen L Webb PhD, MPH · Ingrid H Rutishauser MSc · Geoffrey C Marks PhD, MS, DipNutrDiet · Gregory Masters MSc · Stephen R Leeder PhD, FRACP

Objective: To determine the potential for organ donation in 12 Victorian hospitals.Design and setting: Prospective audit of all deaths in 12 major public hospitals in the state of Victoria between January 2002 and October 2004.Main outcome measures: Number of organ donors and potential organ donors (patients with brain death or likely to progress to brain death within 24 hours if supportive treatment continued), requests for organ donation and consents. Unrealised potential donors (organ donation not requested) were categorised by an independent panel of intensivists as category A (brain death formally diagnosed); B (brain death not formally diagnosed but criteria likely to be fulfilled); and C (potential to progress to brain death within 24 hours).Results: There were 17 230 deaths, 280 potential organ donors and 220 requests for organ donation. The 60 unrealised potential organ donors were classified as category A (3), B (17) and C (40). Consent rate was 53% to 65%, depending on the definition of potential donor (categories A, B and C or category A only). Consent rate was lower when discussions about organ donation were held by trainees or registrars (21%) than when specialists were present (57%) (P = 0.004). A maximum practically achievable organ donation rate for Victoria was estimated to be 15 to 17 donors per million population (current rate, 9 per million population).Conclusions: The potential for organ donation in Victoria is limited by a small organ donor pool. There is potential to increase the number of organ donors by increasing the consent rate (lower than expected from public surveys), the identification of potential organ donors (particularly those likely to progress to brain death if supportive treatment is continued), and requests for organ donation.

Helen I Opdam MB BS, FJFICM · William Silvester MB BS, FJFICM

Objective: To determine the incidence of and risk factors for surgical site infections in general practice.Design: Prospective, observational study of patients presenting for minor excisions.Setting: Primary care in a regional centre, Queensland, October 2004 to May 2005.Participants: 857 patients were assessed for infection.Results: The overall incidence of infection was 8.6% (95% CI, 3.5%–13.8%). Excisions from lower legs and feet (P = 0.009) or thighs (P = 0.005), excisions of basal cell carcinoma (P = 0.006) or squamous cell carcinoma (P = 0.002), and diabetes (P < 0.001) were independent risk factors for wound infection.Conclusion: Our results indicate the high-risk groups for surgery in a general practice setting, such as people with diabetes and those undergoing excision of a non-melanocytic skin cancer or excision from a lower limb. Recognition of these groups could encourage more judicial use of prophylactic antibiotics and use of other interventions aimed at reducing infection rates.

Clare Heal MB ChB, DRACOG, FRACGP · Petra Buettner MSc, PhD · Sheldon Browning DRACOG, DipDerm, FRACGP

Objective: To examine factors influencing the identification of kidney impairment in patients with type 2 diabetes in Australian primary care.Design, setting and participants: 348 general practitioner investigators were asked to estimate kidney function and its severity in 10–15 consecutively presenting patients with type 2 diabetes (n = 3893). They were then asked, for each patient, whether they routinely estimated kidney function. No instruction was provided on how kidney function should be estimated or categorised. Data were collected between April and September 2005.Main outcome measures: Kidney function estimated by the Cockcroft–Gault equation using clinical and laboratory data provided by the GP; estimates of kidney function made by the GP.Results: In 24% of the patients with type 2 diabetes, their GP routinely estimated kidney function. However, few of these patients had impaired kidney function or risk factors for kidney disease. There was a good statistical correlation between the estimates made by GPs and the data-derived estimates (R2 = 0.72). GPs identified patients with data-derived estimates of kidney function < 60 mL/min in over 83% of cases, with a specificity of 90%. Impaired kidney function was reported by GPs in 34.4% of men and 36.4% of women. These figures were discordant with function categorisation using both GP estimates and data-derived values, overlapping in half of the patients. Despite GPs’ ability to assess creatinine clearance, “raw” (unstandardised) serum creatinine levels inappropriately influenced the perception of impairment of kidney function.Conclusion: GPs can accurately assess kidney function, without reporting of estimated glomerular filtration rate (eGFR). However, even in patients at increased risk of chronic kidney disease, routine estimates are seldom made. Our findings underline the value of the recent initiative recommending automatic reporting of eGFR in Australia.

Merlin C Thomas PhD, FRACP · Andrew J Weekes MD · Olivia J Broadley BSc, BCA · Mark E Cooper PhD, FRACP

Objective: To determine whether the adoption of the CONSORT checklist is associated with improvement in the quality of reporting of randomised controlled trials (RCTs).Data sources: MEDLINE, EMBASE, Cochrane CENTRAL, and reference lists of included studies and of experts were searched to identify eligible studies published between 1996 and 2005.Study selection: Studies were eligible if they (a) compared CONSORT-adopting and non-adopting journals after the publication of CONSORT, (b) compared CONSORT adopters before and after publication of CONSORT, or (c) a combination of (a) and (b). Outcomes examined included reports for any of the 22 items on the CONSORT checklist or overall trial quality.Data synthesis: 1128 studies were retrieved, of which 248 were considered possibly relevant. Eight studies were included in the review. CONSORT adopters had significantly better reporting of the method of sequence generation (risk ratio [RR], 1.67; 95% CI, 1.19–2.33), allocation concealment (RR, 1.66; 95% CI, 1.37–2.00) and overall number of CONSORT items than non-adopters (standardised mean difference, 0.83; 95% CI, 0.46–1.19). CONSORT adoption had less effect on reporting of participant flow (RR, 1.14; 95% CI, 0.89–1.46) and blinding of participants (RR, 1.09; 95% CI, 0.84–1.43) or data analysts (RR, 5.44; 95% CI, 0.73–36.87). In studies examining CONSORT-adopting journals before and after the publication of CONSORT, description of the method of sequence generation (RR, 2.78; 95% CI, 1.78–4.33), participant flow (RR, 8.06; 95% CI, 4.10–15.83), and total CONSORT items (standardised mean difference, 3.67 items; 95% CI, 2.09–5.25) were improved after adoption of CONSORT by the journal.Conclusions: Journal adoption of CONSORT is associated with improved reporting of RCTs.

Amy C Plint MD · David Moher PhD · Andra Morrison BSc · Kenneth Schulz PhD, MBA · Douglas G Altman PhD · Catherine Hill MB BS, MSc · Isabelle Gaboury PhD(c)

Vitamin D deficiency has re-emerged as a significant paediatric health issue, with complications including hypocalcaemic seizures, rickets, limb pain and fracture. A major risk factor for infants is maternal vitamin D deficiency. For older infants and children, risk factors include dark skin colour, cultural practices, prolonged breastfeeding, restricted sun exposure and certain medical conditions. To prevent vitamin D deficiency in infants, pregnant women, especially those who are dark-skinned or veiled, should be screened and treated for vitamin D deficiency, and breastfed infants of dark-skinned or veiled women should be supplemented with vitamin D for the first 12 months of life. Regular sunlight exposure can prevent vitamin D deficiency, but the safe exposure time for children is unknown. To prevent vitamin D deficiency, at-risk children should receive 400 IU vitamin D daily; if compliance is poor, an annual dose of 150 000 IU may be considered. Treatment of vitamin D deficiency involves giving ergocalciferol or cholecalciferol for 3 months (1000 IU/day if < 1 month of age; 3000 IU/day if 1–12 months of age; 5000 IU/day if > 12 months of age). High-dose bolus therapy (300 000–500 000 IU) should be considered for children over 12 months of age if compliance or absorption issues are suspected.

Craig Munns MB BS, PhD, FRACP · Margaret R Zacharin MB BS, FRACP · Christine P Rodda MB BS, PhD, FRACP · Jennifer A Batch MB BS, MD, FRACP · Ruth Morley BA, MB BChir, FRCPCH · Noel E Cranswick MB BS, BMedSc, FRACP · Maria E Craig PhD, FRACP, MMed · Wayne S Cutfield BHB, MB ChB, MD · Paul L Hofman MB ChB, FRACP · Barry J Taylor MB ChB, FRACP · Sonia R Grover MB BS, FRACOG · Julie A Pasco BSc(Hons), PhD · David Burgner MB ChB, PhD, FRACP · Christopher T Cowell MB BS, FRACP

Despite astounding advances in scientific knowledge and technological capabilities, modern medicine is not free of significant problems. A persistent high rate of diagnostic errors, the prevalence of medical (iatrogenic) harm and the growing demand for complementary and alternative medicine indicate an urgent need for improvement. An important step is a return to three quintessential clinical paradigms that have become neglected with the advent of high-technology medicine: the need to emphasise prevention and early, presymptomatic diagnosis; the crucial role in decision making of skilful history taking and examination, backed by evidence; and enhanced attention to patient autonomy and emotional factors. Possible reasons for the current neglect of these Altneuparadigms (“old–new” paradigms) are considered, and techniques for restoring their primacy in medicine are discussed.

Ami Schattner MD

Use of locum medical officers is increasing in the NSW hospital system. Locums are expensive, and have highly variable expertise and experience. Locum employment arrangements are ambiguous. Locum work may divert junior doctors from participation in specialist training. Attempts to regulate the locum workforce must be accompanied by measures that increase the appeal of public hospital work and vocational training positions.

Clare A Skinner MB BS, MPH, BA(Hons) · Rebecca L Riordan BNursing, MSc(Public Health) · Kylie L Fraser BAppSc(OT) · John D Buchanan LLB, GradDipEcon, PhD · Kerry J Goulston AO, MD, FRACP

Clinical record A 15-year-old boy developed recurrent, severe occipital headache, with vomiting, lethargy and blurred vision, after a minor head injury without loss of consciousness. The headache was dull and tense, exacerbated by activity and ameliorated by sleep, but unaffected by a change of posture. A week after the head injury, his general practitioner found him to be neurologically normal. The results of a non-contrast cerebral computed tomography scan were also normal. Based on a history of reversible wheeze and cough, the patient had been diagnosed with asthma at 12 years of age. He was strictly compliant with his prescription of inhaled fluticasone propionate (1000 μg/day via accuhaler) and rinsed his mouth after use. As instructed, he doubled the dose of inhaled corticosteroids during acute attacks, which occurred twice a year. There were no interval symptoms. Nonetheless, his dose of inhaled corticosteroids was never reduced. Before being prescribed inhaled corticosteroids, he had been prescribed nasal steroids for allergic rhinitis and nasal polyps, but rarely required this after a successful surgical intervention. He was not taking any other medications or vitamins. A year before presentation, he had had an episode of witnessed collapse with spontaneous recovery. This was attributed to a vasovagal attack at the time. Paediatric assessment 1 month after the head injury showed that he was lean and healthy. His weight (49 kg) and height (160.2 cm) were on the 25th centiles, and his body mass index (19 kg/m2) was on the 30th centile. His head circumference (56.8 cm) was on the 75th centile. Neurological examination gave normal results. A Mini-Mental State Examination identified a short-term memory deficit (2/3 objects recalled). The results of a cardiovascular examination were normal except for hypertension (140/70 mmHg; > 97th centile). There were no renal bruits. Urinalysis gave normal results. He was clinically euthyroid, and had no cutaneous hypo- or hyperpigmentation or signs of Cushing syndrome. He was diagnosed with postconcussion syndrome and reactive hypertension. An ophthalmological assessment identified normal visual fields and acuity and no papilloedema. The fluticasone propionate dose was reduced to 500 μg/day with a view to gradually stopping it altogether. Over the following 2 weeks, his headache worsened in intensity and frequency. An electroencephalogram and cerebral magnetic resonance scan were normal. Amitriptyline (25 mg/day) was prescribed for the headache, without effect. Two months after the paediatric assessment, lumbar puncture, performed under chloral hydrate sedation when relaxed with the back fully extended, showed an elevated cerebrospinal fluid (CSF) opening pressure of 24 cmH2O (reference range, 10–20 cmH2O).1,2 The procedure resulted in immediate headache relief. Examination of the CSF showed no abnormalities. Benign intracranial hypertension was diagnosed and he was prescribed oral acetazolamide (250 mg four times daily). His inhaled corticosteroid therapy was discontinued. Persistent mild hypertension (130/70 mmHg; 97th centile) was again noted on repeated occasions. The results of investigations (including renal function, urinalysis, renal ultrasound, full blood count, inflammatory markers, complement factors, antinuclear antibody, electrocardiography, chest x-ray, thyroid function, plasma renin activity, and urinary catecholamines) were normal and did not reveal a secondary cause of the hypertension. However, the morning cortisol level was undetectable (< 6 nmol/L). A morning short Synacthen test (250 μg) confirmed adrenal suppression, with a suboptimal cortisol rise from baseline of 175 nmol/L to a 1-hour peak of 275 nmol/L (reference, > 600 nmol/L). A paired adrenocorticotropic hormone (ACTH) level was 15 pmol/L, not consistent with primary adrenal disease. In view of the adrenal suppression, he was prescribed oral hydrocortisone (20 mg three times daily) for periods of acute physiological stress. Over the ensuing months, he required multiple therapeutic lumbar punctures, with immediate headache relief on each occasion. The acetazolamide dose was ceased 2 months after initiation. His hypertension resolved 4 months after presentation and the headaches resolved after 9 months. During recovery, he used “stress” hydrocortisone for two episodes of fever and lethargy. Serial (3-monthly) morning short Synacthen tests (250 μg) were performed. By 12 months, he had a normal peak cortisol response (657 nmol/L) and a baseline ACTH level of 7.3 pmol/L. Steroid precautions were then discontinued. Ophthalmological review now identified a cotton-wool spot on the inferotemporal edge of the right optic disc, suggestive of a past axoplasmic blood flow impediment caused by raised intracranial pressure. The temporal association between his headache severity, CSF opening pressures and cortisol levels is shown in the Box. Benign intracranial hypertension has an incidence of 1 per 100 000 children.3 As “benign” belies the potentially debilitating nature of this disorder, including the risk of irreversible blindness,4 the synonyms “pseudotumor cerebri” and “idiopathic intracranial hypertension” are sometimes preferred. Although the pathophysiological basis has not been determined, benign intracranial hypertension has been associated with vitamin deficiencies and toxicities, obesity, head injury,5 adrenal insufficiency,6 and medications, including topical and systemic steroids.7-9 To the best of our knowledge, this is the first reported association between benign intracranial hypertension and inhaled corticosteroids (we conducted a MEDLINE search, using multiple search terms, of English language publications from 1966 to June 2005). The diagnosis of benign intracranial hypertension in our case was based on modified Dandy’s criteria, an elevated cerebrospinal fluid (CSF) opening pressure (using the paediatric reference range),1,2 and repeated instances of immediate headache relief with lumbar puncture. The ophthalmological findings were non-specific but supported the diagnosis. Magnetic resonance venography would have excluded venous sinus thrombosis (an important cause of benign intracranial hypertension), but there was no reason to suspect a hypercoagulable tendency. As the CSF opening pressure was first measured 2 months after the patient sustained a minor head injury, we cannot determine whether this injury was the trigger for benign intracranial hypertension or incidental to it. However, the injury alone cannot explain the benign intracranial hypertension in this patient, given its mild nature and the absence of any radiological evidence of parenchymal brain injury. The adrenal suppression was most likely secondary to inhaled corticosteroid therapy. This is supported by the gradual adrenal recovery after withdrawal of inhaled corticosteroids. The past history of collapse suggests the onset of the adrenal suppression may have occurred 1 year before presentation. Given the absence of papilloedema at presentation, and the fact that the headache and elevated CSF opening pressure were clearly temporally associated with the withdrawal of inhaled corticosteroids, we suggest that the benign intracranial hypertension was related to the withdrawal of the steroid dose. Benign intracranial hypertension has previously only been reported in cases of primary adrenal insufficiency (Addison disease),6 or in association with the use or withdrawal of topical, oral or intranasal corticosteroids.7-9 Clinicians should be mindful of benign intracranial hypertension as a potential side effect when prescribing inhaled corticosteroids. Our case emphasises the importance of using appropriate inhaled corticosteroid dosage and “back titration”. Patients receiving inhaled corticosteroids require regular review. Dosages should be monitored to achieve the minimal dose for symptom control.10 Temporal association between headache severity, cortisol levels, and cerebrospinal fluid opening pressure Medication: F = fluticasone propionate (initially 1000 μg/day); A = acetazolamide (initially 1000 mg/day). CSF = cerebrospinal fluid.

Patrick Patradoon-Ho FRACP · Hasantha Gunasekera MIPH(Hons), FRACP · Monique M Ryan MMed, FRACP · Geoffrey R Ambler MD, FRACP

Clinical record Lesions characteristic of cutaneous anthrax A 48-year-old farmer, with a history of type 2 diabetes, hypertension and hypercholesterolaemia, presented to the local hospital with fever, malaise and inflamed lesions on his hands and arms. He reported that about 5 weeks before his presentation at hospital, he had grazed his arms while clearing the loading auger on a grain harvester on his property in south-western New South Wales. Within days of the initial injury, pimples appeared at the grazed sites and these developed into larger inflamed lesions over the following weeks. He reported that during this time he scratched one lesion and pus was released. The patient was admitted and commenced on intravenous flucloxacillin with a provisional diagnosis of cellulitis. Cutaneous anthrax was not considered as a diagnosis. Oral erythromycin was added to his treatment the following day. Swabs from the lesions were collected on the day of admission and sent to the local laboratory for routine culture. Staphylococcus aureus was isolated, as well as a Bacillus species, which was thought to be a wound contaminant. Two days after admission, a nurse at the hospital noted the appearance of the lesions and discussed the patient with infectious disease control officers in the population health unit. The nurse provided photographs, and the characteristic appearance of the lesions in a farmer from the anthrax belt raised the suspicion of cutaneous anthrax. (Figure) The centres of the lesions were depressed and some had formed black scabs. The surrounding tissue was red and oedematous with extensive swelling. These were noted as characteristic features of cutaneous anthrax eschars.1 The local laboratory then performed a gram stain of the Bacillus sp. This suggested Bacillus anthracis, and the isolate was forwarded the following day to the Institute of Clinical Pathology and Medical Research. The appearance of the lesions in a farmer from an anthrax-prone area and the Gram stain of the organism led to a provisional diagnosis of cutaneous anthrax. Antibiotic treatment was changed 2 days after admission to intravenous penicillin for 6 days. One day before discharge, the patient commenced a course of oral ciprofloxacin. The patient made a full recovery. Reference laboratory investigation Gram stain of the isolate revealed large, square-ended, gram-positive rods containing oval, central to subterminal spores that did not swell the cell. The isolate produced dull, flat, white-grey non-haemolytic colonies, 3–5 mm in diameter on horse-blood agar aerobically after 24 hours’ incubation. The colonies showed characteristic comma-shaped outgrowths along the edge and were sticky when touched with an inoculating loop. The organism was non-motile in broth medium after 4 and 24 hours’ incubation. These findings are consistent with B. anthracis. Bacterial cell wall long-chain fatty acid analysis was performed using the Microbial Identification System (MIS) (MIDI Inc., Newark, Delaware, USA). MIS identified the isolate as B. anthracis when compared with library entries on the Bioterrorism database. On cluster analysis, the isolate clustered with previous isolates of B. anthracis. The B. anthracis-specific polymerase chain reaction (PCR) assay was performed.2 Briefly, this is a PCR system that detects five target gene sequences present in the chromosome, and virulence plasmids pX01 and pX02. The chromosomal marker (Ba813) is present in all B. anthracis strains. Virulent strains of B. anthracis are encapsulated and toxigenic. Plasmids pX01 and pX02 are both required for virulence, and the absence of either results in attenuation. The PCR assay on this organism gave a positive result for all five targets Anthrax is a bacterial infection resulting from endospores of Bacillus anthracis, a gram-positive, rod-shaped bacterium, entering the body through skin abrasions or by inhalation or ingestion.1 The cutaneous form accounts for more than 90% of all human cases of anthrax worldwide.3 Anthrax is a zoonosis, and normally affects grazing animals such as sheep, cattle and goats. Animals usually become infected by ingesting anthrax spores, which remain viable in the soil for many years; the spores are resistant to desiccation and ultraviolet light.4 In New South Wales, anthrax in animals is a notifiable disease under the Stock Diseases Act 1923 administered by the NSW Department of Primary Industries (DPI). The DPI has responsibility for monitoring known anthrax properties and controlling outbreaks of animal anthrax. Most outbreaks in NSW occur in the “anthrax belt”, which runs through the middle of NSW and into northern Victoria. The area is bordered by Moree and Bourke in the north and Albury and Deniliquin in the south. About three properties per year in NSW experience an anthrax outbreak.5 Cutaneous anthrax in humans usually results from direct contact with infected animals or animal products such as wool, meat or hides, and is generally an occupational hazard. The head, forearms and hands are the most common sites of infection.4 The lesions are not usually painful, but pain may result from oedema or secondary infection. The differential diagnosis includes conditions producing papular lesions with regional lymphadenopathy. If the lesions are purulent, staphylococcal lymphadenitis is the most likely cause, although secondary staphylococcal infection may occur with cutaneous anthrax.1 Human cutaneous anthrax is rare in Australia, with the most recent cases reported in 1998 in a forklift driver in Queensland and in 1997 in a knackery worker in northern Victoria.6 The last known case in NSW occurred in 1991, in a worker who was slaughtering sheep on a property in the anthrax belt.7 Lessons from practice Human cutaneous anthrax may occur in Australia, associated with occupational exposure to infected stock or products contaminated with anthrax spores. Cutaneous anthrax eschars characteristically appear as depressed ulcers with black scabs surrounded by oedema, and are painless and not purulent. Secondary infection may produce purulent lesions and mask the clinical and laboratory diagnosis of cutaneous anthrax. Health care workers and laboratory staff in the anthrax belt should be aware of the clinical and laboratory aspects of cutaneous anthrax. The source of infection in our patient remains unclear, despite the farm being an anthrax-prone property. The farmer had not reported any recent animal anthrax on the property, although animal outbreaks had occurred in previous years. The spores might have been introduced directly from the soil at the time of the injury, or entered the wounds later following contact with soil, animals or other products contaminated with anthrax spores. Animal anthrax is more likely to occur following a climate change such as heavy rain after a prolonged drought.5,8 About a week before the farmer sustained the injury, the property, which had experienced drought for number of years, received more than 40 mm of rain. Public health action included discussion of the nature of the infection with the patient and staff of the health facility, who were reassured that cutaneous anthrax is not transmissible from person to person and that the vegetative form of the bacterium is not infectious.1 Contact was made with the regional veterinary officer of the DPI, according to arrangements that provide for the exchange of information on zoonotic diseases. Routine vaccination and surveillance of stock, as well as the rapid identification and management of animal anthrax incidents are key actions required to reduce human cutaneous anthrax. Additionally, a high level of awareness among farm workers and health care staff will support the early diagnosis and appropriate treatment of human cases. Prevention through vaccination is not possible in Australia, as a human anthrax vaccine is not registered in Australia.9 The United States Centers for Disease Control and Prevention (CDC) do not recommend prophylaxis for the prevention of cutaneous anthrax. Active surveillance is recommended where there is a continuing risk of exposure.10 Prophylaxis for inhalation anthrax exposure is recommended as a 6-week course of ciprofloxacin or doxycycline.11 Inhalation anthrax results from breathing in large numbers of anthrax spores. This concentration of spores is not usually reached in soil, and inhalation anthrax has not been described in Australia.8 The characteristic eschar lesions and the absence of fever, pain and pus, along with a history of contact with animals or animal products, should suggest anthrax.1 However, secondary infection, as occurred in our patient, can produce purulent lesions and fever and mask the initial cause of infection. Cutaneous anthrax is generally self-limiting and resolves without complications. The infection responds rapidly to antibiotic treatment, and oral penicillin is generally highly effective for cutaneous disease, rendering lesions sterile after 24 hours.1 The CDC now recommend oral ciprofloxacin or doxycycline for cutaneous anthrax. This recommendation follows the use of spores of a β-lactamase-positive anthrax strain as a bioterrorism agent.12 Spread to regional lymph nodes and septicaemia occurs in about 20% of untreated cases of cutaneous anthrax.3 Intravenous therapy with a multidrug regimen is recommended for cutaneous anthrax with systemic involvement.12 In our patient, anthrax was not initially considered as a diagnosis, which may indicate a need to raise awareness of cutaneous anthrax among health care workers in the anthrax belt. It is also important that staff working in regional pathology laboratories are aware of the tests that can be undertaken at a local level and those that need to be referred to a reference laboratory for the diagnosis of human anthrax.

Anthony Kolbe BAppSc, MPH · Marion G Yuen BAppSc · Bridget K Doyle BAppSc, DipEd

Anaphylaxis is a serious, rapid-onset, allergic reaction that may cause death. Severe anaphylaxis is characterised by life-threatening upper airway obstruction, bronchospasm and/or hypotension. Anaphylaxis in children is most often caused by food. Bronchospasm is a common symptom, and there is usually a background of atopy and asthma. Venom- and drug-induced anaphylaxis are more common in adults, in whom hypotension is more likely to occur. Diagnosis can be difficult, with skin features being absent in up to 20% of people. Anaphylaxis must be considered as a differential diagnosis for any acute-onset respiratory distress, bronchospasm, hypotension or cardiac arrest. The cornerstones of initial management are putting the patient in the supine position, administering intramuscular adrenaline into the lateral thigh, resuscitation with intravenous fluid, support of the airway and ventilation, and giving supplementary oxygen. If the response to initial management is inadequate, intravenous infusion of adrenaline should be commenced. Use of vasopressors should be considered if hypotension persists. The patient should be observed for at least 4 hours after symptom resolution and referred to an allergist to assist with diagnosis, allergen avoidance measures, risk assessment, preparation of an action plan and education on the use of self-injectable adrenaline. Provision of a MedicAlert bracelet should also be arranged.

Simon G A Brown PhD, FACEM · Raymond J Mullins PhD, FRACP · Michael S Gold MD, FRACP

Insects that commonly cause anaphylaxis in Australia are honeybees, jack jumper ants and wasps. Honeybee (Apis mellifera) sting is the most common cause of sting anaphylaxis in most areas of Australia, except where jack jumper ants (Myrmecia pilosula) occur (in native woodland areas in southern and eastern Australia). In some regions of Australia, “European wasps” (Vespula spp) and “paper wasps” (including Polistes and Ropalidia spp) make significant contributions. Who is at highest risk of anaphylaxis to stings and bites? People at highest risk of severe anaphylaxis after a sting are those aged over 35 years (Level II) and those who have had a previous severe reaction (Level II) (see case scenario Box).1,2 Do sting reactions always get worse? Sting hypersensitivity may persist for decades, but reactions to individual stings are highly unpredictable. Individual sensitivity and the amount of venom may vary, and each sting has the capacity to resensitise. Prospective sting studies have shown that the probability of the next sting giving no reaction or a lesser reaction than the presenting reaction is about 80% for Vespula wasps, 50% for honeybees and 30% for jack jumper ants.1-3 Evidence-based practice tip When a definite diagnosis of rapid-onset, systemic sting allergy has been made and the responsible insect identified, insect venom immunotherapy is highly effective at reducing the risk of anaphylaxis to future stings (Level II).* * NHMRC levels of evidence. What are some common pitfalls in practice? Doctors should be aware that: In some patients with sting anaphylaxis, rash may not be observed or may be very transient.4 Over the 60–90 minutes after the sting, many patients with sting anaphylaxis require more adrenaline (as judged by titrated continuous infusion4) than the amount provided by use of one EpiPen syringe.4 Although honeybee venom immunotherapy (VIT) provides significant protection against life-threatening stings, the protection appears to be less complete than that conferred by VIT for Vespula spp and the jack jumper ant.2,3 Case scenario* A 60-year-old farmer was stung on the neck while driving a truck. After parking, he removed a stinger and found a dying bee. Within 5 minutes he experienced flushing, sweating, fading vision and momentary loss of consciousness. He was too confused to call an ambulance on his mobile phone, but managed to stumble out of the vehicle and lie on the ground, where he was found by a passing motorist 10 minutes later. An ambulance was called. Paramedics arrived 20 minutes after the sting event and noted the man was wearing a MedicAlert bracelet. He was arousable, showed no respiratory distress or rash, and had a regular pulse rate of 55 beats/min and systolic blood pressure of 80 mmHg. The patient had brief expiratory wheeze and a blood oxygen saturation level of 88%. His condition improved with horizontal posture and with administration of oxygen, intramuscular adrenaline (0.5 mg injected into the lateral thigh) and nebulised salbutamol. On arrival at a hospital emergency department 30 minutes later, he received another intramuscular injection of adrenaline (0.3 mg), and 2 L normal saline was given intravenously over 20 minutes. A transient urticaria was noted. After another 4 hours’ observation, the patient was discharged with a prescription for an automated adrenaline syringe (EpiPen), an emergency anaphylaxis action plan, and an appointment arranged to visit an allergy clinic. The man had tolerated bee stings on multiple occasions up to the age of 53 years, when, for the first time, he developed flushing and mild wheezing 15 minutes after a sting. At that time, he was provided with an EpiPen. Five years later, he experienced an uneventful bee sting and decided not to renew the EpiPen. Intercurrent problems included mild chronic airflow obstruction, a history of smoking (until 30 years of age) and borderline hypertension (treated with a thiazide). Specialist allergy assessment 1 month after discharge demonstrated serum-specific IgE to honeybee venom at 10 kU/L and a positive skin test to the venom at 0.01 μg/mL (intradermal). Based on sighting the insect, the presence of a stinger and demonstration of high-level specific IgE to honeybee venom, the recent events were attributed to allergy to honeybee sting. The patient’s respiratory and cardiac comorbidities may also have contributed to the severity of the reaction (Level IV). Because of his occupation, the presence of comorbidities, the severe, hypotensive anaphylaxis after the latest sting, and the slow trend to higher failure rates once venom immunotherapy is discontinued,2 the patient was trained in using an EpiPen device and advised to carry one at all times and to undergo honeybee venom immunotherapy indefinitely. He was also given contact numbers for emergency services (dial 000 in Australia, or 112 from a mobile phone anywhere in the world) and advised how to minimise exposure. * This is a fictional case scenario based on similar real-life cases.

Robert J Heddle PhD, FRACP, FRCPA

Intraoral examination revealed rampant dental caries (a), enamel erosion (b) and generalised calculus deposits (c). A 22-year-old man presented with badly decayed teeth and generalised dental pain. Intraoral examination revealed rampant dental caries, enamel erosion and generalised calculus deposits (Box). The patient had a 2-year history of daily use of crystal methamphetamine, administered intravenously or via snorting. He complained of a dry mouth, which led him to consume 2–3 litres of carbonated beverages a day, and reported periods of jaw clenching and temporomandibular joint discomfort on waking. This clinical syndrome in methamphetamine users, which has been referred to as “meth mouth”,1 can be attributed to several factors: the acidic nature of the drug, its xerostomic effect and propensity to cause cravings for sweetened carbonated beverages; drug-related tooth grinding and clenching; and extended periods of poor oral hygiene. Treating dentists should avoid the use of nitrous oxide, narcotics and vasoconstrictors. Early interventions, such as dietary counselling, application of fluoride varnish and dental plaque control, can prevent significant damage to the teeth.

Kishore Shetty

CorrectionsRe: “Rural internship for final-year medical students”, by Tarun Sen Gupta, in the 3 July issue of the Journal (Med J Aust 2006; 185: 54-55). This letter was co-authored by Dr Richard B Murray, Associate Professor and Planning Director, Rural Clinical School, James Cook University, Townsville, QLD. Dr Murray’s details were omitted from the manuscript of the letter submitted to The Medical Journal of Australia, and so were not printed with the letter. The html and pdf versions of this article were corrected on 29 August 2006.

Tarun Sen Gupta · Richard B Murray

CorrectionsRe: “The burden of chronic kidney disease in Australian patients with type 2 diabetes (the NEFRON study)”, by Merlin C Thomas, Andrew J Weekes, Olivia J Broadley, Mark E Cooper and Tim H Mathew, in the 7 August issue of the Journal (Med J Aust 2006; 185: 140-144). In the statement defining microalbuminuria, macroalbuminuria and normoalbuminuria on page 141, the sexes were transposed. The correct statement is: Albuminuria was stratified according to International Diabetes Federation guidelines12,13 as follows: Microalbuminuria — urinary albumin–creatinine ratio (ACR) of 3.5–35 mg/mmol (women) or 2.5–25 mg/mmol (men). Macroalbuminuria — urinary ACR > 35 mg/mmol (women) or > 25 mg/mmol (men). Normoalbuminuria — urinary ACR < 3.5 mg/mmol (women) or < 2.5 mg/mmol (men). The html and pdf versions of this article were corrected before publication.

Merlin C Thomas · Andrew J Weekes · Olivia J Broadley · Mark E Cooper · Tim H Mathew

To the Editor: Trans fats are produced by partial hydrogenation of liquid vegetable oils to produce oils which are more solid at room temperature and have better physical properties for food processing, such as increased shelf-life. Trans fats represent a major dietary cardiovascular disease risk, with as little as 5 g daily increasing the risk of ischaemic heart disease by 25%.1 A recently published survey of the trans-fat content of French fries and chicken nuggets purchased from two international fast food chains in 20 countries emphasises the wide variability of trans fats in different countries.2 This work highlights the potential health risks imposed by the industrially generated trans fats in these food products. Of the sampled French fries and nuggets, 20 of 39 samples from 19 different countries yielded trans-fat levels in excess of 5 g for an average serve. Interestingly, the report included no data from Australia. We have evidence that similar fast foods have substantial quantities of trans fats (putting Australia in the mid-range of the league table). The only available published results are those reported by the Australian Consumers Association (ACA) in 2005.3 The ACA tested 55 foods and found 18 had trans-fat levels greater than 2% of total fat. The interesting issue is that the ACA data show a variation in trans-fat levels of greater than 22-fold (0.8%–22.5% of total fat) in popular fast foods. Data on trans-fat levels should be available on all foods in this country, whether sold in supermarkets or to the food service industry. However, at present, there is no requirement to include trans-fat content on nutrient information panels, except when the manufacturer wishes to make a nutritional claim about cholesterol, saturated, unsaturated or trans-fatty acids.4 Many countries, including the United States, Canada and some European countries, have either placed limits on the permissions for trans fat in processed foods, or, more commonly, mandated labelling requirements. The most notable is Denmark, where legislation restricts maximal industrially produced trans fats to less than 2%.5 Not surprisingly, that country reported markedly lower trans-fat contents in fries and nuggets than those sold in Australia.2 Despite review of the “Australia New Zealand Food Standards Code”, labelling of the trans-fat content of food has not been mandated,4 and consumers and health professionals wishing to reduce their trans-fats intake remain unable to make informed choices.

David Cameron-Smith · Andrew J Sinclair

To the Editor: With obesity reaching epidemic proportions in Australia, professional education needs to reflect this increase in prevalence. Research has shown that health professionals’ lack of knowledge is a common barrier to providing care for overweight and obese individuals.1-3 We investigated the coverage of obesity education in university medical, dietetic and nursing curricula (partially replicating the earlier study in the Journal by Campbell and Welborn4), and also the extent to which obesity was covered in the curricula of selected professional specialist colleges in Australia. Contact hours for obesity were compared against two “control diseases”, diabetes and depression. Surveys were sent to 15 medical schools, and administrators of six dietetic courses and six nursing courses in Australia. The survey asked questions including the total number of contact hours in the course for each topic, and if additional teaching was needed. Most college curricula were publicly available, as were the Continuing Medical Education events and topics. Nine of the 15 medical schools and three of the six dietetic and nursing courses returned the questionnaire. Because of the nature of the data and the very small sample sizes, no statistical analysis was conducted. In the medicine curriculum, while variation in mean contact hours between obesity, diabetes and depression was small, the range was quite large (Box). There was also wide variation in the contact hours for obesity education in the nursing curriculum. The median contact hours for obesity education were half that of both depression and diabetes, and one university reported zero contact hours for obesity education. With the exception of the Royal Australian College of General Practitioners, the colleges surveyed did not include obesity in the prescribed teaching curriculum. Several professional education topics were available on obesity, but considerably more were available relating to diabetes and depression. Our findings indicate that most of the universities appear to provide undergraduate students with adequate contact hours for education about obesity. The professional training provided by the individual specialist medical colleges is not as comprehensive, and lacks specific obesity education. Based on these findings, more systematic research is needed to examine the details of training and to develop programs which better equip health professionals to deal with the growing burden of obesity and related diseases. Future research should not be limited to measuring contact hours — the focus should extend to investigating the content of courses and professional development programs, and to examining the barriers to including obesity education. Summary of median, mean, standard deviation and range for contact hours of teaching about obesity, diabetes and depression in undergraduate courses Course and disease Number reporting contact hours Contact hours Median Mean ± SD Range Medicine Obesity 8 7 13 ± 11 5–30 Diabetes 8 13 20 ± 20 6–64 Depression 6 13 13 ± 5 4–20 Nursing Obesity 3 6 8 ± 10 0–19 Diabetes 3 13 18 ± 8 10–25 Depression 3 13 16 ± 10 8–28 Dietetics Obesity 3 15 14 ± 8 6–22 Diabetes 3 20 17 ± 8 8–23 SD = standard deviation.

Melissa J Hayden · Leon Piterman · John B Dixon · Paul E O'Brien

To the Editor: The discussion of fibrinolysis in the recently published guidelines for the management of acute coronary syndromes 20061 is interesting. The recommendations clearly indicate that second-generation agents should be preferred to streptokinase in all circumstances. The guidelines reference the GUSTO-I trial data2 as the primary support for those recommendations. These data are, at best, debatable in terms of showing any benefit of front-loaded tissue plasminogen activator over streptokinase, and then only in limited circumstances (ie, patients aged less than 75 years with anterior infarcts and within 4 hours of the onset of symptoms).3,4 To my knowledge, there have been no head-to-head trials of this size of the other fibrinolytic agents discussed against streptokinase. Thus there is no justification for the blanket superiority that is accredited to these agents, both by implication and explicitly, in the guidelines. It is a matter of some concern that guidelines from such respected groups should make statements that will be used broadly by clinicians, but that go beyond the evidence base to which they refer. On the balance of information available there is no compelling case to relegate streptokinase from the front line.

Andrew J Bezzina

In reply: As Bezzina states, the GUSTO-I trial is the main source of evidence for the superiority of front-loaded alteplase (rt-PA) over streptokinase, showing a 1% absolute and 15% relative benefit.1 Subgroup analysis suggested that only certain groups benefited, but this is an inappropriate use of subgroups, and the result should be applied overall. A clear mechanistic reason for the advantage of rt-PA — greater 90-minute full coronary patency — has also been demonstrated.2 Meta-analyses of the percutaneous coronary intervention (PCI) trials in acute myocardial infarction have all shown benefit over fibrinolysis. However, the benefit of PCI is greater compared with streptokinase than with plasminogen activators.3 Although not providing a head-to-head comparison, these data also support the superiority of plasminogen activators over streptokinase. The second generation plasminogen activator studies have all compared these with the “gold standard” front-loaded rt-PA. These agents have been shown to not be inferior in relation to mortality,4,5 and tenecteplase showed a decrease in systemic bleeding.5 Administration as a bolus without the adverse reactions commonly seen with streptokinase (such as hypotension) make them much more convenient and safe, particularly in smaller institutions. In addition, streptokinase is an inappropriate choice in Indigenous patients because many have high levels of anti-streptokinase IgG and streptokinase resistance.6

Philip Aylward · Constantine N Aroney · Ken Hossack · Andrew M Tonkin

To the Editor: We are delighted at the attention which the editorial by Hickie and Blashki1 has drawn to our clinical update on the management of bipolar disorder in general practice.2 However, we are bemused by a number of the sentiments, criticisms and statements of fact included in that robustly expressed editorial. We will focus only on a few of the major issues raised. Hickie and Blashki argue that there are too many “worthy” guidelines promulgated to general practitioners by “specialist colleagues” across the range of medical conditions, and that extrapolation from specialist centre studies “may particularly annoy GPs”. On the other hand, they bemoan the fact that “few [guidelines] have targeted general practice”. We are surprised by this insinuation that such issues pertain to our clinical update. Three of the authors of our article are GPs in either clinical or academic practice, and the document has been formally endorsed by the Royal Australian College of General Practitioners. Our article focuses on the practical issues concerning the role of the GP in the management of patients with bipolar disorder, and deals frankly with the respective contributions of the GP, psychiatrist, and psychologist. It is our experience that GPs are enthusiastic in enhancing their skills in the management of mental illnesses such as bipolar disorder in the primary care setting. Therefore, we have little doubt that updates such as ours will be viewed as helpful aids for GPs, who are often the main “port of call” for people with this condition. We strongly contend the statement that we ignore practice-based issues and thereby risk “an overall negative rating from the target audience”. Hickie and Blashki state that “the most useful mental health guidelines tackle the tough issues”, such as sources of self-help, self-monitoring, detailed illness descriptions, family education, quality e-health resources, and guidance when patients become a danger to themselves and others. We fail to understand the implication that our update does not address such issues, as these very practical matters are clearly highlighted in detail in our article. Finally, we are surprised at the negative tone concerning guidance for the management of mental illness in general practice by authors who have argued strongly for the value of evidence-based guidelines in specialist psychiatric practice.3 Although (as we clearly acknowledge) there is currently a limited evidence base for managing such conditions in primary care, there is still a major need for practical guidance for the practitioner in this setting.

Philip B Mitchell · James A Best · Bronwyn M Gould · Ian G Wilson

To the Editor: Hickie and Blashki are to be commended for their view that clinical practice guidelines in mental health should be relevant to a primary care setting.1 Unfortunately, such guidelines have little effect on clinical outcomes, as most general practitioners have not been taught how to use them to their best advantage.2 There is also little known about the best way to implement guidelines in mental health, let alone in primary care mental health settings.3 As a result, more guidelines, even those more attuned to the primary care environment, will be of little benefit to our community. The Royal Australian and New Zealand College of Psychiatrists (RANZCP) is actively promoting the use of clinical practice guidelines4 as a quality improvement tool that will allow mental health practitioners (including GPs) to assess their practice more carefully and measure and analyse variance. The next step is to fund research into how best to implement mental health guidelines at the coalface. It is only through practice-based research that the barriers to successful implementation of evidence-based practice can be identified and overcome. Such research could be funded via a National Health and Medical Research Council (NHMRC) or Australian Research Council (ARC) grant program and coordinated by groups such as the RANZCP or the National Mental Health Working Group Safety and Quality Partnership Group. Mental health has already been identified as a grant funding priority by the ARC.5 Once this has been achieved, then training and mentoring to help practitioners review their practice as part of a quality improvement framework is required, rather than more guidelines per se. Providing well researched, up-to-date and accessible information for GPs on “self-help, self-monitoring, [and] detailed illness descriptions”, as suggested by Hickie and Blashki, is commendable, but is not what is required for guidelines to truly improve the safety and quality of mental health care in Australia.

Andrew J Wilson · David Barton

To the Editor: With Australia’s rapidly ageing population and an explosion in the number of retirement villages and nursing homes, Finn and associates are to be congratulated for ventilating the subject of the interface between residential aged care facilities and emergency departments.1 My experience of emergency department (ED) and aged care facility relations spans over 50 years and I have been involved in both sides of the equation. Firstly as a surgeon, then as director of an ED, and finally, as a resident of a retirement village for over 20 years (including, for my wife, 5 years in the affiliated nursing home), and during that time my wife and I have had at least eight episodes as patients in an ED. Retirement villages and nursing homes are not equipped or organised to handle medical or surgical emergencies. Problems of “disposal” arise after ED assessment and treatment in a public hospital. The hospital may not have an empty bed. The patient’s condition may not be serious enough to require a hospital bed, but the patient may not be well enough to return to his or her retirement village. Privately insured patients may have the option of transferring to a private hospital but usually spend an unnecessarily long time in the ED awaiting such transfer. Matters that need attention are: a standing arrangement between public and neighbouring private hospitals to facilitate quick transfer of suitable patients. the removal of long delays in EDs that occur while waiting for the results of investigations and even longer periods awaiting “higher opinions” after receiving these results. a speedier and more detailed hospital summary addressed to the general practitioner (if known) as well as to the aged care facility concerned.

Keith S Jones

Lance Armstrong effect The “Lance Armstrong effect” does not describe an ability to win multiple, gruelling Tours de France. Rather, it refers to the many patients, who, like Armstrong, have been successfully treated for their advanced metastatic testicular cancer. US researchers now suggest that a better understanding of this “effect” could help patients with other types of disseminated solid tumours, who generally do not do as well. They suggest the core reason for the effect may lie in the unusual thermal hypersensitivity of normal testicular germ cells and cancerous testicular cells — both have a propensity to die when at normal, instead of below normal, body temperature. The researchers also point out that the hyperthermia of fever can give a transient boost to immunity. Although hyperthermia, a very old form of cancer therapy, is not widely accepted as a cancer treatment, new types of thermal delivery systems are inviting another look. Targeted induced hyperthermia, achieved, for example, by applying external magnetic fields to heat up iron-containing nanoparticles or liposomes attached to tumour cell surface markers might help to enhance tumour responses to other cancer therapies. JAMA 2006; 296: 445-448 Male circumcision and HIV Waiting lists for adult male circumcision continue to grow in South Africa, as news spreads about a single, randomised controlled trial.1 The trial, conducted in South Africa and published last year, found that male circumcision could reduce the sexual transmission of HIV from women to men by about 60%.2 Further trials are being conducted in Zambia and Kenya and are due to be completed in 2007. If these trials confirm the earlier findings, the problem of providing safe and effective circumcision services will become even more important. However, the largest potential problem may well be a false perception of security — circumcision may reduce, but it does not remove, the risk of transmission. 1. Bull World Health Organ 2006; 84: 509-511 2. PLoS Med 2005; 2: e298 In bed with Dame Edna ... “As long as I keep the pill between my knees, it’s safe and it works”, quipped Dame Edna Everage on TV in the early 1980s, when then-TV doctor Richard Smith was being interviewed (in a bed) about a complex study which suggested certain contraceptive pills were linked with breast cancer. Dr Smith went on to become editor of the BMJ but never forgot the many challenges and lessons learned in his time with the mass media. At that time, journals were offhand about or even hostile to media coverage. Today, Smith notes that medical journals are courting the mass media with media releases to reach the public as well as their readership and to engage in the political process and increase citations. Smith believes that medical and health issues should be debated in the mass media and, further, that journals can offer a place for a higher level of debate — including for non-doctors. J R Soc Med 2006; 99: 347-352 Surgical outreach Concerns that specialist outreach clinics in remote communities would uncover an overwhelming burden of illness requiring hospitalisation seem to be unfounded, say Australian researchers. Gruen and colleagues studied three Indigenous remote communities in the Northern Territory from 1990 to 2001; each community received some visits from consultants in four surgical disciplines — general surgery, gynaecology, ophthalmology and ear, nose and throat —as part of the federally funded Specialist Outreach Service (SOS). Overall, 2339 new surgical problems presented in 2368 people, and there were 156 opportunistic presentations on outreach clinic days. However, while SOS improved access to (and completion of) consultations and procedures, there was no increase in elective referral rates for hospital outpatient clinics or for inpatient services. Lancet 2006; 368: 130-138 Kidneys a-plenty A Spanish transplant centre has been able to initiate a pre-dialysis transplant program thanks to a surplus of kidney donors — principally, non-heart-beating donors. Since 1989, the Hospital Clínico San Carlos in Madrid, Spain, has been procuring kidneys from adults aged younger than 60 years of age who died suddenly of irreversible cardiac arrest outside the hospital. Declared dead by circulatory rather than neurological criteria, they were transferred to the hospital for the sole purpose of organ donation. Recipient data from 1989 to 2004 showed that kidneys from non-heart-beating donors had a slightly higher incidence of non-viable transplantation and of delayed graft function. However, 1- and 5-year graft survival rates were similar to those for transplants from heart-beating donors younger than 60 years of age (and better than rates for transplants from heart-beating donors older than 60). Ann Intern Med 2006; 145: 157-164 Dr Ann Gregory, MJA

Ann Gregory