Letters

To the Editor: Russell and colleagues recently wrote an open letter to the new Minister for Health.1 In response, we call for the Minister to champion the cause of Indigenous health. Dear Minister,Russell et al raise a number of pressing issues directly relevant to the health portfolio.1 However, of all the challenges that face you, perhaps the greatest is reversing the neglect and extreme health disadvantage experienced by Australia’s Indigenous people.2 Although the task is daunting, and detractors may argue that there are no evidence-based solutions, 150 years of collective experience from across the globe provides compelling support for real investment in the spheres of water, sanitation, housing, education, employment and primary health care to reverse health disadvantage. Tragically, while these basic necessities are taken for granted by most Australians, they remain a dream for many Aboriginal and Torres Strait Islander peoples. The greatest public health gains during the previous two centuries resulted from ensuring that communities had sustained access to clean water, adequate sanitation and appropriate housing.3 It is astonishing that these basic rights should remain on the unresolved agenda of a highly developed country. To our shame, these basic direct health determinants are not yet guaranteed for Aboriginal and Torres Strait Islander Australians.4 Relative poverty, absolute poverty and social exclusion all have a major impact on health.5 In Australia, relative poverty denies many Indigenous communities access to housing, education, transport and other societal benefits. It would be naïve to argue that this inequitable distribution of Australian resources has not been a major determinant of the poorer health of Indigenous Australians.6 The recent Auditor-General’s report on whole-of-government Indigenous service delivery arrangements clearly indicates that current approaches are inadequate and fall short on service delivery.7 There is therefore a critical need for decisive direct investment in basic infrastructure and its maintenance in Indigenous communities, along with comprehensive primary health care. This must extend to equipping and developing individuals and communities through a major investment in education and creation of employment opportunities that engages the community and is developed in true and equal partnership with respected Indigenous leaders and communities You may argue that many of these health determinants fall beyond your direct sphere of accountability. You may choose to point to the small-scale success stories, particularly in Indigenous primary health care. However, as Minister for Health and Ageing, you will continually be confronted by the direct evidence of the deleterious results of these health determinants on the life expectancy and health of Indigenous Australians.8 The time is ripe for a bold national leader to champion this cause in the corridors of power. We encourage you to become that advocate among your Cabinet colleagues. The challenge is yours. Will you have the courage and moral fortitude to make your mark on Australian history?

David N Durrheim · Mark Wenitong · Clare Huppatz · George Rubin

To the Editor: The academic success of the rural clinical schools (RCS) program is clear: community-based clinical placements in rural Australia are able to produce graduates that are academically indistinguishable from tertiary hospital-trained peers.1,2 The effect of the RCS program on the workforce is yet to be established. Eley and Baker recently identified early career choices by Queensland RCS graduates.3 Western Australian RCS students spend an entire academic year in the country. Here, we provide the first data on the return of WA RCS graduates to rural internship positions relative to the intern year cohort as a whole. Based on two cohorts entering postgraduate year 1 (PGY1) positions in 2004 and 2005, 14 of 28 WA RCS graduates have requested and completed the one rural rotation that is permitted during their intern year. Given the limited rural rotations available in WA, some sought rural experience as far afield as Queensland and New South Wales. The substantive uptake of rural experience by RCS interns is in contrast to the intern cohort as a whole (Box). In 2004 and 2005, the three WA tertiary allocation centres placed 135 and 131 interns, respectively. As WA has no whole-year rural internships, a subset of the urban allocation included a 3-month rural rotation. The distribution of graduates of the RCS was different to that of graduates from other programs (χ2 = 7.0693; df = 1; P = 0.008). RCS students were more likely than other graduates to take a rural rotation (odds ratio, 3.1). Furthermore, ongoing postgraduate contact with the RCS cohort has identified that 23 of 28 graduates have chosen to undertake at least some time in the country during their PGY1–3 years. There is also early indication that graduates will sign up for a rural vocational training pathway (3/28). Interestingly, only a small proportion of these graduates are from a rural background (5/28). Our data suggest that rural practice is seen as highly desirable postgraduate clinical experience by WA RCS graduates. These results provide initial evidence that the WA RCS program will increase the rural workforce. Aggregate of 2004 and 2005 rural clinical school (RCS) interns versus non-RCS interns: comparison of postgraduate year 1 (PGY1) intern location choices RCS (n = 28) Non-RCS (n = 238) Total (n = 266) Expected tertiary allocation centre PGY1 internships (statistically even distribution) Urban-only internship 20 174 194 Rural rotation during internship 8 64 72 Actual tertiary allocation centre PGY1 internships (actual distribution) Urban-only internship 14 180 194 Rural rotation during internship 14 58 72

Denese E Playford · Harriet Denz-Penhey · Lesley Skinner · J Campbell Murdoch

To the Editor: The growing problem of tuberculosis in resource-rich countries has been recently highlighted,1 with immigration thought to be an important contributor. To assess a possible increase in incidence, we performed a retrospective case record review of all children who had tuberculin skin tests or who were diagnosed with tuberculosis at The Children’s Hospital at Westmead for 3 years from 2004 to 2006. This period included the establishment of a refugee clinic in May 2005, which routinely tests refugees by tuberculin skin testing. We compared our findings with published data from 1982 to 1991.2 Latent tuberculosis infection was defined as tuberculin skin test induration of ≥ 10 mm (regardless of prior BCG vaccination) and a decision by the treating physician to start isoniazid monotherapy. Proven active tuberculosis disease was defined as a child with a positive isolate of Mycobacterium tuberculosis from culture or positive polymerase chain reaction for tuberculosis or positive tuberculin skin test in association with a clinical picture strongly suggestive of active tuberculosis disease. The number of tuberculin skin tests performed increased through the study period (Box 1), largely because the refugee clinic saw 90 new patients in 2005 and 150 in 2006. The proportion of children with an increased induration response increased over the study period (Box 1 and Box 2). We observed an increase in latent tuberculosis infection, both in absolute numbers and in the proportion of the total caseload. The absolute number, but not the proportion, of cases of active tuberculosis disease increased during the study period. Over the same period, hospital admissions remained static at about 26 000 per year. Extrapulmonary tuberculosis was present in 12 of 23 patients (52%) with active tuberculosis, compared with 34% in the earlier study.2 There was no increase in tuberculosis meningitis. The active tuberculosis cohort ethnicity was consistent with the earlier study, with 22 of 23 patients of non-European origin, and 18 born outside Australia. The predominant ethnic groups were from Africa (eight, all born outside Australia) and the Indian subcontinent (four, two born outside Australia). Our finding of an increase in the proportion of patients with latent tuberculosis infection but not active tuberculosis disease is largely due to increased testing of refugees. It is reassuring that we did not find active tuberculosis. The United Kingdom has reported an increased incidence of tuberculosis in African immigrants.1,3 Australian immigration trends have shown a demographic shift, with increasing numbers of refugees from Africa.4 Although our study is likely to suffer from referral bias, it is the largest review of paediatric tuberculosis from a tertiary centre in Australia. Children very rarely transmit tuberculosis, but it is important to identify and treat latent tuberculosis to prevent progression to active disease.5 We believe our results are encouraging in showing a low incidence of active tuberculosis and indicate the need to screen refugees for latent tuberculosis to direct chemoprophylaxis. 1 Comparison data on tuberculosis among children, 2004–2006 Year Total TSTs performed Results available > 10 mm > 15 mm No. of patients commenced on isoniazid No. with active TB disease No Yes No Yes 2004 116 111 104 7 108 3 5 4 2005 257 247 202 45* 221 26* 25 8 2006 278 263 193 70* 215 48* 36 11 TST = tuberculin skin test. TB = tuberculosis. * P < 0.01 compared with previous year. The discrepancy between the total numbers performed and the cumulative numbers in the categorisation of response is due to patients not returning to have the TST read. 2 Results categorised by size of tuberculin skin test induration

Paul D Robinson · Dianne Dalton · Terri Cripps · Nicholas J Wood · Alison M Kesson · David Isaacs

To the Editor: I write to voice my objection to the use of the term “tsunami” by many in the medical education community when describing the massive increase in medical student numbers that has begun as a result of recent government initiatives.1-5 Although it is true that the government has opened the floodgates, as opposed to the water restrictions they applied in the mid 1990s, I believe describing this increase using the metaphor of a tsunami is a misnomer. I do not disagree with the potential for this phenomenon to wreak mayhem with the quality teaching and supervision of medical students and junior doctors and their access to adequate employment in both prevocational and vocational places. Nevertheless, I believe that we need to adopt a more accurate term to describe this impending event. A term I have heard on a number of occasions is that of the “rising sea levels” of medical graduates, which (although less dramatic) I think is more accurate, because this expression indicates that the change (likely a sea change) is going to be long-lasting, if not permanent, rather than a flash flood, as signified by “tsunami”. If we are going to continue to describe this impending increase as a “tsunami”, the only benefit I can see is that it may encourage us to learn lessons from the real event. As a result of the 2004 Boxing Day tsunami, a tsunami early warning system was put in place to alert citizens of an impending tsunami and give them a small window of time to prepare. A number of medical education groups, including the Australian Medical Association Council of Doctors in Training, have started sounding the alarm about the increase in medical students. It remains to be seen whether this warning will be heeded in time. One only has to look to the fiasco in the United Kingdom this year with junior doctor allocations to see what happens when adequate preparation is not made for impending change. Will a student who enters medicine in 2010 feel that, during their medical education voyage, they are receiving a quality education? And when they graduate, will they find a rich sea of prevocational and vocational opportunities? Or will they feel like a drop in the ocean of fellow graduates scrambling to gain a properly supervised training position?

Andrew W Perry

To the Editor: The Mandometer treatment for patients with eating disorders (Box 1), developed at the Karolinska Institute in Stockholm, was brought to the attention of the public and the profession in Australia through Norman Swan’s Health Report in 2003.1 Since then, 40 patients from Australia have received treatment using this method, 29 of whom were treated in Stockholm, seven in Melbourne and four in San Diego. In a randomised controlled trial, Mandometer treatment brought 75% of patients into remission in an average 14 months, with a relapse rate of about 10% during 5 years of follow-up.2 There are seven criteria for remission, including normal eating behaviour, normal body mass index (BMI) and physical parameters, and remission of psychiatric symptoms. Patients must have returned to school or work, be comfortable with their body weight, and have avoided binge eating and vomiting for at least 3 months. Patients fulfilling five of these criteria are considered in partial remission. Of the 40 (39 female) patients reported here, 27 patients (68%) were diagnosed according to criteria of the Diagnostic and statistical manual of mental disorders, 4th edition, with anorexia nervosa, seven (17%) with bulimia nervosa, and six (15%) with an eating disorder not otherwise specified. The median age of patients was 18.7 years (range, 10.7–39.3 years). On admission, they had been ill for a median period of 5 years (range, 1–25 years) and they had had up to 20 previous unsuccessful episodes of treatment (median, 3 episodes). Fourteen patients (35%) fulfilling five remission criteria have returned to Australia for follow-up, but have not yet returned to school or work. Eight patients (20%) are in full remission, 11 (28%) are in treatment, and seven (17%) withdrew from treatment before completion, some after only a few days. All patients who are treated in Stockholm and San Diego are followed up in Melbourne, and outpatient treatment is also offered at the Melbourne clinic. Of the 27 patients with anorexia, 16 have entered full or partial remission, with marked improvement in all parameters (Box 2). Six achieved remission and 10 partial remission in 220 (range, 129–257) and 220 (range, 168–570) days, respectively. The five anorexic patients who dropped out of the treatment had low BMI (11.3, 13.5, 13.8, 14.6 and 15.7 kg/m2), had been ill for 4, 5, 8, 14 and 25 years, and had had up to nine previous episodes of treatment. This report concerns severely ill Australian patients who had previously undergone many episodes of care. These preliminary results are encouraging, given the known resistence of these disorders to treatment.4,5 Although the outcomes reported here appear to be better, comparison with other methods used in Australia should be made in a randomised controlled trial. The Melbourne Mandometer Clinic is willing to participate in such a trial. 1 Mandometer treatment has four interventions 1. The patient re-learns how to eat using Mandometer, a computerised procedure that provides feedback during meals (Figure). In about 5–6 months, anorexic patients are able to eat a normal meal and perceive a normal level of satiety after practising eating gradually larger meals. Bulimic patients are similarly trained, but they are able to eat normal amounts of food from the beginning of treatment. After using Mandometer, patients are trained to eat socially. 2. Patients rest in warm rooms after each meal. Warmth has an anxiolytic effect and prevents compensatory hyperactivity and vomiting after the meal. 3. Physical hyperactivity is prevented by use of wheelchairs and warming jackets between meals. 4. Patients are trained to re-learn social skills. Manuals describing the details of treatment are used at all clinics. The patient can adapt her or his eating rate to a linear curve displayed on the monitor (A), because she or he sees her or his own eating rate emerging (B). At regular intervals, the patient rates her or his feeling of fullness on a scale that appears on the monitor (C) and adapts her or his ratings to an s-shaped curve (D). The axes have no numerical values during training. 2 Body mass index (BMI) and obsession, anxiety and depression scores in six Australian anorexic patients treated to remission and in 10 anorexic patients treated to partial remission Psychiatric symptoms were evaluated by the Comprehensive Psychopathological Rating Scale.3 Values are medians and ranges. There were no significant differences between the groups.

John Court · Cecilia E K Bergh · Per Södersten

To the Editor: Senna is widely used in laxatives, but the results of its misuse are not inconsequential. We describe a 36-year-old woman admitted with hypercalcaemia and renal failure. She had a 6-year history of anorexia nervosa and ingestion of 50–100 senna tablets daily for weight loss. Examination revealed clubbing of the fingers and toes, a body mass index (BMI) of 17.7 kg/m2 and postural hypotension. Laboratory findings on admission are shown in Box 1. Results of autoimmune studies and protein electrophoresis, and the serum angiotensin-converting enzyme level were normal. Parathyroid hormone-related peptide was absent. She had a bland urinary sediment, trace proteinuria (150 mg/24 h; reference range, < 150 mg/24 h) and a urine pH of 5.0 (physiological range, 4.5–8.0). Computed tomography scans did not detect malignancy or infection, but showed bilateral medullary renal calcifications. Renal biopsy confirmed extensive nephrocalcinosis and the absence of primary glomerular disease. A skeletal survey showed prominent periosteal reaction and new bone formation at the ends of long bones (Box 2, A). A bone scan revealed increased tracer uptake in a pattern consistent with hypertrophic osteoarthropathy (HOA; Box 2, B). Interestingly, bone mineral density (BMD) scans showed increased lumbar and femoral T scores (1.2 and 1.3, respectively). Four years later, the renal failure, clubbing and HOA persisted despite a reduction in senna intake. Low urine volume is a prerequisite for urolithiasis, but hypercalcaemia is the key requirement for nephrocalcinosis.3 The suppressed parathyroid hormone level and elevated serum calcium level excludes primary hyperparathyroidism and made familial hypocalciuric hypercalcaemia unlikely. The serum and urine biochemistry was inconsistent with thiazide diuretic use or renal tubular acidosis. Hence, exogenous calcium is the likely cause of hypercalcaemia. Each of the senna (calcium sennosides) tablets the patient ingested contained 12.5 mg of calcium. Chronic ingestion, in addition to dehydration (with low calcium excretion) and a low BMI may contribute to a vicious circle of calcium phosphate retention, renal failure and nephrocalcinosis. Indeed, hydration increased her calcium excretion to 6.23 mmol/day and normalised her serum calcium and phosphate levels after a week. The association between finger clubbing and senna misuse, and the reversibility of finger clubbing, were reported in 1975.4 Several reports have followed, but only one noted concurrent HOA on plain x-rays.5 We believe that our report is the first to show the extent and distribution of HOA related to this disorder. It remains unknown whether HOA is reversible with abstinence from senna. Patients with anorexia are also more likely to have a low BMD, and the increased BMD scores seen in our patient might be the result of metastatic calcification or periosteal new bone formation. Neither could be conclusively proven. The development of nephrocalcinosis in anorexic patients is more common than is generally appreciated. Clubbing and HOA are useful clues to senna misuse, and BMD measurements should be interpreted with caution in this setting. 1 Biochemical parameters on admission before treatment Parameter Patient value Reference range Serum concentrations of: Sodium (mmol/L) 139 135–145 Potassium (mmol/L) 3.2 3.5–5.0 Magnesium (mmol/L) 0.74 0.80–1.50 Chloride (mmol/L) 102 95–107 Bicarbonate (mmol/L) 24 21–30 Urea (mmol/L) 8.6 2.5–7.8 Creatinine (μmol/L) 166 40–120 Albumin (g/L) 28 35–45 Corrected calcium (mmol/L) 2.93 2.20–2.60 Ionised calcium — pH adjusted (mmol/L) 1.51 1.14–1.29 Phosphate (mmol/L) 2.58 0.80–1.20 25-hyroxyvitamin D (nmol/L) 23 55–108 Parathyroid hormone (pmol/L) 1.0 1.1–7.7 Rates of: Creatinine clearance (mL/min) 20 90–150 Urine sodium excretion (mmol/day) 28 40–100 Urine potassium excretion (mmol/day) 29 50–140 Urine calcium excretion (mmol/day) 1.56 2.0–7.5 2 Ankle x-ray and full bone scan

Andy K H Lim · David H Hooke · Peter G Kerr

To the Editor: The editorial by Woollard and Newman1 discussing the best management of stable coronary artery disease is welcome and timely, but we believe their conclusions undervalue the benefits of optimal medical therapy. The authors correctly point out that the early trials comparing surgery with medical treatment did not include aspirin, β-blockers or lipid-lowering drugs in the medical treatment group. Clearly, the benefits of these treatments have been so dramatic as to make these original trials irrelevant to current practice. After reviewing trials comparing percutaneous coronary intervention (PCI) with coronary artery bypass graft (CABG) surgery and discussing their limitations, the authors conclude that: The best interpretation of currently available data is that, for patients with severe coronary artery disease, the more invasive procedure with a longer recovery time (CABG) has a better long-term clinical outcome and is more cost-effective than the less invasive fast-recovery procedure (PCI). We believe this conclusion is only appropriate for those patients in whom an adequate trial of medical therapy has failed to achieve sufficient relief of anginal symptoms. Based on current evidence, including that from the MASS II2 (comparing medical therapy with both CABG and PCI), COURAGE3 (comparing PCI with medical therapy) and AVERT4 (comparing intensive lipid-lowering therapy using atorvastatin with PCI) trials, we conclude the following about the management of stable coronary artery disease: Irrespective of whether patients under-go CABG or PCI (or neither), aggressive medical therapy combined with appropriate lifestyle measures is the means by which patients are protected from major adverse cardiovascular events. There is no evidence that either PCI or surgical treatment is superior to the current best medical treatment in preventing death or myocardial infarction. Patients with stable angina should be informed accordingly and be reassured that their symptoms are likely to abate or resolve with optimal medical treatment over a period of time. Intervention should be reserved for patients in whom an adequate trial of medical treatment has failed to relieve symptoms or in whom medication is poorly tolerated, and possibly for patients in whom satisfactory anti-atherosclerosis treatment targets cannot be achieved. Performing PCI and CABG in patients with stable angina consumes a significant portion of the Medicare budget. We believe many patients with absent or relatively minor symptoms consent to these procedures in the mistaken belief that their lives will be prolonged. Application of evidence-based guidelines would significantly reduce the number of interventions currently performed and make much needed resources available for prevention strategies and for selected patients with acute coronary syndromes in whom this expensive treatment has proven benefit. Lastly, we strongly support the implication by Woollard and Newman that patients should “undertake consultation with other providers” regarding treatment options — including medical therapy.

Richard W Harper · Esther M Briganti · Brett H R Forge

In reply: We agree with Harper and colleagues that for most patients with stable coronary artery disease, prevention of death or other cardiac events is best achieved with medical treatment. Many patients and even some doctors are so impressed by the restoration of coronary blood flow achieved with angioplasty that they incorrectly assume it will provide a prognostic benefit. However, we do not agree that intervention should be reserved for patients in whom medical treatment has failed. It is unlikely that a trial comparing current medical treatment with coronary artery bypass graft (CABG) surgery for severe left main coronary disease will ever be performed, and current evidence favours CABG at least in this group. In addition, many patients with angina prefer a mechanical intervention to long-term antianginal medications, and why should they be denied that choice?

Keith V Woollard · Mark A J Newman

To the Editor: In Australia, the rising cost of the Pharmaceutical Benefits Scheme (a comprehensive system for subsidising prescription medicines for the whole population) has led to the Australian Government instituting a number of cost-saving strategies. One of these strategies is to encourage the use of generic medicines. However, generic prescribing is a contentious issue among Australian prescribers. The debate has centred on issues related to bioequivalence, quality and safety.1,2 Further, previous studies have shown that changing existing prescribing behaviour is difficult.3,4 To avoid the difficulty of having to change existing prescribing behaviour, education about the benefits of generic prescribing should be aimed at medical students, who are the prescribers of the future.5 To explore and evaluate senior (final-year) medical students’ perceptions of and knowledge about generic medicines and generic prescribing, we undertook a nationwide, web-based survey of senior medical students in Australian universities from 18 June to 18 September 2004. Of 1497 senior medical students in 10 universities throughout Australia, 400 (26.7%) responded to the survey. The first part of the questionnaire required students to select the correct bioequivalence limits allowed by the Therapeutic Goods Administration when comparing a generic medicine with a brand-name medicine. We compared responses to detect differences according to sex, graduate (students with a Bachelor or higher degree on entry to medical school) versus non-graduate status, and between the 10 universities using Fisher’s exact test. For the first question, six options were given, with the correct answer being 80%–125%. Most respondents (64%) did not answer the question, and only three (0.8%) answered it correctly. Responses to other individual questions in the web survey are shown in the Box. Seventy-one per cent of respondents (287) thought that they needed more information on how bioequivalence tests are conducted. More than 90% of respondents did not believe that generic medicines registered in Australia are equal in terms of quality and efficacy to their brand-name counterparts. A similar proportion of respondents thought that generic medicines would cause more side effects than brand-name medicines. Eighty-seven per cent highlighted pharmacists as one of the most important health care professionals to advise them on generic medicines. Almost 92% of respondents agreed that their future prescribing habits would be likely to be influenced by their senior colleagues and medical consultants. More than 60% believed that their respective university did not adequately cover the topic of cost-effective prescribing in their medical curriculum. Our survey clearly shows that medical students in Australia need to be better taught about issues relating to generic medicines and generic prescribing. Modifying existing curricula to include education about generic medicines will be a critical strategy towards promoting rational use of generic medicines by medical practitioners. Responses to questions assessing knowledge and perceptions of senior medical students in Australia about generic medicines and bioequivalence Responses P (Fisher’s exact test) Survey question/statement Strongly agree Agree Neutral Disagree Strongly disagree Sex Graduate status* University† All generic products of a particular medicine that are rated as “generic equivalents” are: therapeutically equivalent to the innovator brand product 66 (16.5%) 276 (69.0%) 18 (4.5%) 36 (9.0%) 4 (1.0%) 0.692 0.094 0.096 therapeutically equivalent to each other 4 (1.0%) 49 (12.3%) 23 (5.8%) 263 (65.8%) 61 (15.3%) 0.846 0.998 0.107 I have not been introduced to the issues of bioequivalence for generic drugs during my medical education 120 (30.0%) 153 (38.3%) 47 (11.8%) 69 (17.3%) 11 (2.8%) 0.703 0.893 0.010 I need more information on how bioequivalence tests are conducted for generic medicines 73 (18.3%) 214 (53.5%) 74 (18.5%) 34 (8.5%) 5 (1.3%) 0.036 0.572 0.195 A generic medicine is bioequivalent to a brand-name medicine 79 (19.8%) 270 (67.5%) 18 (4.5%) 30 (7.5%) 3 (0.8%) 0.574 0.943 0.827 A generic medicine must be in the same dosage form (eg, tablet, capsule) as the brand-name medicine 27 (6.8%) 106 (26.5%) 52 (13.0%) 196 (49.0%) 19 (4.8%) 0.751 0.677 0.152 A generic medicine must contain the same dose as the brand-name medicine 47 (11.8%) 171 (42.8%) 44 (11.0%) 126 (31.5%) 12 (3.0%) 0.919 0.905 0.048 Generic medicines are of inferior quality to brand-name drugs 148 (37.0%) 226 (56.5%) 20 (5.0%) 5 (1.3%) 1 (0.3%) 0.006 0.944 0.054 Generic medicines are less effective than brand-name medicines 151 (37.8%) 228 (57.0%) 19 (4.8%) 1 (0.3%) 1 (0.3%) 0.029 0.124 0.010 Generic medicines produce more side effects than brand-name medicines 148 (37.0%) 222 (55.5%) 26 (6.5%) 3 (0.8%) 1 (0.3%) 0.052 0.619 0.023 Generic medicines are less expensive than brand-name medicines 165 (41.3%) 215 (53.8%) 9 (2.3%) 9 (2.3%) 2 (0.5%) 0.190 0.154 0.072 Brand-name medicines are required to meet higher safety standards than generic medicines 112 (28.0%) 219 (54.8%) 42 (10.5%) 22 (5.5%) 5 (1.3%) 0.747 0.843 0.162 * Comparing graduate (those with a Bachelor or higher degree on entry to medical school) with non-graduate students. † Differences between the 10 universities represented.

Mohamed Azmi Ahmad Hassali · Kay Stewart · David C M Kong

To the Editor: In a recent article, Youl and colleagues provided information about the ability of doctors to accurately diagnose skin lesions that they excise or biopsy.1 We wish to offer some comments about their comparison between general practitioners and skin cancer clinic doctors. First, in the study by Youl et al the behaviour of GPs and patients in mainstream practice was different from that of doctors and patients in skin cancer clinics, as indicated by the comparative frequency of whole body skin examinations performed (GPs, 30.4%; skin cancer clinic doctors, 73.2%).1 The study did not indicate the circumstances under which each decision to excise took place. Did patients become aware of a new or changing skin lesion and bring it to the attention of the doctor, or did the diagnosis result from a whole body skin check that might have revealed an earlier, previously unnoticed and more subtle lesion? It may be useful to separate basal cell carcinoma (BCC) and squamous cell carcinoma (SCC) into histological subtypes, as early superficial BCC and intraepidermal SCC may be more difficult to diagnose than other subtypes.2 Second, the casemix of non-melanotic skin cancers in the two groups of doctors was quite different, with a BCC : SCC ratio of 1.1 : 1 for GPs and 2.1 : 1 for skin cancer clinic doctors. The difference in casemix was reflected in a study of our own3 in which we described the histology of 1247 lesions excised by doctors, including 76 lesions removed by one doctor in a designated skin cancer clinic. In an unpublished sub-analysis, we divided the results into two settings for comparison (Box). Like Youl et al, we found that the casemix of non-melanotic skin cancers was significantly different for the two groups of doctors (P < 0.001), but in our study the BCC : SCC ratio was much higher for skin cancer clinic doctors (4 : 1) than for GPs (0.6 : 1). We believe this most likely reflects an increased pick-up of BCC in skin cancer clinics, owing to the higher frequency of full body skin examinations and the consequent detection of lesions of which the patient is unaware. Third, the reported sensitivity and specificity in the study by Youl et al refers only to excised lesions. There is no information given about the lesions that practitioners decided not to excise. The sensitivity and specificity of skin examinations can only be determined if all relevant skin lesions are assessed, thereby giving an accurate representation of the number of true- and false-negative diagnoses. However, this would require multiple excisions, which would be clinically unacceptable. An important limitation of the study is that it does not assess or compare how many skin cancers each group of doctors missed. In conclusion, although the study by Youl et al provides comprehensive information about diagnostic accuracy, we do not feel — based on the information available — that a meaningful comparison between the two groups of doctors can be made. Comparison of lesion excisions in skin cancer clinic and general practice settings Mean patient age (years) Proportion of excised lesions that were malignant* BCC : SCC ratio Number needed to treat† Skin cancer clinic 56.5 76% (58/76) 4 : 1 (44/11) 4.7 (14/3) General practice 56.9 45% (512/1145)‡ 0.6 : 1 (190/305) 9.0 (154/17) BCC = basal cell carcinoma. SCC = squamous cell carcinoma. * BCC, SCC or melanoma. † Benign or dysplastic naevi excised per melanoma. ‡ There were 26 cases in which histology results were missing. All were in the general practice setting.

Clare Heal · Beverly Raasch

To the Editor: Youl and colleagues are to be commended for their research into the performance of special-interest skin cancer clinicians.1 However, the conclusion that the performance of general practitioners and skin cancer doctors in the diagnosis of skin cancer is similar is highly questionable. To truly compare the two groups and their diagnostic accuracy, it must be established that the participants were representative of the groups they are supposed to represent. The fact that the participating GPs were largely self-selected, perhaps on the basis of their personal interest in the subject, was a potential flaw that the authors acknowledge. Further, an examination of diagnostic accuracy should also take into account whether the lesions found were of similar type, size and stage. There was no determination of any qualitative differences (eg, in size or thickness) between the tumours seen and diagnosed by the two groups. When a patient presents, specifically, with a large, tender, hyperkeratotic squamous cell carcinoma (SCC), there is no real test of diagnostic skills. On the other hand, detecting a small early posterior-thigh melanoma or a superficial BCC on a whole body examination is a challenge. Overall, one would expect at least two to three BCCs to be diagnosed for each SCC found2 — however, in the study by Youl et al, GPs found a similar proportion of each type of lesion, suggesting that perhaps a large number of BCCs were not being detected at all in the GP group. The lower incidence of whole body examinations in the GP group suggests that a higher proportion of asymptomatic lesions may have been missed by the GP group and thus not included in their sensitivity/specificity data. This possibility could have been examined by noting the site of lesions found: identification and diagnosis of lesions in areas covered by clothing or footwear may be more likely on whole body examination. Although Youl et al reported that the diagnostic sensitivity for melanoma among skin cancer clinic doctors was twice that of GPs, a re-examination of the data with all of the above in mind may well reveal that skin cancer clinic practitioners are performing even better than suggested. That being said, the number of melanomas found per week by the skin cancer group (0.25 melanomas/doctor/week) in the study by Youl et al is much lower than in our own dedicated primary care skin cancer clinic (1.47 melanomas/doctor/week, based on audit data gathered between February and September 2007). Good medical care relies on accurate diagnosis and appropriate treatment, and the earlier the diagnosis is made, the less costly and less invasive the treatment and the greater likelihood of a cure. Surely we should all be working together — skin cancer clinic practitioners and GPs, alike — towards achieving optimum outcomes for our patients. To that end, a wider range of higher education in skin cancer medicine is being encouraged by the University of Queensland’s Master of Medicine program3 and the Skin Cancer College of Australia’s fellowship program,4 and bodies such as the Skin Cancer Society of Australia and the Royal Australian College of General Practitioners have been working together to introduce an accreditation process for those practising primary care skin cancer medicine.5,6

Jeffrey Keir

In reply: We thank Heal and Raasch, and Keir for their comments and suggestions. First, as acknowledged in our article, it is possible that general practitioners with an interest in skin cancer may have been over-represented in our sample. It is also the case that the comparison between GPs and skin cancer clinic doctors may have been affected by characteristics of patients attending each type of practice. Patients attending skin cancer clinics are self-selected (often worried about a specific skin lesion), while for those attending GPs in mainstream practice, skin lesions are more likely mentioned during a consultation for something else.1 Case selection, which helps increase diagnostic ability by Bayesian principles (improving pretest probability of malignant lesions), may thus be more likely for skin cancer clinic doctors than GPs.2 Our study represented a broad cross-section of skin cancer clinics, and the number of melanomas excised per doctor per week ranged from 0 to 1.7. The ratio of basal cell carcinomas to squamous cell carcinomas excised ranged from 0.1 to 6.0 for mainstream GPs and 0.8 to 8.0 for skin cancer clinic doctors. Whether clinical and histological features of skin lesions or the type of skin examination undertaken influence diagnostic accuracy was beyond the scope of our initial study. This question will be the subject of future analyses. It has been suggested that a limitation of our study was that it did not assess or compare the number of skin cancers each group missed. However, as specifically stated in our article, the aim of our study was to examine diagnostic accuracy of excised or biopsied lesions. To examine the sensitivity and specificity of all lesions (excised and non-excised) and of screening examinations would require a different study design. We disagree that meaningful comparisons between the two groups cannot be made from the data collected in our study. Our prospective study included over 11 000 skin excisions or biopsies from a large group of mainstream GPs and from doctors working in a variety of skin cancer clinics. One of the most important outcome measures of our study was the degree of accuracy of skin cancer diagnoses within the primary care setting. We have demonstrated that primary care practitioners, whether mainstream GPs or skin cancer clinic doctors, diagnose skin cancer with similar, high levels of accuracy. This is a reassuring result, particularly in a country with the world’s highest incidence of skin cancer.

Philippa H Youl · Peter D Baade · Monika Janda · Christopher B Del Mar · David C Whiteman · Joanne F Aitken

To the Editor: Dabadghao and colleagues have drawn attention to the high prevalence of diabetes mellitus and impaired glucose tolerance (IGT) among Australian women with polycystic ovary syndrome (PCOS).1 Quite rightly, both in this article and in the accompanying editorial by Teede and Stuckey,2 the authors have stressed the need to test women with PCOS for diabetes, and that long-term follow-up and lifestyle modification are important. However, it is important to note that many of these women were attending a reproductive endocrinology clinic for treatment of infertility. Therefore, two other points need to be highlighted. First, the high prevalence of glucose intolerance among this group is of immediate concern, as type 2 diabetes can pose a considerable risk to a pregnancy.3 The dangers include miscarriage, fetal malformation and perinatal mortality. Data from New Zealand indicate that perinatal mortality in pregnancies of women with type 2 diabetes was about triple that in pregnancies of women without diabetes; and, in women with type 2 diabetes which was not detected until after pregnancy was achieved, perinatal mortality was 4.5 times that of women without diabetes.4 Screening for diabetes therefore provides an opportunity to ensure that women with glucose intolerance are adequately prepared for pregnancy, thereby minimising the risks. Important measures include high-dose folate therapy and tight glycaemic control, in accordance with guidelines previously published in the Journal.5 Second, the risk of adverse pregnancy outcomes related to the women’s obesity (mean body mass index, 35.1 kg/m2) is also of concern. Therefore, stringent efforts should be made to take the opportunity presented to normalise the weight of women with PCOS before conception.6 The data presented by Dabadghao and colleagues also provide insights into the most appropriate means of screening for diabetes in this high-risk population. They have shown that screening by fasting glucose level alone will miss a third of cases of diabetes. Furthermore, at least 81% of IGT will also be missed! Glucose tolerance testing is therefore far superior to fasting glucose level for the detection of diabetes and IGT. This is a critical point, given the potential ramifications of undetected diabetes. Given these findings, we urge clinicians to routinely perform a glucose tolerance test for women with PCOS attending for fertility treatment, so that all cases of abnormal glucose tolerance are detected, and appropriately managed, before and after conception. It would be a tragedy for a woman to achieve a pregnancy through in-vitro fertilisation, only to develop complications from undiagnosed diabetes.

N Wah Cheung · Jeremy J N Oats

To the Editor: Drug-eluting coronary stents (DES) were approved for use in Australia in 2002, after they were found to be effective in limiting the incidence of restenosis.1 However, their cost is three to four times that of traditional bare-metal stents, and most Australian states restricted their use in the public health system. In 2006, we reported in the Journal that DES were used in 45% of patients undergoing percutaneous coronary intervention (PCI) in Victorian public hospitals, and that they were largely reserved for patients with risk factors for restenosis, such as diabetes, small vessels, and complex lesions.2 However, in the private health system, DES can be claimed as prostheses from insurance funds, so their use is not limited by financial constraints. In early 2007, multiple reports from around the world noted a small but significant increase in late stent thrombosis occurring 1–4 years after implantation of DES.3,4 This caused concern, as stent thrombosis has a mortality rate approaching 50%. It is now recommended that all patients with DES maintain aspirin and clopidogrel therapy for at least a year, although the appropriate duration of this treatment remains unknown.5 Further, long-term dual antiplatelet therapy is associated with an increased risk of bleeding and is problematic for patients requiring surgery.4,5 We examined the use of DES from February 2006 to June 2007 in 674 patients undergoing PCI by 10 operators in a Victorian private hospital. Monthly use dropped dramatically, from a peak of 91% of patients in July 2006 to 34% in June 2007 (Box). In 118 PCIs performed between March and May 2007, DES were used more often in: diabetic than non-diabetic patients (67% v 47%; P = 0.05); small (≤ 2.5 mm) vessels compared with large vessels (74% v 53%; P = 0.04); and long lesions (> 20 mm stent length) compared with shorter lesions (62% v 52%; P = not significant). This mirrors the pattern previously seen only in patients treated in the public health system.2 Given the issue of late stent thrombosis and the need for prolonged dual antiplatelet therapy, reserving DES for patients who are at high risk of restenosis in both the public and private health systems seems very appropriate. Utilisation is likely to change again with a better understanding of the long-term safety of DES, but a cautious approach should be maintained. Drug-eluting stent (DES) use per month (February 2006 – June 2007) in patients in a Victorian private hospital

Suzanne M McLean · David J Clark

To the Editor: After the recent New South Wales parliamentary inquiry into Royal North Shore Hospital, one might ask whether the deliberations will deliver real change to the hospital system. An article in the Lancet draws attention to some adverse outcomes in London hospitals.1 One such case is reported from the evidence and conclusions of the Islington Coroners’ Court into the death of a man as a result of his outpatient management. The report states: The jurors have . . . taken into consideration the large number of patients attending Guy’s Hospital as casual patients, [and] are of the opinion that the skilled supervision is insufficient, and should be increased; but they . . . believe the deceased received all the care and attention which the present arrangement of the hospital affords. The report goes on to discuss the effect of the excessive numbers of patients each house surgeon is expected to manage daily. The writer is of the mind that “the reported cases are now so frequent that something will have to be done”. As for immediate solutions, the report suggests: The officers of hospitals will have to be a little more careful. The committees and governors of hospitals must not impose impracticable quantities of work on one man’s shoulders; they must take steps either to diminish the amount of casual out-patient work, or they must increase the number of competent and qualified persons. The article concludes that “there is another party in these questions who is not entirely free of blame — viz., the public”. The writer supports the public’s right to “express indignation at defective diagnosis and accommodation in hospitals” but implores the public to be more liberal with hospital funding. The conclusions are straight to the point: The committees of hospitals cannot multiply beds and qualified house-surgeons without funds, any more than one can make bricks without straw. And if the maimed and the diseased are not to be denied skilled advice and accommodation at hospitals, or are not to be passed from one hospital to another, the public must be far more liberal in its support of hospitals than it has been. This report was published in the Lancet on 20 August 1881. Has nothing changed?! The public has now relinquished its responsibility for running public hospitals to the state. I hope more permanent solutions to the problem will follow the recent inquiry! As the writer suggested in 1881, “something will have to be done”.

Catherine E Storey

To the Editor: We report elevated mercury levels in three infants, each the only child of Chinese parents living in Sydney. All three children had eaten fish congee (a rice and fish porridge) as a weaning food and ate fish regularly as toddlers. Their parents had sought medical advice for either developmental delay or neurological symptoms in the children. A 2-year-old boy had demonstrated increasingly aggressive behaviour for the past 6 months. A general practitioner had diagnosed mercury poisoning in the boy’s father 2 months earlier, following investigation for complaints of allergies, rashes, abdominal pain and diarrhoea. The family ate fish (usually salmon, barramundi or snapper) at least five times a week, and had used unspecified herbal medicines in the past. The child had eaten fish regularly since weaning. The boy’s blood mercury level was 158 nmol/L (normal range [NR], < 50 nmol/L), a random urine mercury/creatinine (Hg/Cr) ratio was 9 nmol/mmol (NR, < 6 nmol/mmol*), and his hair mercury level was 1.42 mg% (NR, < 0.18 mg%). The boy’s father and mother (who was pregnant) also had elevated hair mercury levels, of 4.3 mg% and 6.0 mg%, respectively. The father and child were treated with chelation therapy elsewhere. * The two laboratories reported different normal ranges for the urine Hg/Cr ratio. A boy aged 2 years and 10 months presented with delayed speech and some autistic features. Since weaning, he had eaten fish (barramundi, sea perch, salmon and rock cod) up to eight times a week. He had no history of herbal medicine use, and his thyroid function, blood lead level, and a DNA screen for fragile X syndrome were normal. The child’s blood mercury level was 350 nmol/L and urine Hg/Cr ratio was 14 nmol/mmol (NR, < 10 nmol/mmol*). The boy’s father did not eat fish, and his blood mercury level was 19 nmol/L. The child’s mother did eat fish, and had a blood mercury level of 27 nmol/L. Two weeks after removing fish from the diet, the child’s blood mercury level had fallen to 99 nmol/L and his urine Hg/Cr ratio to 7 nmol/mmol. However, his behaviour did not improve, and he was subsequently diagnosed with classical autism. A 15-month-old boy presented with delayed development since birth. Fish had been introduced to his diet at 8 months of age, and he had since continued to consume fish four to five times a week. He had recently eaten either ling or salmon. The boy’s mother had consumed ling three to four times a week after the fifth month of her pregnancy. The child’s thyroid function, DNA screen for fragile X syndrome, chromosome karyotype, and urinary metabolic screen were normal. His blood mercury level was 143 nmol/L, but fell to 19 nmol/L over a period of 1 year after ceasing fish intake. His longer-term developmental status is unknown. Fish congee, made with either freshwater species or locally caught fish, is a common weaning food in coastal regions of southern China and South-East Asia.1 Adding fish to the weaning diet has health benefits,1,2 such as reducing anaemia, and is actively promoted. However, fish, particularly the large pelagic (open ocean) species more likely to be bought in Australia, may also contain mercury. Excessive consumption of mercury has been associated with neurological impairment.3-5 The Box shows that the consumption by infants of fish congee made from portions of large fish species may exceed the provisional tolerable weekly intake (PTWI)7 for methylmercury of 1.6 μg/kg bodyweight/week (the limit considered sufficient to protect a developing fetus). Food Standards Australia New Zealand’s most recent risk assessment concluded that median-level consumers of fish are unlikely to exceed the PTWI for methylmercury,6 but frequent consumers might if all their consumption is of predatory or long-lived fish species, which tend to acccumulate higher concentrations of mercury. It has been previously noted in the Journal that public health policy regarding fish consumption needs to balance the health benefits for cardiovascular disease and anaemia with the possible ill effects of mercury on neurological development in infants.8 We recommend that multilingual information about fish and mercury be made available to pregnant women and mothers, especially targeting groups who are likely to be frequent consumers of fish and who use fish in weaning and infant foods. Regulatory and health promotion activities could also be informed by surveillance of blood or hair mercury levels in infants from ethnic groups at high risk of mercury intoxication, and of the frequency of fish consumption in this age group (by type of fish). Estimated weekly mercury intake in infants consuming fish congee Mean mercury concentrations in fish tissue* (μg/kg fish) Child’s estimated weekly mercury intake† (μg/kg bodyweight/week) Fish fillets‡ Maximum 50 1.25 Median 16 0.40 Minimum 5 0.13 Barramundi Maximum 310 7.75 Minimum 40 1.00 Snapper Maximum 520 13.00 Minimum 52 1.05 * For species consumed in Australia.6 † For a 12-month-old child weighing 10 kg; weekly fish consumption is estimated to be 0.25 kg, assuming 50 g servings five times per week. Provisional tolerable weekly intake for methylmercury = 1.6 μg/kg bodyweight/week.7 ‡ Average concentrations of all fillets purchased.

Stephen J Corbett · Christopher C S Poon

To the Editor: Prostate cancer screening remains controversial. There is currently a lack of evidence that treating prostate cancers identified by screening leads to prolonged survival,1 and the main screening test (serum prostate-specific antigen [PSA] concentration) has poor sensitivity and specificity. Patients with an elevated PSA level usually undergo a transrectal ultrasound-guided (TRUS) prostate biopsy for definitive diagnosis. TRUS biopsy is associated with risks that include bleeding, urinary tract infections, prostatitis and bacteraemia. Infective complications can occur even when prophylactic antibiotics are administered.2 Gram-negative bacteraemia is usually associated with a mortality rate of at least 5%.3 We reviewed prostate biopsy-associated bacteraemia in Australian Capital Territory residents using data collected over the 5-year period from 2002 to 2006. All patients in ACT public hospitals who have positive blood culture results are followed as part of a bloodstream infection surveillance program. Although all prostate biopsies are performed in the private sector, nearly all patients who develop major complications are cared for in public hospitals. The number of biopsies performed on ACT residents during the study period was obtained from Medicare Benefits Schedule statistics (item numbers 37215, 37218 and 37219).4 Over the 5-year period, we identified 19 episodes of bacteraemia following 1843 prostate biopsies, representing a 1.0% risk of bacteraemia (95% CI, 0.6%–1.6%) (Box). No patients died, but five required admission to the intensive care unit. A 1% bacteraemia rate associated with prostate biopsies would be an underestimate of the true rate, for a number of reasons: some patients were also admitted with severe sepsis syndromes but with negative blood cultures; some patients may have been treated in another state; and some patients may have presented to private hospitals, specialists or general practitioners and been treated empirically without blood cultures being collected. Three additional cases of bacteraemia following TRUS biopsies were excluded from our rate calculations because they involved New South Wales residents. It has been estimated that, if a million men over 50 years of age were screened, about 110 000 would have raised PSA levels.5 Of these, about 90 000 would undergo a biopsy and 20 000 would be diagnosed with prostate cancer. Our data show that TRUS biopsies carry a risk of serious infective complications and, notably, about 75% of these complications would occur in men without prostate cancer. We believe that data on the likely complication rates resulting from PSA testing and TRUS biopsies must be factored into considerations of the benefits versus risks and the cost-effectiveness of any prostate cancer screening program. Prostate biopsies and associated bacteraemia episodes, Australian Capital Territory, 2002–2006* 2002 2003 2004 2005 2006 Total Prostate biopsies performed 322 324 465 377 355 1843 Bacteraemia episodes 6 3 2 4 4 19 Proportion of biopsies resulting in bacteraemia 1.9% 0.9% 0.4% 1.1% 1.1% 1.0% * Three New South Wales residents who were diagnosed and treated for bacteraemia in the ACT were excluded (two had biopsies in Sydney and one in the ACT). Two patients had two episodes of bacteraemia (each more than 1 month apart). Fifteen episodes of bacteraemia were caused by Escherichia coli and one each by Klebsiella pneumoniae, Group C Streptococcus, Citrobacter freundii and Proteus mirabilis.

Francis J Bowden · Jan Roberts · Peter J Collignon

To the Editor: The report by Risson and colleagues of a fatal case of necrotising pneumonia caused by community-associated methicillin-resistant Staphylococcus aureus (CA-MRSA)1 appropriately highlights the emerging issue of CA-MRSA infections in Australia,2 and the possibility that severe S. aureus sepsis may follow recurrent furunculosis. We wish to draw attention to the association between severe staphylococcal pneumonia and a preceding influenza-like illness. In September 2006, a 56-year-old woman of European background with a history of chronic back pain and depression presented to the Royal Darwin Hospital after a 4-day influenza-like illness characterised by cough, fever and sore throat. She then developed dyspnoea and pleuritic chest pain, followed by an abrupt respiratory deterioration. She was intubated and admitted to the intensive care unit with severe sepsis. A chest x-ray showed widespread bilateral pneumonia. We began broad-spectrum antibiotic therapy with piperacillin–tazobactam and azithromycin as per the hospital’s dry-season protocol for severe community-acquired pneumonia. Further history from her husband revealed an episode of boils 1 month previously, which responded to antibiotic therapy. We added vancomycin to the therapy and, when sputum and blood cultures showed MRSA 48 hours after admission, we also added rifampicin and gentamicin. On Day 5 of admission, her clinical condition deteriorated further and we replaced rifampicin and gentamicin with linezolid. Complement fixation testing of serum taken on admission showed an influenza A antibody titre of 128, consistent with recent acute infection. Despite ongoing intensive supportive care, the patient died from refractory respiratory failure 10 days after admission. Typing of the S. aureus isolates from blood and sputum showed that their single nucleotide polymorphism and variable gene profile was consistent with the Queensland clone (ST93-MRSA-IV) of CA-MRSA, and that the Panton–Valentine leukocidin gene was present. S. aureus has long been a recognised cause of influenza-associated pneumonia. Of concern, two recent reports from the United States identified 25 patients with CA-MRSA associated with severe pneumonia following influenza-like illnesses.3,4 Most of these patients were young (median age of 21 years3 and 17.8 years,4 respectively) and otherwise healthy. Combined mortality in these two studies was 40%. With an increasing prevalence of CA-MRSA in areas of Australia,2 CA-MRSA pneumonia should be suspected in patients presenting with worsening respiratory status and sepsis following an influenza-like illness. We stress the importance of annual influenza vaccination for those at increased risk of influenza-related complications.5

Steven Y C Tong · Nicholas M Anstey · Gary D Lum · Rachael A Lilliebridge · Dianne P Stephens · Bart J Currie

To the Editor: The recent editorial by Collignon and colleagues emphasised the importance of infection control mechanisms in reducing patient harm from antibiotic-resistant organisms.1 It focused on disinfection of the hands of health care workers in hospitals. However, a vigorous education and surveillance program in a hospital in Victoria failed to achieve compliance among health care workers of even 50%.2 Top of the list of self-reported factors leading to poor compliance is “skin irritation and dryness associated with the use of hand hygiene agents”.3 There have been no properly controlled trials, with clinically important endpoints, of currently recommended hand-hygiene practices. With the likely poor compliance rates, such trials would likely fail. A different approach might be more effective. Reducing skin contact between health care workers, patients and their immediate environment seems logical. Data show that skin contact produces two-step transfer of material in 82% of cases.4 The Victorian study did include gloving as an alternative to disinfection in measuring hand-hygiene compliance.2 However, in what might be a backward step, a recent study concluded that physicians should be encouraged to shake hands with patients!5 Perhaps an educational campaign to avoid skin contact with environmental surfaces and other health care workers, with use of disposable gloves for patient contact, could be the basis of a successful trial to address more effectively the transmission of antibiotic-resistant organisms in hospitals.

Keith V Woollard

To the Editor: The magnitude and distribution of the problem of health care-associated methicillin-resistant Staphylococcus aureus (MRSA) in Australia can be gauged from the report of a forum on MRSA control conducted at the Australasian Society for Infectious Diseases (ASID) in March 2007.1 This report contrasted approaches to control of health care-associated MRSA and quantified the population incidence rate of health care-associated MRSA bacteraemia across Australia from data derived from direct surveillance systems (Box). Reporting of MRSA infections is thought to be complete from all jurisdictions except Victoria and New South Wales. Figures for Victoria were extrapolated from accurate surveillance data representing 50%–60% of events. The degree of incompleteness of reporting in NSW could not be determined, and a range based on reports to NSW Health over 3 years was used. Overall morbidity of health care-associated MRSA in Australia is much higher, as only a minority of MRSA infections lead to bacteraemia. The ASID report estimated that between 699 and 924 cases of bacteraemia would be prevented if other states and territories reduced their incidence of MRSA bacteraemia to that of Western Australia through implementation of more stringent infection control measures. The mortality of MRSA bacteraemia is 8%–50% (average, 29%).2 A recent study showed that more than half (59%) of such deaths were directly attributable to MRSA3 rather than other non-infective causes. These outcome proportions provide a minimum estimate of between 120 and 158 preventable deaths per annum in Australia directly caused by health care-associated MRSA — comparable to the annual South Australian road toll. As identified recently in the Journal by Collignon and colleagues, there are significant structural barriers to achieving infection control — especially inadequate isolation resources and pressure on bed stock.4 Other dimensions of the MRSA problem include the high incidence of MRSA in many aged care facilities, the epidemic emergence of community strains of MRSA (best described in the recent report from WA on MRSA notification data up until 20025), the possibility of significant zoonotic reservoirs,6 and the role played by imprudent antibiotic use. MRSA colonisation or infection needs to be made a nationally notifiable disease, with a system in place to enable typing of isolates. As in WA, such a system would enable more effective identification of MRSA carriers before hospital admission, the detection of emerging epidemic strains, and timely investigation of MRSA outbreaks occurring in community groups, such as in aged care facilities. Most importantly, all states and territories need to adopt, and provide resources for, consistent, stringent approaches to surveillance, prevention and control of health care-associated MRSA that are in keeping with internationally recommended approaches. Given the scale of preventable injury occurring in many states, MRSA control must be made one of the highest priorities for patient safety. Relative burden of health care-associated MRSA morbidity across Australia1 Area Health care-associated MRSA bacteraemia events Year(s) of data Rate per 100 000 population Darwin 16 2006 13.3 New South Wales/ACT*† 437–602 2003–2005 6.2–8.5 Queensland* 133 2005 3.4 South Australia* 37 2006 2.4 Tasmania* 3 2006 0.6 Victoria*† 270–330 2000–2006 5.4–6.6 Western Australia* 22 2006 1.1 Total 918–1143 4.5–5.7 MRSA = methicillin-resistant Staphylococcus aureus. ACT = Australian Capital Territory. * Figures from these jurisdictions include private hospital event estimates. † Figures from NSW and Victoria are minimum estimates, because of incompleteness of current reporting in these states.

John K Ferguson · Helen Van Gessel

To the Editor: The recent editorial by Collignon and colleagues challenged Australian physicians and health care leaders to confront the rising burden of methicillin-resistant Staphylococcus aureus (MRSA).1 Compared with Australia, the United States has a bigger problem with MRSA; more than 60% of all hospital-acquired S. aureus infections are now caused by MRSA.2 Appropriately, the medical community has made an urgent call for action. For example, the Institute for Healthcare Improvement (a not-for-profit organisation based in the US that aims to improve health care throughout the world) incorporated specific MRSA prevention measures into its recent 5 Million Lives Campaign (see http://www.ihi.org/ihi). One of these prevention measures, a recommendation for active surveillance, has generated controversy. Specifically, the cost-effectiveness of this strategy is still vigorously debated in the infection control literature.3 At present, it is unclear what surveillance testing method should be used in the laboratory, and whether testing should be done for all patients or just those identified as high risk. The cry for help from community activists in the US and the United Kingdom has reached the ears of their legislative representatives. In two northern US states, lawmakers are considering bills that require universal active surveillance in their hospitals. Mandating resource-stretched health systems to implement obligatory active screening is not a prudent use of resources. The Society for Healthcare Epidemiology of America and the US Association for Professionals in Infection Control and Epidemiology recently published a joint position paper opposing this legislative activity, noting that data in support of active surveillance have been restricted to high-risk populations.4 We support this position and remind readers that active surveillance does not obviate the need for adherence to basic and consistent hand-hygiene practices. Complacency and lack of clinical leadership remain the greatest challenges in the efforts to reduce the transmission of MRSA. Why do we accept this epidemic as a fact of life as our health care workers complacently contribute to the nosocomial transmission of MRSA? By implementing simple prevention policies, feedback of data on nosocomial transmission of MRSA, and increased infection-control education, we have achieved a 22% reduction in MRSA infections in our network of community hospitals.5 Still, we acknowledge the absence of a zero-tolerance approach to failures in hand-hygiene practices. More needs to be done. We challenge our clinical leaders to demand higher standards for hand hygiene. Most cases of nosocomial MRSA transmission represent failures of basic hygiene practices. The problem is surmountable. Infection control is not a skill of a few, but the responsibility of every team member. The onus is on all of us.

Luke F Chen · Deverick J Anderson · Keith S Kaye · Daniel J Sexton

In reply: We thank Woollard for his comments on hand hygiene. While important, hand hygiene is just one component of what is needed to decrease the spread of methicillin-resistant Staphylococcus aureus (MRSA) in hospitals. Decontamination of the environment, contact precautions for colonised patients, active surveillance and screening, effective programs to prevent common infections such as intravascular catheter sepsis, good antibiotic stewardship and better hospital design are also indispensable.1 We do not accept that “There have been no properly controlled trials, with clinically important endpoints, of currently recommended hand-hygiene practices”. For instance, the recent study quoted by Woollard showed that hand hygiene reduced infections hospital-wide, using the clinically important endpoint of serious bloodstream infections (MRSA and antibiotic-resistant gram-negative bacteria).2 Thus, at least two large peer-reviewed studies show that alcohol-based hand-hygiene programs reduce hospital-acquired MRSA infections2,3 (Level III evidence4). There is nothing wrong with using disposable gloves, as suggested by Woollard, provided they are changed every time a health care worker moves between patients. Otherwise, gloves spread MRSA just as efficiently as unclean hands. Applying good-quality hand-hygiene products is less cumbersome than changing gloves and also allows direct human contact, which Woollard reminds us is important to patients. How MRSA is spread in hospitals is now well known — our problem is getting health care workers to remember to practise good hand hygiene all the time, every day, before and after every patient contact. The data shown by Ferguson and Van Gessel reaffirm how common and serious a problem we have with MRSA bacteraemia. They estimate there are about five episodes per 100 000 people annually across Australia. However, we believe that the true rate is double this. The rate of S. aureus bacteraemia (ie, MRSA and methicillin-sensitive S. aureus combined) in Australia is around 35 per 100 000 per year,5 with 27% caused by MRSA.5 This crudely translates to an MRSA rate of 9.5 per 100 000. The rates are probably much higher in the states with more health care-acquired MRSA (New South Wales and Victoria). More recent data suggest that 36% of hospital-acquired invasive S. aureus infections were caused by MRSA, with the highest percentages in NSW (41%) and Victoria (39%).6 It is also worth noting that, when MRSA bacteraemia became notifiable in England in 2001, there was a 50% increase in reported S. aureus bacteraemia episodes.1 This suggests that under-reporting is common in any voluntary reporting scheme, and is also likely for Australian data. Worryingly, Chen and colleagues point out that MRSA is an even bigger problem in the United States than in Australia. However, we are not far behind.6 Although we share their concerns about legislative impositions, some external controls and measurements can be an advantage. Western Australia, the only Australian state where MRSA is notifiable, has the lowest rates of health care-associated MRSA. While we do not want imposed “one size fits all” legislated controls, we do need change: the practice of the past 40 years has not worked. Every institution needs to have an effective MRSA control program, with components chosen according to the local situation. Institutions should measure MRSA and report centrally, especially if their rates are high and not falling continually (eg, over a year). Shop-floor quality improvement programs with empowered workers are much better than top-down management-imposed regulation (eg, from government). We need to shake our complacency and that of our health care colleagues, accept clinical leadership and take control. Otherwise, legislative controls will be imposed on us.

Peter J Collignon · M Lindsay Grayson · Paul D R Johnson

To the Editor: Xue and colleagues reported an interesting randomised, single-blind trial of acupuncture for persistent allergic rhinitis (PAR) and concluded that acupuncture is an effective treatment for this condition.1 I am not entirely sure that this is true. The authors state that “once needling sensation (known as de-qi) was obtained, the needles were manipulated . . .”. In the sham group they inserted needles at non-acupuncture points where, according to acupuncture theory, no de-qi can be elicited. Thus the intervention patients were experiencing de-qi, and the control patients were not. This means that neither the patients nor the therapist were blinded. Consequently, the difference in outcome between the two groups could be unrelated to acupuncture itself, and caused by patient expectation, therapist expectation or both. In addition, the statement of Xue et al that “no other randomised controlled trial of acupuncture in adults with PAR has been reported in the English medical literature”1 was misleading. Our review of this topic2 included no fewer than six randomised controlled trials, all either published in English or, in one case, with an English abstract. Interestingly, three of them suggested acupuncture to be effective, while three failed to do so. I fear that the study by Xue et al does little to resolve this intriguing discrepancy.

Edzard Ernst

In reply: Ernst is concerned about the sham acupuncture procedure used in our trial.1 Although not universally agreed, the sham/placebo control we adopted has been described as best practice.2,3 The assumption that, without de-qi, participants would not be blinded is not supported by the literature. In fact, a sham procedure without de-qi was used in a recent trial reported by Ernst and colleagues on acupuncture for subacute stroke rehabilitation.4 In that study, as in ours, to increase the credibility of blinding, participants with previous experience of acupuncture were excluded, and those assessing the outcomes of treatment were blinded. With regard to previous randomised controlled trials of acupuncture for adults with persistent allergic rhinitis in the English medical literature, Ernst failed to distinguish between seasonal allergic rhinitis (SAR) and persistent allergic rhinitis (PAR).5 Of the six studies included in his review,5 five were on SAR, and the sixth, which was cited in our report (reference 14), was on children with PAR. These reports, therefore, are not inconsistent with our findings.

Charlie C L Xue · David F Story

To the Editor: Cannabis use is implicated in serious social disruption in many Northern Territory Aboriginal communities.1 Rising levels of cannabis use were first reported in Aboriginal communities in Arnhem Land in 2002, along with associated concerns about escalating social impacts and mental health effects compounded by other substance use.2 A random sample of 164 people in Arnhem Land initially interviewed and assessed in 2004 were followed up between October 2005 and June 2006. Their cannabis use was measured using health worker assessments and self-reports from interviews. Ethical approval was granted by the NT Health Department, Menzies School of Health Research, and James Cook University. Despite a modest decline in cannabis use in this population between 2002 and 2004,3 the 2005–2006 data indicate persisting high rates, with 61% of males and 58% of females (aged 13–34 years) using cannabis at least weekly. In a subsample of 60 cannabis users opportunistically recruited for in-depth interviews in 2005–2006 (37 male and 23 female, aged 13–42 years), 92% of males and 78% of females used cannabis daily; 88% reported cannabis dependence symptoms. These figures appear to be far higher than national rates, although national data for similar age groups are not available.4,5 Research has found that, nationally, 6% of males and 3% of females (aged ≥ 14 years) reported using cannabis in the past week; 18% of males and 13% of females smoked cannabis daily;4 and 21% of adults (aged ≥ 18 years) using cannabis were dependent.5 Beyond high rates of cannabis use in Arnhem Land communities, we also found local characteristics and perceptions that illustrate the drug’s distinctive context of use (Box). Quantities of cannabis used appear to be higher than in the general population; unemployment among users is higher; and violence related to diminished supply is common. One Indigenous community leader described attitudes to cannabis use: “... if there’s a bowl of it on the table, it is smoked until gone, morning to night”. Interestingly, some respondents reported that using cannabis prevents them from engaging in criminal activity (Box). While key community members may believe that cannabis is a tool for social control — “good for calming down people” — they are increasingly recognising the significant social and mental health problems it causes: People get chained by [cannabis], they don’t go hunting with family ... lots of fights when they can’t get any ... [Cannabis] becomes the boss. Continued concerns about adverse mental health consequences for Aboriginal people in Arnhem Land who use cannabis seem to be warranted. Cannabis appears to be firmly entwined in these isolated communities in a manner not seen nationally. High levels of concurrent drug use, particularly tobacco, raise additional health concerns. Resources are urgently needed for prevention programs and targeted interventions for chronic cannabis users and those with psychiatric comorbidity. If these patterns of use continue, the implications for compounding of pre-existing mental illness and the potential mental health burden are disturbing. Characteristics and perceptions of cannabis use in Arnhem Land Aboriginal communities (57 males and 49 females, aged 13–42 years*) in 2005–2006 compared with available national data from 1997† and 2004‡ National4,5 Arnhem Land* National4,5 Arnhem Land* Number of cones smoked Per cent unemployed current users / daily users 3.2 (average per day)‡ 7.4 (average per occasion) 25.6%† / nd 60% / Males, 41%; females, 94% Concurrent drug use Motivations for use Alcohol (86.2%); stimulants§ (8.2%–27.9%); none (10.8%); analgesics (6.6%); antidepressants (5.7%); tranquillisers/sleeping pills (4.4%); other (3.9%)‡ Tobacco (100%); alcohol, restricted access (40%); kava (15%); petrol (5%)¶ nd Socialisation (tempted, lonely, copying friends); mood altering (“calms me down”, “gets me going in the morning”, “makes my mind straight”); drug substitution (from alcohol or petrol); prevents criminal activity (stealing or other trouble) Drug substitution (when cannabis unavailable) Motivations for ceasing/moderating use Alcohol (60.4%); no substitution (34.2%); ecstasy/designer drugs (1.3%); painkillers/ analgesics (0.8%); tranquillisers/sleeping pills (0.5%); heroin (0.3%); antidepressants (0.2%); cocaine/crack (0.1%); other (1.1%)‡ No substitution (83%); kava (7%); alcohol (5%); petrol (5%)¶ nd Limited supply; starting a family (females); “sick of fighting when cannabis runs out”; “made me sick”; “mind not straight”; expenses and time spent looking for cannabis; employment (males) * Self-report interview data from an opportunistically recruited sample (using age and sex quotas) of respondents, including people who had never used cannabis as well as current and former cannabis users. † People aged ≥ 18 years.5 ‡ People aged ≥ 14 years.4 § Including ecstasy. ¶ There have been no reliable reports of stimulant, benzodiazepine or barbiturate use in these communities. nd = data not available.

K S Kylie Lee · Alan R Clough · Katherine M Conigrave