Volume 193 Issue 7

Results of two new trials will drive further research into the “andropause hypothesis” Two important research articles, published in a recent issue of the New England Journal of Medicine (NEJM), bracket the topic of prescribing testosterone for older men.1,2 Testosterone treatment for older men is based on considering male ageing as analogous to either menopause or pathologically based hypogonadism. The former is a false analogy — menopause has a unique natural history featuring complete failure of female reproductive hormones in mid-adult life, contrary to all other human hormonal systems which decline gradually and modestly with ageing. The latter is based on the superficial resemblance of non-specific symptoms in ageing men with those of most hormonal deficiencies or chronic diseases. This “andropause hypothesis” is not well substantiated, with the 2004 United States Institute of Medicine’s authoritative review3 concluding that available evidence did not justify a major study of testosterone replacement in older men comparable to the Women’s Health Initiative (WHI) study of oestrogen replacement in menopause. Nevertheless, the past two decades have seen an approximately 20-fold increase in testosterone prescribing despite no proven new indications. This is largely confined to the US, with minimal changes in Australia4 and other regional markets; however, that bandwagon could certainly be viewed as having left the station, fuelled by heavy direct-to-public drug advertising in the US. One of the NEJM articles reports the European Male Ageing Study (EMAS) — a large observational study of male ageing involving more than 3300 men aged 40 years and over from population-based sources in eight European cities, and the European counterpart to the seminal Massachusetts Male Ageing Study (MMAS).1 The EMAS evaluated the relationship of non-specific physical and mental symptoms to serum testosterone levels, an approach originating from clinic-based studies5,6 and used in the analysis of the population-based data of the MMAS,7 to which the EMAS adds a large sample size. Crucially, the statistical power of this study can neither overcome its logical flaws nor the inconsistency between its findings and conclusions. After noting statistically insignificant associations of low testosterone levels with clinically relevant physical or psychological features, the researchers focus on three sexual symptoms — erectile dysfunction, frequency of morning erections and sexual desire. Each symptom shows a weak association with serum testosterone levels, featuring a shallow breakpoint (at levels of 8–11 nmol/L) together with high false-positive and negative rates. Although the only consistent significant association is between all three sexual symptoms and a serum testosterone level threshold of 8 nmol/L, the researchers inexplicably propose a “definition” of “late-onset hypogonadism” requiring the presence of all three sexual symptoms plus a serum testosterone level of less than 11 nmol/L. The proposal is further undermined by the study’s findings that all the associations of the three sexual symptoms with low serum testosterone levels are nullified by adjustment for age, obesity and co-existing illnesses, indicating that they are attributable to confounding rather than any authentic correlation. The study’s authors, in effect, overinterpret cross-sectional data to imply causality. Such quasi-longitudinal interpretation is especially unsafe when blood test results are not steady for the population. This is the case for serum testosterone, with evidence of downward temporal trends in America and Europe,8 probably due to progressive population increases in obesity. The impact of implied causality should not be underestimated — despite its ritual caveats against testosterone prescribing, this article is likely to encourage more overuse of testosterone. The article resonates with past mistakes, notably the decades of excessive oestrogen prescribing, encouraged by overinterpreted observational studies and curbed only by the first placebo-controlled randomised trial of hormonal replacement therapy. Lest we sleepwalk down that same path, let us hope that it will not take decades before the “andropause hypothesis” undergoes rigorous testing. The second NEJM article reports the early termination of a randomised, placebo-controlled clinical trial for excess adverse cardiovascular effects associated with testosterone supplementation.2 The study evaluated the somatic benefits of 6 months of daily use of testosterone gel in frail men over 65 years with low serum testosterone levels. As expected, this population had a high prevalence of cardiovascular disease, which would explain their low serum testosterone levels as a non-specific effect of chronic illness. The progressive excess of adverse cardiovascular effects in testosterone-treated men (23 men receiving testosterone v 5 receiving placebo in the trial’s total of 209 men) was unexpected but persisted despite the use of various definitions of adverse cardiovascular events (the original definition was broad and included unexplained syncope and peripheral oedema). As the study design was conventional in regard to testosterone dosage, titration and monitoring, these findings show a low cardiovascular safety margin for testosterone supplementation in frail older men. They differ from the findings of previous comparable placebo-controlled studies of testosterone use in older men, which reported no excess cardiovascular events,9 even in 12-month studies of men with cardiac failure.10 Nevertheless, these adverse findings cannot be considered surprising given the earlier onset and greater severity of cardiovascular disease in men, together with the refutation of the long-dominant hypothesis that oestrogen provides women with a degree of cardiovascular protection.11 As this second article2 highlights, the effects of a treatment that excessively increases risk for the most common cause of death — cardiovascular disease — overwhelm even substantial improvements in less common or non-fatal disorders. Hence, the study’s finding of a benefit (improved limb muscular strength) was overshadowed by adverse cardiovascular profiles, as has happened with other treatments, such as the WHI’s reductions in bone fractures and colorectal cancer, and the highly effective, gastric-sparing cyclooxygenase-2 (COX-2) inhibiting analgesics and the oral antidiabetic glitazones. A corollary is that overinterpreting the regular association of low serum testosterone in men with cardiovascular disease as a risk factor that might be ameliorated — rather than as a consequence — led to a misplaced focus on prostate cancer as the major risk of testosterone treatment in older men. Taken together, these two studies can be construed as pressing the accelerator and the brakes at the same time on testosterone prescribing for older men . . . with probably the usual effect of spinning wheels. However, several reflections arise for Australian clinical practice. First, Australia’s national guidelines for testosterone use,12 developed in 2000 and adopted by the Pharmaceutical Benefits Scheme as the criteria for subsidised testosterone prescriptions, are vindicated. They aim to restrict testosterone prescribing for age-related androgen deficiency without hindering it for pathologically based androgen deficiency. European and US guidelines (produced in 2005 and 2006, respectively) were republished recently, essentially unchanged.13,14 The northern hemisphere guidelines, in contrast to Australia’s, blur the distinction between pathologically based and age-related androgen deficiency, loosen the diagnostic criteria and lack regulatory force. Thus they minimise the diagnostic hurdle, leading to the perverse outcome of potentially encouraging rather than deterring unproven overuse of testosterone. Second, the EMAS definition of “late-onset hypogonadism” is not suitable for implementation in practice. Findings of the MMAS confirm clinical experience and are reflected in all clinical guidelines:12-14 serial serum testosterone levels in older men are sufficiently variable7 to require the results of at least two blood samples taken weeks apart to establish a sustained low level of serum testosterone. Hence, a definition reliant on a sample of single testosterone measurements is likely to be highly error-prone, due to regression to the mean and other sources of variability. It also relies on testosterone measurements by mass spectrometry, a superior technology not yet available in Australian pathology laboratories despite recognised limitations of testosterone immunoassays.15 Similarly, the so-called “free” testosterone variable, also used in the study but not recommended, is calculated by an inaccurate formula unsuitable for individual diagnosis.16 Third, reflection on these two NEJM articles makes evident the need to reinforce bans on direct-to-public advertising of testosterone supplementation. This critical protection depends on industry self-regulation. Without this protection, an avalanche of misguided testosterone prescribing awaits us, analogous to the massive, unregulated marketing of non-proprietary impotence drugs advertised on billboards and in the media, apparently beyond regulatory reach. Finally, testosterone prescribing for older men is best restricted to clinical trials where ethical oversight ensures appropriate design and warnings to participants. Age alone may not prove a valid indication, but this should not limit essential further research within the framework of placebo-controlled clinical trials aiming to define the benefits and risks of testosterone use in patients with the comorbidities of ageing, such as obesity, diabetes, metabolic syndrome and other chronic conditions.

David J Handelsman MB BS, FRACP, PhD

Introducing an innovative, evidence-based resource for use in the emergency department In early 2008, the National Institute of Clinical Studies (NICS) Stroke Clinical Reference Group was formed to develop an acute stroke care resource for use in emergency departments (EDs) in Australian hospitals. The NICS reference group used a care bundle approach to develop a guideline implementation tool based on specific recommendations from the 2007 National Stroke Foundation (NSF) Clinical guidelines for acute stroke management relevant for ED care.1 Although these guidelines were already available, there are well known barriers to guideline implementation in the ED. These include increasing demand and acuity, and the broad diversity of clinical presentations. Clinical information provided for ED clinicians needs to be concise and relevant to the emergency care context. The nine-member NICS reference group represented a collaboration between stroke and ED specialists, prehospital providers and managers of state-based stroke networks, with additional guidance from the NSF. Over the following 12 months, a consultative process took place, with a combination of face-to-face and teleconference meetings and email exchanges. The reference group used a Delphi process to reach consensus. In December 2009, the NICS released two documents — the Emergency department stroke and TIA care bundle: information and implementation package and the accompanying Summary for clinicians. These are available on the National Health and Medical Research Council (NHMRC) website (http://www.nhmrc.gov.au/nics/programs/emergency/stroke_tia.htm). This editorial presents a précis of the care bundle. Care bundles have already been shown to improve guideline compliance and lead to improved patient outcomes in several settings, including the ED.2-6 A care bundle is made up of a small number of best-practice recommendation components, is not as comprehensive as a guideline, and aims to identify critical recommendations relating to areas in which there is a significant practice gap or to act as a trigger to other best practice.7 The NICS care bundle needed to bring together several components to help clinicians provide quality care to adult patients who present to the ED with suspected stroke or transient ischaemic attack (TIA) by reducing morbidity and mortality and optimising patient outcomes (Box 1). The criteria for a component’s inclusion in the care bundle were determined by the model developed by the Institute for Healthcare Improvement in the United States:7 each component must be based on sound evidence; the delivery of each component must need improvement; the delivery of each component must be achievable in terms of universally available resources; no component should be a major source of controversy; and the delivery of each component must be measurable. Two components — stroke unit care and thrombolysis — are not included in the care bundle. We acknowledge the importance of stroke unit care — and thrombolysis for patients who meet the criteria for its use — when appropriate resources are available. The NSF recommendations, along with similar international guidelines, state that thrombolysis should only be given under the authority of a specialist team with expert knowledge of stroke management and with pathways and protocols in place to guide the acute phase.1,8-10 Although there is level I evidence that thrombolysis and stroke unit care are effective early interventions for stroke,2 currently thrombolysis is only offered in 28% of acute hospitals that manage stroke patients, and stroke unit care is only available in a third of hospitals across Australia.8 The reference group considered all of these factors and decided, by consensus, not to include thrombolysis and stroke unit care in the care bundle, as the necessary resources to support these are not universally available. However, the reference group believes that an emphasis on the first component of the bundle — a rapid initial stroke screen — could lead to earlier referral to stroke specialists and rapid access to computed tomography or magnetic resonance imaging to confirm the diagnosis and develop a management plan that would consider thrombolysis if clinically appropriate.11 This illustrates how the components of the care bundle may trigger additional best-practice recommendations as a natural consequence and establish joint clinical decision making with other disciplines to improve patient care (Box 2). The NICS clinical reference group is planning to collaborate with key stakeholders in 2010 to evaluate the effectiveness of the care bundle, both as a format for providing specific guideline recommendations to a target audience and in terms of the impact on stroke care in the ED. An implementation plan and auditing tool have also been developed to assist in the uptake of the recommendations. The NICS care bundle is based on the 2007 NSF clinical guidelines,1 and its recommendations are consistent with the current draft of the 2010 NSF guidelines. It is intended that the care bundle will evolve to ensure that recommendations relevant to the ED remain current. 1 Components of the NICS care bundle Rapid initial stroke screen (grade C; level II)* ABCD2 assessment† for suspected TIA (grade B; level II) Urgent‡ CT or MRI (grade A; level I) Nil by mouth until bedside swallow screen (within 24 hours) for stroke (grade C; level I) Aspirin as soon as possible,§ if haemorrhage excluded (grade A; level I): 150–300 mg one-time loading unless contraindicated Physiological monitoring and treatment Neurological status (grade C; levels II and III-2): regular monitoring to establish baseline and identify change Blood glucose (grade B; level II): cautious treatment of markedly elevated blood glucose levels; early, intensive maintenance of euglycaemia is not recommended. Avoid hypoglycaemia Blood pressure (consensus¶): cautious lowering by no more than 10%–20% if extremely high (≥ 220/120 mmHg); monitor for neurological deterioration Hydration status (grade B; level II): maintain euvolemia NICS = National Institute of Clinical Studies. TIA = transient ischaemic attack. CT = computed tomography. MRI = magnetic resonance imaging. * Evidence-based grades and levels as per 2007 National Stroke Foundation clinical guidelines.1 † A seven-point score calculated from age, blood pressure, clinical features, duration of symptoms, and diabetes status. ‡ “Urgent” means as soon as possible, but certainly within 24 hours.1 § “As soon as possible” means within 48 hours.1 ¶ Recommended best practice based on clinical experience and expert opinion. 2 Application of the NICS care bundle* Case study 1: a 68-year-old man presents to a hospital emergency department (ED), having woken with marked weakness of his left arm. Enquiry establishes that he was fine when he went to bed 7 hours earlier. The patient’s blood pressure (BP) at triage is 186/99 mmHg. The triage nurse is concerned that the patient is having a stroke. Case study 2: a 74-year-old woman with a history of type 2 diabetes mellitus and hypertension presents to a metropolitan tertiary hospital ED. She is unable to speak and has no strength in her right arm or right leg. Her friend states that the symptoms started only 2 hours ago. The patient’s BP is 170/95 mmHg; her heart rate is 80 beats/min and the heart is in sinus rhythm; and her blood glucose level is 9 mmol/L. The following care is provided for these patients, consistent with the use of the care bundle: as part of the patient’s assessment, and based on clinical findings, conduct a rapid initial stroke screen using a validated stroke screening tool to determine whether the patient is likely to have had a stroke. If a stroke is suspected, promptly refer the patient for expert stroke management — this may include referral to a stroke unit, or thrombolytic treatment (which is likely for the patient in case study 2); order an urgent computed tomography (CT) scan of the brain; ensure no oral intake until the patient undergoes a swallow screen for dysphagia; maintain hydration via intravenous or nasogastric fluids; administer aspirin (150 mg) within 48 hours if the brain CT scan excludes haemorrhage (if the patient in case study 2 proceeds to thrombolysis, delay aspirin treatment until 24 hours after thrombolysis); monitor the patient’s neurological status, blood glucose level, BP and hydration status to prevent further deterioration. NICS = National Institute of Clinical Studies. * The NICS care bundle was written for the care of stroke patients while in the ED. If the patient is transferred out of the ED early in his or her care, it is anticipated that the remaining components of the bundle will still be provided in the new setting.

Jayantha I Weeraratne MB BS, FACEM · Annette J Lenstra BSc, GradDip(Gov) · Andrew W Lee MB BS, MPH, FRACP · Kelvin M Hill BAppSci(Physio), GradDip(BusComm) · Susan D Huckson BAppSci, RN, ICU(Cert) · Jodie L Clydesdale BNurs, GradDip(ClinNurs)

Objective: To evaluate the impact of psychiatric inpatient bed closures, accompanied by a training program aimed at enhancing team effectiveness and incorporating data-driven practices, in a mental health service.Design and setting: Retrospective comparison of the changes in services within three consecutive financial years: baseline period — before bed reduction (2006–07); observation period — after bed reduction (2007–08); and intervention period — second year after bed reduction (2008–09). The study was conducted at Cramond Clinic, Queen Elizabeth Hospital, Adelaide.Main outcome measures: Length of stay, 28-day readmission rates, discharges, bed occupancy rates, emergency department (ED) presentations, ED waiting time, seclusions, locality of treatment, and follow-up in the community within 7 days.Results: Reduced bed numbers were associated with reduced length of stay, fewer referrals from the community and subsequently shorter waiting times in the ED, without significant change in readmission rates. A higher proportion of patients was treated in the local catchment area, with improved community follow-up and a significant reduction in inpatient seclusions.Conclusion: Our findings should reassure clinicians concerned about psychiatric bed numbers that service redesign with planned bed reductions will not necessarily affect clinical care, provided data literacy and team training programs are in place to ensure smooth transition of patients across ED, inpatient and community services.

Tarun J Bastiampillai MB BS, BMedSc, FRANZCP · Niranjan P Bidargaddi BE(CompSc), PhD · Rohan S Dhillon MB BS, FRANZCP, MClinSc · Geoffrey D Schrader FRANZCP, PhD · Jörg E Strobel MD, FRANZCP · Philip J Galley RMN, DipCPC, BA(Hons)

Objective: To determine the proportion of older Australian men who meet the Pharmaceutical Benefits Scheme (PBS) criteria for osteoporosis treatment and are receiving effective treatment.Design and setting: A population-based, cross-sectional analysis of the baseline phase of the Concord Health and Ageing in Men Project (CHAMP), a large epidemiological study focusing on the health of older men. Data were collected through questionnaires and clinical assessments. Bone mineral density (BMD) of the hip and spine was measured by dual x-ray absorptiometry (DXA). Vertebral deformities were identified from DXA lateral vertebral fracture assessment images. The study was conducted at Concord Hospital, Sydney, between January 2005 and May 2007.Participants: 1705 community-dwelling men aged 70 years or over from a defined geographical region around Concord Hospital.Main outcome measures: Prevalence of vertebral deformities; previous minimal trauma fractures; BMD T-scores ≤ – 3; falls in the previous 12 months; use of bisphosphonates and calcium and vitamin D supplements.Results: Of the 1705 men seen at baseline, 1626 completed all DXA scans and 401 (25%) met one or more of the PBS criteria for osteoporosis treatment. Ninety per cent of the men who met the PBS criteria were unaware they had osteoporosis. Of the men eligible for PBS-subsidised treatment, 39 (10%) reported use of a bisphosphonate, 56 (14%) had taken calcium supplements, and 28 (7%) had taken vitamin D supplements. Only three men had taken calcium, vitamin D and bisphosphonates in combination.Conclusions: Despite a high prevalence of osteoporosis in elderly Australian men, awareness, diagnosis and treatment of the condition remain very low.

Kerrin Bleicher BSc, PostGradDipPhysio · Vasi Naganathan MB BS, FRACP, PhD · Robert G Cumming MB BS, MPH, PhD · Markus J Seibel MD, FRACP, PhD · Philip N Sambrook MD, LLB, FRACP · Fiona M Blyth MPH, FAFPHM, PhD · David G Le Couteur FRACP, GradCertEd, PhD · David J Handelsman MB BS, FRACP, PhD · Louise M Waite MB BS, FRACP, PhD · Helen M Creasey MB BS, FRACP

Objective: To compare trends in invasive pneumococcal disease (IPD) in non-Indigenous people in north Queensland before and after the introduction of funded pneumococcal vaccines, and to examine the proportion of cases that occurred after vaccine roll-out that could be vaccine-preventable.Design, setting and participants: In 2005, a 7-valent pneumococcal conjugate vaccine (7vPCV) for non-Indigenous children and a 23-valent pneumococcal polysaccharide vaccine (23vPPV) for non-Indigenous adults aged ≥ 65 years were made freely available. Trends in IPD in the non-Indigenous estimated resident population in north Queensland (about 581 850 in 2006) were compared between the 4 years before (2001–2004) and after (2006–2009) the vaccines were rolled out.Main outcome measures: Incidences and serotypes of IPD in non-Indigenous people.Results: After the introduction of the vaccines, there were significant declines for all ages in the average annual incidence of IPD (− 34%; P < 0.05) and 7vPCV serotype IPD (− 77%; P < 0.05). In children aged < 5 years, there was a 91% decline in the incidence of 7vPCV serotype IPD (P < 0.05); in adults aged 15–64 years and ≥ 65 years there were 62% and 77% declines, respectively, in 7vPCV and 23vPPV common-serotype IPD (P < 0.05). There was a 188% increase in 23vPPV-only serotype IPD in adults aged 15–64 years (P < 0.05), whereas there was no significant change in adults aged ≥ 65 years. Serotype 19A was the most frequently identified serotype in 2006–2009, causing 19% of all IPD in those 4 years.Conclusions: There is circumstantial evidence that 7vPCV has had a powerful indirect effect in preventing IPD in adults in north Queensland; 23vPPV may have had a direct effect in adults aged ≥ 65 years. It is likely that with combined direct and indirect effects, newer conjugate vaccines could prevent more IPD than could be prevented with the two current vaccines.

Jeffrey N Hanna MPH, FAFPHM · Jan L Humphreys RN · Denise M Murphy DipMedTech · Helen V Smith GradDipPH, BApplSci, MASM

E-health is the health care buzzword of the moment, with a person-controlled electronic health record funded in the 2010 federal Budget and legislation to introduce health identifiers recently passed by Parliament. E-health can ease the patient journey, improve quality of care and reduce costs. Australia’s health care system lags behind all other sectors of our economy in the use of computerised systems. While general practice and community pharmacy are highly computerised, the hospital sector is not. Adopting e-health is likely to result in higher quality practice, but general practice and hospitals need a mechanism for securely sharing patient data. Uncoordinated implementation of differing, incompatible systems within and between hospitals compounds a dire lack of national coordination of effort. Multiple funding streams and jurisdictions and the lack of an implementation strategy have slowed e-health development. Government programs underestimate the costs of change management and the need for training and technology. Confusion reigns about responsibilities, but governments must ensure connectivity between health care providers and recognise that the benefits will accrue into the future. The National E-Health Transition Authority has developed national open-access standards, and its foundation projects and the National Broadband Network are now coming into place. To ensure the clinical relevance, utility, safety and acceptability of e-health systems, health professionals urgently need technical capacity and expert guidance.

Christopher Pearce PhD, MFM, FACHI · Mukesh C Haikerwal MB ChB, FRACGP, DipIMCRCS

Investment in information and communication technology (ICT) in the health sector can bring important benefits. To date, the focus has been on automating clinical work practices such as ordering tests and prescriptions, which significantly improves efficiency and safety. Uptake of ICT has been slow and the results less favourable than anticipated for various reasons, including poor integration of systems into complex clinical work processes, limited training, and the intermittent nature of ICT funding. As a result, many health care organisations have been operating hybrid paper and computer systems that introduce new patient risks, staff frustration, and outcomes below expectation. The focus must shift from automation of clinical work to innovation; from evolutionary application of ICT to revolutionary uses. Health professionals must embrace ICT as a “disruptive technology” that will produce significant changes in their roles and responsibilities and lead to real health reform with new, innovative models of health care delivery. As other industries have shown, substitution and role changes are areas in which ICT can lead to the greatest gains.

Johanna I Westbrook PhD, FACMI, FACHI · Jeffrey Braithwaite MBA, PhD, FCHSM

Objective: Design, setting and participants: Vaccination data for Western Australians aged 10 years and older were obtained from two sources: the WA Pandemic Influenza Vaccination Database (PIVD; which collected reports of pandemic influenza vaccinations from vaccination providers statewide) for the period 30 September 2009 to 31 January 2010, and the WA Health and Wellbeing Surveillance System (HWSS; a continuous population-based telephone survey) for the period 1 December 2009 to 31 January 2010. Data from the PIVD was used to impute vaccination coverage estimates for at-risk subpopulations not assessed in the HWSS interviews.Main outcome measures: Vaccination coverage of Western Australians aged 10 years and older and of subgroups targeted by the national pandemic (H1N1) 2009 influenza vaccination campaign.Results: A total of 171 789 pandemic influenza vaccinations were reported to the PIVD by 31 January 2010 and 88% of these were administered by 1 December 2009. Based on HWSS data, vaccination coverage of persons aged 10 years and older was 14.5% (95% CI, 12.6%–16.6%) and of persons aged 18 years and older was 15.3% (95% CI, 13.3%–17.6%). Based on PIVD data, coverage in adults ranged from 10.3% in pregnant women to 52.8% in health care workers.Conclusions: Our estimate of pandemic influenza vaccination coverage in the adult population of WA is comparable to the national estimate of 19%, but it did not reach levels considered sufficient to interrupt community transmission. Future influenza vaccination programs should target groups at increased risk of severe influenza, such as pregnant women.

Donna B Mak MB BS, MPH, FAFPHM · Alison M Daly BA(Hons), BA(Ed) · Paul K Armstrong MBBS, MAE, FRACP · Paul V Effler MD, MPH, FAFPHM

Objective: Design, setting and participants: Cross-sectional study of consecutive patients attending the Joondalup Health Campus public antenatal clinics in WA in January 2010.Intervention: Audit of uptake of the H1N1-specific vaccine.Main outcome measures: Rate of H1N1-specific vaccination, and reasons for not being the vaccinated.Results: 479 of 541 women who attended the clinics (88.5%) were included in the audit. Three women had been infected with pandemic influenza in the preceding influenza season, leaving 476 women who were eligible for vaccination in pregnancy. Of these 476 women, only 33 (6.9%) had been vaccinated. Of the remaining 443 women who were eligible to receive the vaccine but had not been vaccinated, 63.9% had not been offered vaccination despite multiple visits to their general practitioners during pregnancy, 19.6% had been advised by their GPs against vaccination in pregnancy, and 61.6% stated that they would decline vaccination if offered because of safety concerns.Conclusions: Uptake of H1N1-specific influenza vaccine in pregnant women was poor. Reasons for this relate both to vaccination not being offered to or actively sought by the women, as well as concerns — of both the women and their GPs — about vaccine safety in pregnancy. Uptake in this setting may improve if vaccination is offered through public antenatal clinics with concurrent safety education for obstetricians and vaccination providers.

Scott W White MB BS · Rodney W Petersen MB BS, MBA, FRANZCOG · Julie A Quinlivan MB BS, PhD, FRANZCOG

We identify key lessons learned from the international experience of pay-for-performance and use them to formulate questions for Australia to consider before such a scheme is introduced. Discussion of lessons learned is based on a narrative review of the literature. We examined international evidence on factors to consider when designing pay-for-performance schemes, and the impact of these schemes on primary care practitioner behaviour and on primary care funding. Pay-for-performance schemes evolve over time, and usually involve several complex interventions including accreditation, education, quality improvement programs, investment in information technology and data collection systems, professional support and regional structures. These are all necessary conditions for linking financial incentives to quality of care. There is a strong argument for changing the existing service incentive payments program and investing the resources into revised outcome payments that provide rewards for annual improvements in numbers of patients receiving completed cycles of care. If pay-for-performance is to be introduced in Australia, several key lessons should be learned from the experiences of other countries. Pay-for-performance should be used as part of a wider strategy for quality improvement; it should not be seen as a panacea. Pay-for-performance should be used to drive quality improvement, not simply to reward those who are already providing high-quality care.

Stephen M Campbell BA(Hons), MA(Econ), PhD · Anthony Scott BA(Hons), MSc, PhD · Rhian M Parker BSc(Econ)(Hons), MSc, PhD · Lucio Naccarella BSc(Hons), GDipMHS, PhD · John S Furler MB BS, FRACGP, PhD · Doris Young MB BS, FRACGP, MD · Peter M Sivey BSc(Hons), MSc, PhD

We report a series of cases of thyroid dysfunction in adults associated with ingestion of a brand of soy milk manufactured with kombu (seaweed), and a case of hypothyroidism in a neonate whose mother had been drinking this milk. We also report two cases of neonatal hypothyroidism linked to maternal ingestion of seaweed made into soup. These products were found to contain high levels of iodine. Despite increasing awareness of iodine deficiency, the potential for iodine toxicity, particularly from sources such as seaweed, is less well recognised. Clinical recordsCases of thyroid dysfunction associated with ingestion of soy milkIn November 2008, a 36-year-old woman (Patient 1, Box 1) presented with a mildly elevated serum thyroid-stimulating hormone (TSH) level detected during screening for in vitro fertilisation. As she tested negative for thyroid antibodies, her urinary iodine level was measured to exclude iodine deficiency; this level was markedly elevated at 4445 μg/L (reference range [RR], < 200 μg/L). The source of the excess iodine was unclear until the patient did an internet search and identified that the soy milk she had been drinking (Bonsoy) contained kombu1,2 — a type of seaweed. The patient ceased drinking the soy milk, which resulted in rapid normalisation of her TSH level. Three months later, a 38-year-old man (Patient 2, Box 1) presented with florid thyrotoxicosis. Minimal uptake of technetium on a thyroid scan and absence of TSH receptor antibodies essentially excluded Graves disease. The scan result, in combination with his elevated urinary iodine level (1278 μg/L), indicated that iodine toxicity was the most likely cause of the thyrotoxicosis. He drank brands of soy milk other than Bonsoy, but also drank Bonsoy in takeaway coffee. After he ceased drinking all soy milk, his symptoms rapidly abated and his serum TSH level normalised 3 months later. No further cases of suspected iodine toxicity were seen until approximately 1 year later, when six additional patients presented to one of us (B A C) over a 6-week period (Patients 3–8, Box 1). These patients presented with thyroid conditions ranging from subclinical hyperthyroidism to florid thyrotoxicosis. Patient 3 had already been diagnosed with thyrotoxicosis due to underlying iodine toxicity (urinary iodine level, 11 427μg/L); however, the source of excess iodine was not identified until she sought a second opinion. One month after she ceased consuming Bonsoy milk (which she had been consuming for the previous 8 years), her serum TSH level normalised. An aliquot of Bonsoy milk was analysed for iodine content using a plasma mass spectrometer (Department of Biochemistry, Royal Prince Alfred Hospital, Sydney, NSW), which showed an iodine concentration of 25 000 μg/L. In comparison, the levels of iodine in other soy milks that were analysed ranged from 15 μg/L to 281 μg/L (Box 2). Two weeks later, the same laboratory received a second aliquot of Bonsoy milk for analysis, due to a case of neonatal hypothyroidism. The newborn screening program had identified a baby with an elevated TSH level (28 mIU/L; RR, < 20 mIU/L; heel-prick blood sample). Additional testing 19 days after birth showed further elevation of the baby’s serum TSH level (163 mIU/L), as well as a low level of serum free thyroxine (3.7 pmol/L; RR, 10–25 pmol/L). Exposure to exogenous iodine from a maternal source was suspected because of the marked rise in serum TSH level. Urinary iodine levels were subsequently found to be elevated in both the mother (5415 μg/L) and the baby (9797 μg/L). During the last trimester of pregnancy, the mother had been drinking about 500 mL of Bonsoy milk daily. She had been breastfeeding since delivery. The iodine concentration of the second aliquot of this soy milk (27 580 μg/L) was similar to that of the previously analysed sample. The baby was initially treated with thyroxine but, after the mother ceased ingesting the soy milk, the baby’s thyroid function normalised. Independent analysis of the soy milk (Division of Analytical Laboratories, NSW Health, Sydney, NSW) again revealed an extremely high iodine concentration (31 000 μg/L). Cases of neonatal hypothyroidism associated with maternal ingestion of seaweed soupTwo cases of neonatal hypothyroidism related to maternal ingestion of seaweed have been reported recently by two of us (P J E and M M J).3 The first case involved a Korean mother who, during pregnancy and the puerperium, consumed soup made with overseas-bought dried seaweeds. Her baby, born at 36 weeks’ gestation, had a normal TSH level at the time of newborn screening (heel-prick blood sample). However, the baby subsequently developed jaundice and, at 3 weeks of age, a repeat TSH test showed elevation of the baby’s serum TSH level (39 mIU/L; RR, 0.4–5.0 mIU/L) as well as a low level of serum free thyroxine (9.7 pmol/L; RR, 13–30 pmol/L) and an elevated urinary iodine level (690 μg/L). The baby was initially treated with thyroxine but, after the mother ceased ingesting seaweed soup, the baby’s thyroid function normalised. Dried samples of two different seaweed compounds, analysed by a commercial pathology company (Sullivan Nicolaides Pathology, Brisbane, QLD), showed iodine concentrations of 291 μg/g and 424 μg/g. The second case involved an infant born at 27 weeks’ gestation who had a normal TSH level at the time of newborn screening, and an elevated serum TSH level (24 mIU/L; RR, 0.06–7.14 mIU/L) when a routine repeat TSH test was carried out at 1 month of age. This infant’s mother had also been ingesting seaweed soup — made with Heng Fai seaweed, imported from China, to increase her breast milk supply. The baby’s urinary iodine level at the time of maternal seaweed ingestion was elevated (454 μg/L). The iodine concentration in the mother’s breast milk at the time of seaweed ingestion was elevated at 878 μg/L; 4 weeks after she ceased consuming the seaweed, the concentration dropped to 188 μg/L. NSW Health was notified and testing of the Heng Fai seaweed by the NSW Food Authority revealed high levels of iodine (4450 μg/g), which resulted in voluntary withdrawal of Heng Fai seaweed by the importers in March 2010.4 DiscussionIodine toxicity causes a spectrum of thyroid disorders, ranging from hyperthyroidism to hypothyroidism.5,6 Reasons for the variable effects are unclear, but may relate to age, pre-existing autoimmune thyroid disease, and amount and duration of iodine ingestion.5,6 The adults described here did not appear to have underlying nodular goiters or Hashimoto disease (Box 1). In iodine toxicity, thyroid technetium uptake scans usually show absent or low technetium uptake in thyrotoxicosis and increased technetium uptake in neonatal hypothyroidism. Graves disease and autonomous nodular thyroid disease are more common causes of thyrotoxicosis but can be excluded primarily by scan results, lack of a goitre and absence of TSH receptor antibodies. However, as urinary iodine levels are not routinely measured in clinical practice, iodine toxicity may be underdiagnosed. This series of cases of thyroid dysfunction led to a national recall of Bonsoy milk on 24 December 2009,7 and the distributer agreed to voluntary withdrawal of the product from sale in Australia. This brand of soy milk was fermented in seaweed, which is thought to improve the flavour, and is promoted as having wide-ranging health benefits.1 The NSW Health alert for Bonsoy milk stated that in a child, ingestion of only 5 mL, and in an adult, only 30 mL, would exceed the safe upper limit of iodine intake.8,9 It is unclear whether changes to the manufacturing process of the Bonsoy product may have increased its iodine content. However, after removal of kombu from the manufacturing process, the iodine content was reduced markedly (15 μg/L) and the product returned to the Australian market in April 2010. The World Health Organization was notified of the iodine toxicity of the Bonsoy milk, which was withdrawn from sale in a number of other countries.10 Between January and June 2010, 48 retrospective Australian cases of thyroid dysfunction associated with this brand of soy milk were also notified to local public health authorities (Katrina Knope, Coordinating Epidemiologist, OzFoodNet, Office of Health Protection, Department of Health and Ageing, June 2010, personal communication). A cluster of cases of thyrotoxicosis, linked to iodine toxicity from an unidentified soy milk, was also reported in New Zealand in 2005.11 The common practice by women from Japan and Korea of ingesting seaweed made into soup, sometimes in large quantities, to promote wellbeing in the mother and stimulate breast milk supply, does not appear to be widely known in the medical community. However, due to iodine transmission through breast milk, transient or even persistent hypothyroidism has been reported in neonates born to mothers who undertake this practice.12,13 If left undiagnosed and untreated, neonatal hypothyroidism can have devastating clinical consequences, including impaired intellectual development. Although newborn screening tests will help to identify hypothyroidism during the first week of life, there is no subsequent routine screening of thyroid function in term babies whose TSH level may not increase until after 1 week of age, as seen in one of the neonatal cases described here and in a Korean study of preterm infants.12 Our findings demonstrate the importance of: considering iodine toxicity in patients who present with thyrotoxicosis in the absence of TSH receptor antibodies and low or absent uptake on a thyroid technetium uptake scan; measuring urinary iodine level in cases where thyrotoxicosis is not explained by conditions such as autoimmune or nodular thyroid disease; and actively seeking a history of maternal seaweed consumption during pregnancy and lactation in cases of neonatal hypothyroidism. Finally, although iodine deficiency is a documented and serious concern in Australia,14,15 these cases highlight the risks of excess iodine intake from dietary sources. The food industry is not strictly regulated (eg, imported products are not usually tested to confirm their contents), and contamination of food and drink is only detected when unusual or severe clinical events ensue. There is a strong public health argument for monitoring iodine levels in imported foods and commercially available seaweed preparations. 1 Characteristics of a cluster of eight adult patients in whom thyroid dysfunction was attributed to consumption of Bonsoy, a brand of soy milk manufactured with seaweed, November 2008 to December 2009* Sex; age (years) SerumTSH level (mIU/L) Serum fT4 level (pmol/L) Serum fT3 level (pmol/L) Serum TRAb test result Serum TPO/Tg Ab test result Technetium uptake on thyroid scan Urinary iodine level (μμg/L) Thyroid ultrasound result RR 0.4–3.5 9–19 2.5–5.7 0.5%–3.5% < 200 Patient 1 F; 36 4.63 9.7 Not done Not done Negative Not done 4 445 Not done Patient 2 M; 38 < 0.02 59.4 16 Negative Negative Negligible 1 278 Normal size, single nodule (3 mm diameter), normal vascularity Patient 3 F; 46 < 0.005 50 39 Negative Negative < 0.5% 11 427 Mild enlargement, reduced vascularity Patient 4 F; 36 < 0.04 30 12 Negative Negative 0.5% 777 Normal Patient 5† F; 37 < 0.0005; 12.4 29; < 5 5.6; 3.4 Negative Negative Not done 6 208 Normal Patient 6‡ F; 29 0.04 16 4.9 Negative Negative 0.5% 48 Tiny nodules (< 3 mm diameter) Patient 7 F; 33 0.08 18 6.6 Negative Negative 1.3% 5 022 Normal Patient 8 M; 47 0.07 17 4.9 Negative Negative 0.1% 320 Single nodule (5 mm diameter) TSH = thyroid-stimulating hormone. fT4 = free thyroxine. fT3 = free triiodothyronine. TRAb = TSH receptor antibody. TPO/Tg Ab = thyroid peroxidase and thyroglobulin antibodies. RR = reference range. F = female. M = male. * Reported daily intake of Bonsoy milk ranged from < 100 mL/day to 1000 mL/day. † Patient 5 had blood tests for thyrotoxicosis performed at 5.5 months postpartum, and repeated at 7 months postpartum (when she had developed hypothyroidism). ‡ Patient 6 ceased consumption of the Bonsoy milk about 2–3 months before testing. 2 Iodine concentration in various milks, assayed in December 2009* Brand Iodine (μμg/L) Soy milks Bonsoy 25 000, 27 580 Sanitarium So Nice 27 Vitasoy 19 Coles Soy Drink 19 So Natural Original 15 Other milks Woolworths Lite (low-fat cows milk) 281 Pura Milk (full-fat cows milk) 215 So Good Rice Milk 29 * All testing was carried out at the Department of Biochemistry, Royal Prince Alfred Hospital, Sydney, NSW.

Bronwyn A Crawford PhD, MB BS, FRACP · Christopher T Cowell MB BS, FRACP · Phillip J Emder BSc(Med), MB BS, FRACP · Diana L Learoyd PhD, MB BS, FRACP · Elizabeth L Chua PhD, MB BS, FRACP · John Sinn MB BS, MMed(ClinEpi), FRACP · Michelle M Jack PhD, MB BS, FRACP

IgG4-related systemic disease (IRSD) is a recently described entity with protean manifestations. We describe a patient who developed inflammation and fibrosis in multiple organs over 20 years, sequentially involving his pancreas, bile ducts, gallbladder, submandibular and lacrimal glands, and kidneys. He had an elevated serum IgG4 level. Retrospective analysis of biopsies showed strongly positive tissue immunostaining for IgG4, confirming the diagnosis of IRSD. This case illustrates the natural history of partially treated IRSD and its varied clinical presentations. Early diagnosis and treatment is important, as the condition is highly steroid-responsive. Clinical recordA 55-year-old man who abstained from drinking alcohol and was not taking any medications initially presented to a hepatobiliary surgeon in 1989 for recurrent abdominal pain and jaundice. On examination, he was icteric, with mild epigastric tenderness, but was afebrile and had no other signs of chronic liver disease. His serum amylase level was mildly elevated, and results of liver function tests (LFTs) showed abnormal levels of albumin (29 g/L; reference range [RR], 36–48 g/L), alkaline phosphatase (284 U/L; RR, 32–91 U/L), γ-glutamyltransferase (98 U/L; RR, < 38 U/L), alanine transaminase (50 U/L; RR, < 34 U/L) and bilirubin (102 μmol/L; RR, < 18 μmol/L). An abdominal computed tomography (CT) scan showed a 6 cm mass in the head of the pancreas consistent with a carcinoma. At laparotomy, his pancreas was noted to be hard and thickened. An intraoperative cholangiogram revealed multiple bile duct strictures. A cholecystectomy and side-to-side choledochojejunostomy were performed. Biopsy of the pancreatic mass showed mature fibrocollagenous connective tissue with scattered plasma cells, but no evidence of malignancy. The liver biopsy showed features consistent with large bile duct obstruction. Histological examination of the cystic duct and gallbladder showed diffuse lymphoplasmacytic infiltration, with intervening areas of dense fibrosis, causing marked wall thickening (Box 1, A). No gallstones were found. Rectal biopsies performed later were normal, excluding ulcerative colitis. A diagnosis of chronic idiopathic pancreatitis and primary sclerosing cholangitis (PSC) was made. His symptoms waxed and waned with no specific treatment. In 1993, the patient presented to a rheumatologist with dry mouth, without dry eyes, and was noted to have bilateral submandibular gland swelling and exophthalmos from bilateral lacrimal gland enlargement. He was otherwise well, with no history of rash, myalgia or arthritis. His erythrocyte sedimentation rate (ESR) was 29 mm/h (RR, 7–18 mm/h). No antinuclear antibodies or antineutrophil cytoplasmic antibodies were detected. A percutaneous Trucut biopsy of the submandibular gland showed a dense lymphoplasmacytic infiltrate and prominent fibrosis. Lymphoepithelial lesions were present without morphological features of lymphoma. A diagnosis of seronegative Sjögren’s syndrome was made, and the patient was given a trial of steroid treatment, starting with 30 mg of prednisolone daily. His exophthalmos and submandibular gland swelling resolved over 3 months, and no abnormality was seen on a repeat orbital CT scan. Incidentally, his abdominal pain resolved and his LFT results returned to normal. The prednisolone dose was reduced over 12 months to 5 mg/day. In 2002, with a serum creatinine level of 180 μmol/L (RR, 64–104 μmol/L) and a urea level of 13.5 mmol/L (RR, 2.5–8.3 mmol/L), the patient was referred to a nephrologist. Antinuclear antibodies were detected at a high titre of 1 : 640, and his ESR was 45 mm/h. The patient had eosinophilia (0.9 × 109 cells/L; RR, 0–0.5 × 109 cells/L), with an otherwise normal full blood count. His serum IgE level was normal. His spot urine protein–creatinine ratio was 89 mg/mmol (RR, 15–35 mg/mmol), consistent with proteinuria. A renal biopsy showed interstitial oedema; marked fibrosis with a dense infiltrate of lymphocytes, plasma cells and eosinophils; and moderate tubular atrophy, indicating severe active chronic tubulointerstitial nephritis. The patient’s prednisolone dose was increased to 60 mg/day, resulting in a dramatic improvement in renal function and normalisation of the spot urine protein–creatinine ratio. His steroid dose was tapered over the next few years. In 2009, the patient was referred to our gastroenterology clinic. His history of a benign pancreatic mass with biliary strictures was typical for autoimmune pancreatitis, and a diagnosis of IgG4-related systemic disease (IRSD) was considered. His IgG level was elevated (23.3 g/L; RR, 5.0–16.0 g/L), but IgM and IgA levels were in the normal range. In particular, the IgG4 subclass was markedly elevated, at 12.2 g/L (RR, 0.04–0.86 g/L). Levels of all other IgG subclasses fell within the normal range. The patient’s earlier biopsy samples were retrieved and immunostained for IgG4 using a recently commercially available monoclonal antibody (Zymed Laboratories Inc, San Francisco, Calif, USA). Strongly positive immunostaining for IgG4 was seen in the gallbladder, cystic duct (Box 1, B), submandibular gland (Box 2, A), lacrimal gland and kidney (Box 2, B) biopsies compared with control staining for total IgG. The proportions of IgG4-positive plasma cells (relative to total IgG-positive plasma cells) in the cystic duct, submandibular gland and kidney were 89%, 45% and 66%, respectively. These high proportions of IgG4-positive plasma cells (> 40%) confirmed the diagnosis of IRSD.1 DiscussionRecurrent abdominal pain and elevated serum amylase levels had first developed in our patient 20 years previously. He was given the diagnosis of chronic idiopathic pancreatitis, as the concept of autoimmune pancreatitis was not reported until 5 years later.2 Autoimmune pancreatitis may present with recurrent abdominal pain and jaundice, and a pancreatic mass indicative of carcinoma is commonly seen on imaging. The diagnosis of autoimmune pancreatitis can be suggested by an elevated serum IgG4 level3 and confirmed by biopsy showing a dense lymphoplasmacytic infiltrate with numerous IgG4-positive plasma cells.1,4 Patients usually respond well to treatment with steroids. Our patient’s cholestatic LFT results and biliary strictures were ascribed to PSC. However, gallbladder and cystic duct histology revealed a lymphoplasmacytic infiltrate rich in IgG4-positive cells with associated obliterative phlebitis, typical of lymphoplasmacytic sclerosing cholangitis.5 An elevated serum IgG4 level is seen in 75% of patients with this condition, and steroid therapy usually normalises LFT results and biliary strictures, a feature not seen in PSC.6 Our patient also had chronic sclerosing cholecystitis. Although our patient’s chronic sclerosing sialadenitis and dacryoadenitis were attributed to Sjögren’s syndrome, his exocrine gland histology was more consistent with Mikulicz’s disease, which differs from Sjögren’s syndrome both clinically and histopathologically. Mikulicz’s disease is associated with prominent infiltration of IgG4-positive plasma cells into lacrimal and salivary glands, as well as a favourable response to corticosteroid treatment.7 Also, lacrimal gland tear production is preserved in Mikulicz’s disease but not in Sjögren’s syndrome,8 as was noted in our patient. Tubulointerstitial nephritis is the most common form of IgG4-related renal disease. Characteristically, the renal tubulointerstitium is infiltrated by large numbers of IgG4-positive plasma cells and eosinophils, resulting in cortical fibrosis and tubular atrophy.9 This condition is frequently associated with eosinophilia and proteinuria, and responds well to high-dose corticosteroid treatment, leading to improved renal function and reduced proteinuria. IgG4-associated membranoproliferative glomerulonephritis has also been reported. IRSD can manifest in varied clinical presentations (Box 3), unified by characteristic histological findings of diffuse IgG4-positive plasma cell infiltration of multiple organs, resulting in obliterative phlebitis and tissue fibrosis.10 Our case illustrates the potentially chronic natural history of partially treated IRSD leading to presentations to different medical specialists over 20 years, emphasising the importance of diagnosing IRSD in order to institute treatment with steroids. Competing interestsNone identified. 1 Biopsy samples of the patient’s cystic duct A: Cystic duct biopsy sample showing thickening of the wall, with a diffuse dense lymphoplasmacytic infiltrate accompanied by fibrosis (haematoxylin–eosin stain; original magnification, × 20). B: Cystic duct biopsy sample showing numerous IgG4-immunopositive plasma cells (monoclonal antibody stain; original magnification, × 400). 2 Biopsy samples of the patient’s submandibular gland and kidney Biopsy samples showing numerous IgG4-immunopositive plasma cells (monoclonal antibody stain). A: Submandibular gland (original magnification, × 200). B: Kidney (original magnification, × 400). 3 Manifestations of IgG4-related systemic disease Autoimmune pancreatitis* Lymphoplasmacytic sclerosing cholangitis* Lymphoplasmacytic sclerosing cholecystitis* Chronic sclerosing sialadenitis* Chronic sclerosing dacryoadenitis* Chronic tubulointerstitial nephritis* Membranoproliferative glomerulonephritis Autoimmune hepatitis Riedel’s thyroiditis Interstitial pneumonia Pseudotumours of the lung Lymphoplasmacytic aortitis Hypophysitis resulting in hypopituitarism or diabetes insipidus * Feature was present in our patient.

Eu Jin Lim MB BS · Prithi S Bhathal MB BS, FRCPA · Peter P Tagkalidis MB BS, PhD, FRACP · Antony G Speer MB BS, FRACP

Clinical records Patient 1 A 59-year-old man infected with HIV presented to hospital with sudden onset of tinnitus, vertigo and hearing loss in his right ear. An audiogram showed moderate bilateral sensorineural hearing loss, which was worse on the right. A magnetic resonance imaging scan of his brain showed no abnormalities. He was diagnosed with Meniere’s disease and managed symptomatically. Symptoms worsened over the following months, and he was reviewed in the neurology and ear, nose and throat (ENT) clinics of another tertiary hospital. Both clinics agreed with the diagnosis of Meniere’s disease. Serological tests for syphilis were performed 12 months after symptom onset. The rapid plasma reagin (RPR) and Treponema pallidum particle agglutination (TPPA) test results were reactive, with the RPR test showing a titre of 1:128. Cerebrospinal fluid (CSF) examination showed a mild lymphocytic pleocytosis and a reactive TPPA test result but a negative RPR test result (Box 1). A diagnosis of otosyphilis was made and the patient was treated with intravenous benzylpenicillin 2.4 million units 4 hourly and oral probenecid 2 g daily for 2 weeks, followed by three doses of weekly benzathine penicillin 2.4 million units intramuscularly. His symptoms stabilised but did not improve. Patient 2 A 30-year-old man infected with HIV presented to an HIV clinic having had tinnitus, hearing loss and imbalance for 3 months. He was referred to ENT clinics in two tertiary hospitals, both of which diagnosed Meniere’s disease. Audiological tests showed mild right sensorineural hearing loss. Serum RPR and TPPA test results were reactive, with an RPR titre of 1:516. CSF examination showed a mildly elevated protein level, but no other abnormalities (Box 1). A diagnosis of otosyphilis was made, and the patient was treated for 2 weeks with benzylpenicillin 2.4 million units 4 hourly. His symptoms resolved completely, and an audiogram performed 6 months after treatment showed that his hearing had returned to normal. The recent increase in early syphilis infections in Australia has been accompanied by the re-emergence of disease manifestations unfamiliar to modern clinicians. Otosyphilis is a rare cause of sensorineural hearing loss and dizziness, and is important for clinicians to consider because the hearing loss is potentially reversible with early diagnosis and treatment. We report two cases of otosyphilis occurring in patients infected with HIV. In both cases, the diagnosis of otosyphilis was initially missed, despite review by several specialist medical units. Cochleovestibular dysfunction is a well described complication of congenital and acquired syphilis. In acquired syphilis, it can occur at any stage of infection. In the pre-penicillin era, hearing loss was reported in 17% of patients with early latent infection and in 80% with symptomatic neurosyphilis.1 Cochleovestibular symptoms of neurosyphilis can occur via two main mechanisms. First, the eighth cranial nerve may be affected, for example in acute syphilitic meningitis. In these situations, hearing loss is usually accompanied by other neurological deficits, and findings on cerebrospinal fluid (CSF) examination will usually be abnormal. Second, and more commonly, hearing loss and vestibular symptoms present without features of coexisting neurosyphilis. These symptoms may occur at any stage of syphilis and are thought to result from direct damage to the vestibulocochlear apparatus. During dissemination, spirochaetes invade the inner ear perilymph, leading to inflammation of the labyrinthine structures and otic capsule. CSF parameters are usually found to be normal but, histologically, fibrosis and ischaemic necrosis of labyrinthine structures are seen2 and endolymphatic hydrops is common. These pathological findings are identical to those of Meniere’s disease, explaining the similar clinical features. Symptoms may be sudden or insidious in onset, and include bilateral (but often asymmetrical) sensorineural hearing loss, tinnitus and vestibular symptoms ranging from dizziness to severe vertigo. These symptoms closely resemble those of Meniere’s disease. Audiological testing shows sensorineural hearing loss, classically affecting low or high frequencies while sparing middle frequencies, and speech discrimination is poor. Without treatment, otosyphilis will progress to profound deafness over months to years. Symptoms can fluctuate markedly over time, but the overall course is one of deterioration.3 There is no established case definition for otosyphilis, but the diagnosis should be made on the basis of a typical clinical presentation and positive serological test results for syphilis. This approach is purposely “over inclusive”, as otosyphilis is a potentially reversible cause of hearing loss. The optimal treatment for otosyphilis is not established. The published literature is limited, consisting of case reports and small case series, but indicates that intravenous therapy is required. Intramuscular penicillin penetrates the perilymph poorly, and there are numerous reports of treatment failure when patients with otosyphilis are treated with penicillin regimens for latent syphilis. In one report, spirochaetes were recovered directly from a patient’s perilymph after treatment.4 Intravenous penicillin G at a dose of 18–24 million units per day, administered as 3–4 million units every 4 hours for 14 days, is the regimen recommended by the United States Centers for Disease Control and Prevention for treatment of otosyphilis. Probenecid is sometimes added, as are subsequent courses of intramuscular or intravenous penicillin.5 There is no high-level evidence to support any of these approaches. Steroids are commonly coadministered, although there are few supporting clinical data. The rationale is that inflammation of the endolymphatic duct appears crucial to the pathogenesis of otosyphilis. A typical steroid treatment regimen is prednisolone at a dose of 0.5–1.0 mg/kg tapered over 1–2 months. Regardless of the penicillin regimen used or whether steroids are employed, treatment outcomes are uniformly poor. Studies consistently show that auditory symptoms abate for only 30% of patients, while 7%–15% have improved results in audiological or speech discrimination tests. Tinnitus and dizziness have better outcomes, with 70%–80% of patients reporting improvement. Factors associated with better outcomes include duration of symptoms less than 5 years, age less than 60 years and fluctuating hearing loss.6 Both of these patients had HIV infection. Rates of syphilis are known to be substantially higher in the HIV-positive population.7 Otosyphilis has previously been described in patients infected with HIV, but relevant published literature is sparse. Patients co-infected with HIV and syphilis appear no different to HIV-negative patients in their clinical features, severity of disease or likelihood of developing this manifestation of syphilis. In both of the cases we report, the diagnosis of otosyphilis was missed despite review by several specialist medical units. In the past few years, rates of early syphilis have risen markedly in Australia, predominantly among homosexual men.8 Relevant practitioners should be aware of this diagnosis in patients presenting with the symptoms described here, especially those at risk of syphilis, such as sexually active homosexual men, including those with HIV infection. Current guidelines recommend syphilis screening in sexually active homosexual men at least annually (up to every 3 months in those at higher risk) and at regular intervals in individuals infected with HIV.9 1 Cerebrospinal fluid and serological test results for two patients with HIV and otosyphilis Patient 1 Patient 2 Cerebrospinal fluid Appearance Clear, colourless Clear, colourless White cell count (× 106/L) 1 polymorph 8 lymphocytes 0 polymorphs 1 lymphocyte Red cell count (× 106/L) 0 0 Protein (g/L) (reference range, 0.15–0.45 g/L) 0.52 0.7 Glucose (mmol/L) (reference range, 2.5–5 mmol/L) 2.7 2.7 Microbiological culture and sensitivity Nil Nil Treponema pallidum DNA polymerase chain reaction test Not detected Not detected Serum Rapid plasma reagin (titre) Reactive (1:128) Reactive (1:516) T. pallidum particle agglutination Reactive Reactive Lessons from practice Consider the possibility of otosyphilis in any patient (especially those with HIV infection or at risk of syphilis and/or HIV infection) presenting with auditory symptoms, particularly sensorineural hearing loss with tinnitus and vestibular dysfunction. Positive serological test results for syphilis will establish the diagnosis. All sexually active men who have sex with men should be screened for syphilis at regular intervals. Otosyphilis must be treated with intravenous penicillin regardless of findings of cerebrospinal fluid testing.

Janet M Pasricha MB BS(Hons) · Tim R Read MB BS, FAChSHM · Alan C Street MB BS, FRACP

To the Editor: We report a case of measles in a 24-year-old man who returned from the Fédération Internationale de Football Association (FIFA) World Cup in South Africa in July 2010. Despite Australian health alerts about measles in South Africa,1 the patient had received no pretravel medical advice or vaccinations. Six days after returning home to the Northern Territory, the patient developed fever, headache and myalgia. The following day he developed vomiting, diarrhoea, and productive cough, with a widespread rash appearing the subsequent day. He visited two general practitioners, was prescribed doxycycline and then admitted to hospital on Day 6 of his illness. On examination, his temperature was 38.9ºC; pulse, 112 beats per minute; blood pressure, 133/72 mmHg; and oxygen saturation, 96% on room air. He had conjunctivitis, a widespread blanching maculopapular rash involving his face, trunk, limbs, hands and feet (Box) and cervical lymphadenopathy. He had bilateral basal lung crackles and tender hepatomegaly. Investigations showed thrombocytopenia (platelet count, 148 × 109/L; reference range, 150–450 × 109/L); hyponatraemia (sodium concentration, 131 mmol/L; reference range, 132–142 mmol/L); and abnormal liver function test results (alanine transaminase concentration, 417 U/L [reference range, < 40 U/L]; alkaline phosphatase concentration, 236 U/L [reference range, 39–117 U/L]). His chest x-ray was normal. The following day, measles virus RNA was detected from a throat swab, and the patient was put into respiratory isolation and therapy with doxycycline ceased. He made a full recovery. The patient reported receiving childhood vaccinations, and while he thought he may have received one measles, mumps and rubella vaccination, he had not had two. Follow-up was required for 84 identified contacts, with no measles cases subsequently notified in the NT. The FIFA World Cup is the world’s largest single-sport event, with an attendance this year of over 3 million people. Mass gatherings may be associated with outbreaks of communicable diseases such as meningococcal disease, measles and pandemic (H1N1) 2009 influenza, in addition to an increased risk of sexually transmitted diseases.2 These risks should be considered when seeing patients who have travelled to such events, in addition to country-specific health risks. Recent data from the GeoSentinal surveillance network showed that a systemic febrile illness was the most common presenting syndrome among travellers returning from South Africa. Most of these cases of illness (54.5%) were due to spotted fever group rickettsiosis.3 The risk of acquiring rickettsiosis increases among travellers visiting game parks, with an incidence of African tick bite fever (Rickettsia africae) among short-term safari tourists of 4.0%–5.3%.4 Measles has rarely been reported in travellers returning from South Africa,2 but the country is in the midst of a measles epidemic, with 17 354 confirmed cases between January 2009 and 12 August 2010.5 Measles presents with fever, cough, rhinorrhoea and conjunctivitis, followed by a widespread rash. The incubation period is usually 7–10 days, but may be up to 18 days. The virus is highly infectious, from 5 days before to 4 days after the onset of rash. Young adults from non-endemic countries such as Australia are at particular risk, as they may only have had one childhood vaccination for measles, with consequent inadequate protection. Although measles has been eliminated in Australia,6 sporadic outbreaks continue to occur,7 and travellers returning from overseas create an ongoing potential for the re-establishment of endemic measles. Given the public health implications of a delayed diagnosis, doctors must be alert to possible cases of measles in travellers returning from endemic countries. The patient’s widespread maculopapular blanching rash

Bridget E Barber · Krispin M Hajkowicz · Vicki L Krause · Kevin G Freeman · Bart J Currie

To the Editor: Allergen avoidance is the mainstay of food allergy management. Consumers with food allergies rely on accurate labelling of foods to avoid ingestion of allergens and subsequent allergic reactions. Current Australian legislation states that ingredients derived from common allergens (peanuts, tree nuts, eggs, wheat, cows milk, soy, fish, shellfish and sesame) must be clearly labelled.1 However, use of shared processing facilities can result in cross-contamination of other ingredients with these allergens. This has led to the use of advisory statements such as “may contain traces of” by manufacturers. A recent Food Standards Australia New Zealand survey found that consumers with food allergies are frustrated by such labelling.2 There is also confusion among the medical profession about whether to advise patients with food allergies to avoid all foods with allergen avoidance advisory statements. The perception by some in the general population and in the medical community is that these statements are so widely used that avoidance would be overly prohibitive. However, there are currently no published data on the extent of advisory labelling use in Australia. We aimed to assess the prevalence of advisory labelling for three common food allergens — peanuts, tree nuts and eggs — on the packages of products for which these allergens were not listed as ingredients. Products containing one type of tree nut (eg, macadamia nut) could still have advisory labelling for other tree nuts (eg, almond). All products were therefore examined for advisory labelling for any tree nut which was not listed as an ingredient. Packages of non-refrigerated processed foods were examined between August and September 2008 at a large supermarket in Melbourne. All product types were examined within each category (eg, for the savoury biscuits category, we examined rice crackers, flavoured wheat crackers and water crackers) and, for each product type, one flavour per brand was selected for examination. Single-ingredient foods, such as flour, sugar, fruit and vegetables, were excluded. Advisory statements included, but were not limited to, “may contain traces of”, “processed on the same line as” and “made on equipment that also processes”. Overall, 761 products were examined. Of these, 384 (50%) carried an advisory statement for one or more tree nuts. Of 737 products that did not list peanut as an ingredient, 348 (47%) carried an advisory statement regarding peanut. Of 641 products that did not list egg as an ingredient, 146 (23%) carried an advisory label for egg. The presence of advisory statements varied between categories of food, with sweet biscuits most likely to carry labelling for peanut and tree nuts and bakery items most likely to carry labelling for egg (Box). Advisory statements have been widely adopted by manufacturers across a range of products, and are likely to limit food choices for consumers with food allergies who avoid all foods labelled with advisory statements. Unfortunately, there is no evidence on the frequency of trace allergen contamination in Australian products, although studies in the United States found that only 10% of 179 products with advisory labelling for peanut contained detectable levels3 and that, as for Australia, this labelling was widely used for some product categories.4 There is also no evidence regarding the proportion of consumers with food allergies who will develop an allergic reaction (including anaphylaxis) to trace contamination of food. Scientific assessment of the risks posed to consumers with food allergies by trace contamination is urgently required. This evidence would allow the development of informative labelling guidelines, including changes to legislation where required, to allow consumers with food allergies to safely manage their allergies, hopefully without the need to avoid entire categories of common foods. Allergen avoidance advisory statements on packaged processed foods at a large supermarket in Melbourne, August–September 2008 Number (%) of products with an advisory statement* Category of food Peanut Tree nuts Egg Sweet biscuits (n = 130) 117 (93%) 120 (92%) 48 (70%) Chocolates (n = 60) 43 (80%) 49 (82%) 2 (4%) Bakery items (eg, cakes) (n = 35) 24 (71%) 30 (86%) 10 (71%) Muesli bars and snack bars (n = 27) 13 (67%) 20 (74%) 4 (15%) Dinner bases and stocks (n = 32) 19 (59%) 9 (28%) 5 (17%) Savoury biscuits (n = 41) 23 (56%) 23 (56%) 20 (51%) Lollies (n = 55) 29 (56%) 25 (45%) 1 (2%) Breakfast cereals (n = 63) 25 (41%) 37 (59%) 3 (5%) Instant noodles (n = 18) 7 (39%) 6 (33%) 8 (50%) Pasta sauces (n = 15) 5 (33%) 4 (27%) 0 Bread (n = 16) 5 (31%) 5 (31%) 5 (31%) Soups (n = 20) 3 (15%) 3 (15%) 4 (21%) Cake mixes (n = 30) 4 (13%) 20 (67%) 11 (58%) Tinned meals (n = 17) 2 (12%) 2 (12%) 1 (7%) Baby foods (n = 30) 3 (10%) 3 (10%) 1 (4%) Pasta (n = 13) 0 0 4 (39%) Chips (n = 20) 0 0 0 Other (eg, tinned fish, breadcrumbs, sauces, custard powder) (n = 139) 26 (19%) 28 (20%) 19 (14%) * Number of products that had an advisory statement but did not have the allergen of interest listed as an ingredient. The denominators used to calculate percentages were the numbers of products within each category of food that did not have the allergen of interest listed as an ingredient.

Jennifer J Koplin · Nicholas J Osborne · Katrina J Allen

To the Editor: Prescribing and dispensing generic medicines is an option to reduce costs in the community and is also common practice in public hospitals. In addition to the issues discussed by McLachlan,1 we have noted a concerning trend in the “branding” of many new generic medicines that has the potential to add to the confusion for patients, prescribers and pharmacists. Medicines with special release properties are branded with suffixes as a reminder of their longer duration of action, such as Sustained Release (eg, Tramal SR; CSL Limited) or eXtended Release (eg, Efexor-XR; Wyeth Australia). Different formulations may also use a suffix to indicate distinguishing properties, such as “dispersible” in Rulide D (Sanofi-Aventis) or “osmotic release oral system” in Adalat OROS (Bayer Schering). However, these suffixes, while meaningful, can be a source of misunderstanding about dosing intervals and length of action, leading to errors.2 A standard nomenclature does not exist, even for formulation descriptors. Adding to this confusion, generic medicine manufacturers are now marketing products with prefixes or suffixes, not to denote a modified formulation but to place their “brand” on the medicine. Ascent Pharmaceuticals has three different suffixes/prefixes for its generic products, reflecting the names of previous manufacturers (eg, Quinapril-GA, GN-Carvedilol), and two suffixes for simvastatin (GA and DP). Spirit Pharmaceuticals adds its name to some products (eg, Simvastatin-Spirit). Taking an oral medication history becomes challenging when patients state they use “Spirit” medication for their cholesterol, “GN” tablets for their heart and “GA” tablets for their high blood pressure. With the many generic “brands” now available (eg, 22 simvastatin products), packs with similar labelling lined up on the pharmacy or patient’s shelves increase the risk of wrong selection. Although no reports have been published to date, similarities in brand prefixes may cause medication errors in electronic prescribing and dispensing systems, when an incorrect medication is selected from a dropdown menu. Entering “Apo” in some systems selects more than 40 products manufactured by Apotex, which uses a naming pattern of the generic drug prefixed by Apo (eg, Apo-Alendronate), as well as APO-go (Hospira), which is apomorphine (and does not contain “go”). This situation could be avoided in future if the National E-Health Transition Authority’s Australian Medicines Terminology and its editorial principles are adopted in systems. These rules require that all medicines are represented by descriptions that list the generic name first, with the sponsor’s name following in parentheses (Paul Frosdick, Chief Terminologist, National E-Health Transition Authority, personal communication, 25 August 2010). The Food and Drug Administration in the United States and the Therapeutic Goods Administration in Australia have developed documents outlining approved medicines terminology,2,3 but there are no official lists of approved suffixes or prefixes. The Institute for Safe Medication Practices, a non-profit US-certified patient safety organisation internationally regarded as an expert in medication safety, has recognised this problem and maintains a list of products with drug name suffixes and their meanings.4 With more generic options available, good communication between medical and pharmacy clinicians and patients is essential to clarify the indication, along with the specific name, of medications prescribed. Pharmaceutical manufacturers need to consider the impact of meaningless prefixes and suffixes, which add to the confusion and potentially contribute to medication errors. National authorities should improve overall governance of labelling of generic medicines, to prevent errors reaching patients.

Linda V Graudins · Michael J Dooley

To the Editor: In recent years, medical school selection criteria have continued to evolve as new information and evidence have come to light.1,2 Increasingly, data are becoming available that shed light on local selection processes, and we commend Laurence and colleagues for their contribution to this.3 Diversity in selection processes is desirable because of the varying educational cultures and emphases between medical schools, and for maintaining variety among students who gain entry. The pursuit of “ideal” evidence-based selection criteria ought to continue, although the effect that a single, standard national selection process would have on the diversity of students entering medicine is unknown. However, a national approach to offering places for medical school could have considerable advantages. Mutual recognition of admission processes by medical schools is one approach suggested by Laurence et al.3 This may not have as significant an impact on student diversity as might a single selection process, and it could reduce the substantial financial burden that the application process places on both applicants and medical schools.3 Students from low socioeconomic backgrounds are often unable to apply to interstate medical schools due to the travel costs associated with attending interviews. This both reduces the number of opportunities they have of gaining a place at a medical school, and prevents them from gaining valuable interview experience, further reducing the likelihood of a successful application.3,4 Mutual recognition of and sharing of selection information would potentially level the playing field for applicants who cannot afford to attend multiple interviews. Students commonly receive medical school offers as late as midway through the first semester — a consequence of students receiving offers from multiple universities, and universities having to make a number of rounds of offers until their quotas are complete. Universities already rank their applicants, and, if students were able to submit their medical school preferences, applications could be coordinated nationally through a centralised admissions system. This would result in fewer rounds of offers and a shorter admission process, reducing the cost for universities and enabling students to enrol earlier, with greater certainty. This may avoid the need for students to relocate and miss out on vital parts of their first year of study.

Christopher X J Wong · Adam J Nelson · Ross L Roberts-Thomson

To the Editor: In their recent editorial, Bower and colleagues effectively summarised the problem of apparent governmental indifference to congenital anomalies.1 This is not unique to Australia and probably exists worldwide. National systems to collect congenital anomalies data were set up in many countries in the mid 1960s, including the National Congenital Anomaly System in the United Kingdom and the Canadian Congenital Anomalies Surveillance System. This followed the thalidomide tragedy and exemplified that it often takes an acute crisis to stimulate politicians into action. However, due to a lack of leadership, foresight and finances,2-4 these systems gradually declined to the extent that they became of very little value, lacking in accurate ascertainment and pregnancy termination data. As a result, regional registries were set up in England and Wales, and Canada was left with only two provinces (British Columbia and Alberta) collecting data. Prevention is one of the new driving forces for collecting good data, and the advent of using folic acid to effectively reduce neural tube defects brought a new urgency to the need for comprehensive useable data. Accordingly, the Canadian government set up a task force and formed a new entity in 2002, the Canadian Congenital Anomalies Surveillance Network, with a mandate to provide logistical and financial help to all 10 provinces and three territories. While progress has been slow, it has been very encouraging, with three additional provinces and one territory developing new surveillance systems this fiscal year (April 2010 – March 2011). The Network has set up guidelines and standards to enable all provinces and territories to collect data in a comparable format,5 which can then be forwarded to a central database in the national capital, Ottawa. The quality of the data should be improved because they are gathered at a local level. This model could be adapted for use in Australia because, according to Bower et al,1 a nucleus of good data from at least three states already exists.

R Brian Lowry

To the Editor: I agree with Emery and colleagues1 that family history can add much to downstream clinical interventions that provide tangible benefits to the presenting patient and his or her kin. The pedigree chart has some advantages over simple narrative recording of the same details, including the ability to instantly visualise relationships between individuals and the ease with which the chart can be updated and annotated.2 There is a familial aggregation (commonly an affected first-degree relative) in up to 25% of presenting cancer patients, while around 5% will harbour a highly penetrant genetic predisposition to cancer. In the cancer clinic, an acceptably detailed family cancer pedigree can typically be obtained in even less time than the 30 minutes suggested by Emery et al.1 Steps in drawing pedigrees have been outlined elsewhere2,3 and typically involve collecting information such as simple demographics, naming and symbolising different cancer types, and recording age of onset and age of death (if relevant) for the different individuals, starting with the presenting patient. The time required to construct a three-generation pedigree would typically be around 10 minutes, making it an attractive addition to routine history taking. If a cancer pattern emerges, the pedigree should be extended as far as possible. Of course, diagnosis verification (through death registries, etc) would be required before surveillance, prophylaxis and therapy decisions are addressed in the familial cancer setting. A convenient refresher for doctors who do not routinely draw pedigrees might be to first construct their own family tree, with reference to the symbols and relationship illustrators commonly used.2,3 It is likely that pedigrees in most clinics will be drawn by hand, at least initially, although there are software programs for pedigree creation available (eg, Family Tree Builder; <http://www.myheritage.com/family-tree-builder>). Over the years, I have informally asked my specialist oncology trainees to routinely construct family pedigrees. I do not recall, among these highly clinically skilled doctors, one that was able to correctly draw a family pedigree until we had worked on it together. Given the benefits of family pedigree analysis across many disease categories, a little practice in pedigree drawing may be a useful exercise for many of us. Even without the requisite confirmation of disease status of the individuals represented, a simple pedigree chart, combined with quick reference to information on the potential clinical significance of any patterns it shows (eg, by consulting National Health and Medical Research Council guidelines),4 can help prioritise referrals to busy familial cancer clinics.

Michael J McKay

To the Editor: Langlands and colleagues show that family history is poorly taken in patients presenting to an acute medical unit at a major tertiary hospital.1 However, their article and its companion editorials2,3 do not adequately stress the settings in which family history taking may be both easily achievable and most cost-effective. The nihilism that Thomas and Thompson2 convey about recording family history in the acute setting is of more concern, given that a tertiary hospital may offer the best opportunity to initiate the process of case detection for a number of heritable and lethal diseases, such as autosomal dominant familial hypercholesterolaemia (FH), the most common monogenic cause of premature coronary artery disease (CAD). We previously demonstrated that a family history of cardiovascular disease was almost never recorded by coronary care unit medical staff at Royal Perth Hospital.4 For 509 patients aged < 60 years presenting with symptomatic CAD to the coronary care unit, we found that 70% had insufficient clinical data documented in the medical records to enable a diagnosis of FH. In a follow-up study of 103 patients with premature CAD admitted to the coronary care unit, a nurse practitioner was able to record a positive family history of premature cardiovascular disease in a primary relative for 43% of patients, of whom 95% had phenotypic FH based on a recognised clinical diagnostic tool (the Dutch Lipid Clinic Network score).5 Patients detected in this way in Western Australia are now referred to a statewide FH program run by staff from a lipid clinic.5 In this program, where detailed pedigree drawing and family tracing is performed by trained nurses, we have found a causative mutation for FH in up to 85% of patients with a clinical phenotype strongly suggestive of FH. Additionally, we find that for every index case so detected, we can additionally diagnose at least three new cases of FH in the patient’s relatives, many of whom are young. This method of case detection and subsequent treatment with cholesterol-lowering therapies is highly cost-effective and, more importantly, enables therapy to be targeted at younger patients, thereby maximising the potential for preventing CAD. Our experience illustrates that for a lethal condition such as FH, an accurate family history recorded by a nurse can spark a cascade of action that leads to a definitive diagnosis of FH in the index patient and the subsequent detection of otherwise undiagnosed FH in the community, with significant associated cost savings.6 Hence, we propose that nursing staff can efficiently bridge this gap in medical care while we are getting our house in order by training medical staff to effectively take a family history.

Timothy R Bates · Elissa B Poulter · Frank M van Bockxmeer · Gerald F Watts

To the Editor: I read the article by Langlands and colleagues1 with some dismay, and a sinking heart. Their report is further evidence of the dangers of moving away from the basic skills of comprehensive history taking and performing a detailed physical examination. Interestingly, in the United States, the debate regarding performing a physical examination has come full circle, from virtually ignoring its importance to now telling us how vital it is and how to perform it.2 But it is the taking of a comprehensive history that, as one of my mentors told me, “is where the money is”, and a detailed family history is an integral part of this. Eliciting a detailed family history is arguably more important for paediatric patients, who have a longer potential life span and hence have more to gain from this information. In my paediatric practice, for example, I see numerous overweight children, some of whom have a strong family history of hypercholesterolaemia, vascular disease or type 2 diabetes mellitus, which places them at considerable risk of cardiovascular disease in their adult years. A detailed family history may also “unmask” the genetic contribution to a child’s history of deafness or learning disability. Time constraint is the main impediment to taking a comprehensive family history, but it is worth keeping in mind that it is time well spent and that, in the paediatric population, it may make a significant contribution to the long-term health of the child. With the impending advent of personalised genomic screening, there will be an even greater imperative to formalise the gathering of family history details.3,4

Simon E P Hauser

To the Editor: Pridmore1 describes a case in which a man’s recent actions suggested suicidal intent. The man told police he had no ongoing suicidal plans and they took his words at face value. He subsequently killed himself. The High Court exonerated the police of any responsibility, a decision that seemed based on two premises: that suicide does not presuppose mental disorder; and that “There is no general common law duty of care to rescue a person from harm, including self-harm”.2 There is sufficient grey in both those inter-related premises to make a black-and-white judgment suspect. I share Pridmore’s position that suicide does not always equal mental disorder, although one wonders whether situational crises in individuals with subtle vulnerabilities could be subsumed under such a label. More important is the question of individual autonomy. The reason we have no common law requiring us to “rescue” another adult is because we set such a high value on autonomy. Although we assume a person’s competence, our curiosity about it should be aroused when people behave in unexpected ways. When the behaviour is strikingly different, and has potential for serious harm, are we not obliged to intervene or procure assessment? This is, I would have thought, a moral rather than legal or medical concern. This is what we would do for a child wandering on the road or a demented person lost at night. We might be free of legal or medical censure for ignoring them, but we would be embarrassed to publicly admit our failure to act if we might have done something useful at the time. Making serious preparation to kill yourself is strikingly different behaviour, and should raise questions about both your competence and autonomy. Perhaps not 100% but surely more than 50% of such people have a disorder. A person preparing to suicide is “more likely than not” mentally ill, and that is how the “common person” would surely see it: “guilty”, as it were, till proven innocent. Expertise is called for to make that final determination. The police certainly do not have the expertise. Why, then, would they not seek it? Furthermore, if the “rational, cooperative and responsible” man in this case had indicated his plans to “rationally” commit suicide to escape an intolerable predicament, it is hard to believe the police would have walked away, even though he may have been competent to make such a decision!

Paul T Dignam

To the Editor: The editorial on the Human Variome Project by Cotton and Macrae1 neatly lays out the reasons for government funding for gene mutation databases for inherited diseases. Most of these diseases are rare, but collectively, they are common. The article highlights the key principles of detecting gene mutations and establishing pathogenicity in order to offer individuals (or couples) relevant health information for themselves and/or their (future) offspring. Cotton and Macrae mention a number of specific diseases, but do not mention the commonest life-shortening inherited disease affecting Australian children — cystic fibrosis (CF). Far from being theoretical, nearly all of the principles outlined by Cotton and Macrae are already in place for CF, including clinical databases (in Australia, the Australian Cystic Fibrosis Data Registry), an international gene mutation database (at http://www.sickkids.on.ca, which is contributed to by Australian genetics laboratories) and programs to offer carrier screening to the population. Unfortunately there is very little government funding for these initiatives, so they are not coordinated. In particular, screening for CF carriers in the population, which is of considerable clinical utility, has only small, fee-for-service programs that reach very few people.2,3 These programs are inequitable in that many people are unaware of the existence of such screening programs and, of those who are, many cannot afford the cost of testing. CF provides an excellent model for the development of a coordinated approach to inherited disease screening and, given that 800 000 Australians are carriers of CF mutations, funding CF screening should be a major government priority.

R John Massie · Martin B Delatycki

To the Editor: Costello and Murphy’s lament1 about the discoverer of prostate-specific antigen (PSA), Richard Ablin, describing PSA testing as a “hugely expensive public health disaster”2 contains several egregious claims that require correction. They write that “since the introduction of PSA testing in the 1980s, we have seen a 25% reduction in mortality” from the disease. Thirty years ago (in 1980), before PSA testing was possible, the age-adjusted mortality rate from prostate cancer in Australia was 33.4/100 000. In 2007, it was 31.0/100 000, a decline of 7.2%. Over the same period, prostate cancer incidence rose 110%, from 80.8/100 000 to 170/100 000,3 thanks to the aggressive promotion of PSA testing. Recent New South Wales data show that 3 years after radical prostatectomy, 77.4% of men are impotent and 12.3% have urinary incontinence, compared with 22.3% and 1.0%, respectively, of controls.4 Many of these men are, to use Costello and Murphy’s word, “overtreated”1 — they underwent unnecessary surgery and now live with the consequences. From a 2009 European trial,5 the take-home message for a man being tested today is: There is a one in 50 chance that, in 2019 or later, he will be spared death from a cancer that would otherwise have killed him. And there is a 49 in 50 chance that he will have been treated unnecessarily for a cancer that was never a threat to his life.6 This is what Ablin called a public health disaster. With reference to this European trial,5 Costello and Murphy claim that “PSA testing has led to greatly reduced mortality”. This “great reduction” was from 4.2 to 3.3 deaths per 10 000 person-years. Costello and Murphy propose that “those with a PSA level well below the median for men in their 40s . . . (the vast majority at this stage) could be reassured . . .”, but below the median lie half the men, not the vast majority. This policy would mean that the other half of men aged 40 (not the small minority, as implied by Costello and Murphy) would be above the threshold. They would not be reassured; presumably they would be offered more frequent testing and follow-up. Labelling half the population of men aged 40 as higher risk has enormous implications for the men (anxiety, inconvenience, cost) and the health system that would be called on to fund more appointments and tests. Such a proposal does indeed sound like a public health disaster.

Simon Chapman · Alexandra Barratt

In reply: Chapman and Barratt’s statements about prostate cancer highlight the dramatic difference between the views of sociologists and the views of clinicians who deal daily with the burden of prostate cancer diagnosis, treatment and mortality. In 1851, prostate cancer was considered a rare disease.1 Chapman and Barratt’s views seem to remain consistent with this thesis. However, in 2010 in Australia, prostate cancer is the most common cancer diagnosed in men, and the most common cancer causing death in men. Our own data do not support the contention that treatment for prostate cancer produces dreadful outcomes. We can reassure men that there is a high likelihood of cure through early treatment for prostate cancer, with a less than 5% chance of becoming incontinent and a 70% chance of retaining erectile function.2 Clearly, Chapman and Barratt have not taken into account a recently reported Swedish study which showed that in 20 000 men randomly allocated to prostate-specific antigen (PSA) screening or a control group between 1994 and 2008, there was a 50% reduction in mortality from prostate cancer for men in the screened arm.3 The numbers needed to screen (293) and treat (12) from this study are almost exactly the same as those from breast cancer screening studies. We repeat our recommendation for the early use of a single PSA test4 in men aged in their 40s. This is most beneficial because the background noise from benign prostatic hyperplasia development does not occur, and PSA level is very discriminatory for detecting those who go on to develop significant prostate cancer.

Anthony J Costello · Declan Murphy

To the Editor: The long-term functional outcomes in the three cases of swimming pool filter-induced transrectal evisceration described by Price and colleagues1 are excellent and significantly better than many other cases described in the literature. However, it may be possible to improve further on such results, or at least decrease short-term morbidity, by expediting the reduction of the eviscerated bowel. In all three cases described,1 the children presented initially to a local hospital and were subsequently transferred to a tertiary care facility before operative reduction was initiated. It is likely that earlier operative reduction of the eviscerated bowel would decrease secondary venous congestion of the prolapsed segment, thus minimising further ischaemic changes already initiated by the mesenteric arterial and/or venous tear. This in turn may improve the perfusion of the affected segment; lessen the development of hypothermia, hypovolaemia and sepsis; and possibly increase the final length of viable bowel. A “damage control” laparotomy2,3 performed at an appropriate hospital of initial presentation (one with a surgeon and anaesthetist available), with the principal aim of early reduction of the prolapsed bowel into the abdominal cavity, may improve outcomes in such cases.

Phillip J Carson

To the Editor: We read with interest the recent meta-analysis of rates and methods of suicide in Australia by Large and Nielssen.1 Previous research has shown a marked increase in hanging suicides in women in South Australia over the 15-year period 1986 to 2000,2 which concurs with the findings of Large and Nielssen at a national level between 1988 and 2007.1 One of the features of this increase in hanging suicides, which was not specifically addressed in their excellent overview, relates to trends in specific age groups such as the young. A previous analysis of suicides in individuals aged 16 years and under and for whom autopsies were performed at the Forensic Science Centre (now Forensic Science South Australia; FSSA), in Adelaide,3 included a total of 19 cases of suicide for the 5 years 1985 to 1989, of which seven deaths (37%) were from hanging. The age range of the hanging victims was 14–16 years (mean age, 15 years), with a male to female ratio of 6:1. This compares with a total of 10 suicides among young people, recorded in FSSA files for the 5 years 2005 to 2009, of which nine (90%) were by hanging. The age range of the hanging victims in this group was 10–16 years (mean age, 14.7 years), with a male to female ratio of 4:5. The difference between these two periods was statistically significant (P < 0.005).3 Thus, although suicide in those aged 16 years and under remains an uncommon event,3 developments in deliberate self-harm among the general population may also be reflected in the young. While there was a fall in the total number of young suicide victims, from 19 to 10, between the two periods in the SA study, this has not been matched by a decline in hanging suicides (ie, a significantly greater percentage of suicides in the young in SA now involve self-suspension).3 This finding may be of use to those studying specific issues and trends in youth suicide.

Roger W Byard · Amy Austin · Corinna van den Heuvel

Making sense of acute medicine. A guide to diagnosis. Paul F Jenkins, Paula H Johnson. London: Hodder Arnold, 2010 (xii + 308 pp). ISBN 9780340984253. Medical diagnosis is traditionally learnt at the patient’s bedside in the presence of an experienced clinician. Time and workload pressures restrict the teacher–student relationship within a hospital. Sometimes, the clinical signs and wisdom are pieced together independently or, at best, with fractured feedback. Clinical decisions can be delayed by plentiful, irrelevant and risky investigations. Jenkins and Johnson’s book provides the inexperienced doctor or medical student with memorable material to facilitate clinical decision making. It has competition internationally but is peerless. The authors’ excellent work focuses in a very practical way on history taking and physical examination, and the rational use of investigations. Both authors have experience in Australian general medicine. Jenkins was also one of the founders of acute medicine — the specialty responsible for diagnosis and initial management of most admissions to hospital — in the United Kingdom. Acute medical units have transformed hospital general medicine overseas and are doing the same in Australia and New Zealand. Such units are a great environment for learning about diagnosis. The text is structured to cover, in a systematic way, common medical presentations (eg, “dizziness”, “breathlessness” and “weakness”). It is well organised and indexed and has easy-to-read hints and highlights that keep the reader engaged. The lists are not exhaustive but are clinically relevant to an Australian readership. X-rays illustrate common pathological lesions in the differential diagnosis. The intended readership includes medical students and junior medical trainees rather than senior clinicians. The book has limited scope; it will assist diagnosis of common conditions and prevent unnecessary investigations rather than guide prognosis, inform treatment or detail complicated pathophysiology. Overall, it is an excellent balance between affordability, portability, readability, erudition and necessary detail. I have difficulty in wresting it away from any student who visits my office.

Campbell Thompson

Primary hyperparathyroidism (PHT) complicated by osteitis fibrosa cystica (OFC) — the “classical” form of PHT — is rarely seen today. A 41-year-old woman of Sri Lankan descent presented with persistent pain in her right distal forearm 2 days after chopping vegetables. X-ray revealed a fracture through a lucent lesion within the midshaft of the right ulna. A whole-body bone scan showed numerous abnormalities of the major long bones consistent with OFC (Figure, A). Skeletal x-rays showed widespread lytic lesions with osteopenia and subperiosteal erosions, illustrated here by x-ray of the right hand (Figure, B) where marked subperiosteal bone resorption can be seen; note the ill-defined phalangeal cortex (thick arrow) and erosion of the terminal tufts of the distal phalanges (thin arrow). Skull x-ray showed “salt-and-pepper” demineralisation (Figure, C), best appreciated by the lack of visible vascular markings on the calvarium; lytic lesions are present (arrows). The serum calcium level was 4.22 mmol/L (reference range, 2.13–2.63 mmol/L). Our patient was successfully treated with a parathyroidectomy — 5 months after surgery, bone turnover markers were normal, and 14 months after surgery, bone mineral density, tested at the hip, had increased by 22%.

Anna Lih · Mridula Lewis · John Carter

The casket was not carried out, instead was wheeled. Appalled family walked either side, his mother riding in a chair behind, reaching out a black-clad arm to touch then hold the lid somewhat obscured by flowers. He was forty-eight. Voices resistant to the medications, having torn his soul with strident assertions, instructed him how to plait his hair, to tie the knot, to place the chair and kick it out. This is the best that we could do — we are short of resources for mental health. We have to fund forces in Afghanistan, invest in new buildings, do countable things. He was an artist at times pencilling the Blue Mountains, arcades in The Rocks. After years his wife left him, a frowning alien cut off from relatives who found him, cut him down.

Stephen Leeder

Hair-raising stress levels Hair samples may provide insight into stress levels preceding myocardial infarction (MI), according to a Canadian study. Although the association between acute stress and MI is well recognised, research into the role of chronic stress has been limited by recall bias and lack of a reliable biological marker. The stress hormone cortisol, which accumulates in hair, may be just such a marker. As hair grows at a rate of about 1 cm per month, the researchers measured cortisol levels in the proximal 3 cm of scalp hair to reflect stress levels in the previous 3 months. The study, claimed to be the first to use hair samples for this purpose, found that median cortisol levels were significantly higher in the 56 MI patients (P = 0.006) compared with controls. On multivariate analysis, hair cortisol levels had a stronger correlation with MI than other risk factors such as body mass index and cholesterol levels (OR, 17.4; 95% CI, 2.15-140.5). Patients with raised hair cortisol levels may therefore benefit from more aggressive management of cardiovascular risk factors. Stress 2010 online. doi:10.3109/10253890.2010.511352 Warts and all Treatment options for warts include cryotherapy with liquid nitrogen, topical salicylates and sometimes a “wait and see” approach, due to their benign natural course — but which is best? A Dutch study has shown that cryotherapy is the most effective, at least for common warts. The study randomised participants to one of three arms — cryotherapy every 2 weeks, topical salicylic acid (40%) or a “wait and see” approach, which allowed the use of over-the-counter treatments only. At 13 weeks, cure rates for common warts were 49% (95% CI, 34-64), 15% (95% CI, 7-30) and 8% (95% CI, 3-21) respectively. However, there was no difference between any of these treatments for plantar warts, which tended to be more difficult to treat, especially in adolescents and adults, where the best cure rate was only 5%. Despite better treatment outcomes for common warts with cryotherapy, it was associated with a higher incidence of side effects such as pain and blistering. CMAJ 2010. doi:10.1503/cmaj.092194 Double take on D-dimers Elevated D-dimer levels (normal range, < 500 ng FEU[fibrin equivalent units]/mL) are routinely used to help exclude venous thromboembolism (VTE), but may also indicate underlying malignancy in patients without VTE, according to a UK study. The incidence of underlying malignancy was found to be 27.5% in those with very high levels (> 8000 ng FEU/mL). D-dimer levels over 8000ng FEU/mL were also associated with reduced overall survival (35 months v 65 months). Although expressing caution related to poor standardisation of assays, the study authors suggest that the presence of a very high D-dimer level in the absence of VTE should alert for the presence of underlying malignancy. J Clin Pathol 2010; 63: 818-822 doi:10.1136/jcp.2010.076349 Adherence a sticky issue Medication adherence can be a challenging problem in children and adolescents with chronic illness. In a systematic review of 17 trials, Queensland researchers have found that adding behavioural interventions — such as monitoring, goal-setting, reward, problem solving and linking medication taking with established routines — were more likely to improve medication adherence than education alone. These findings suggest that adherence may be improved by incorporating strategies such as regular follow-up and simple behavioural techniques into existing practice. However, few of the trials included patients with established adherence problems. Arch Dis Child 2010: 95; 717-723 doi 10.1136/adc.2009.175125 Need for needle exchange Yale researchers have found that hepatitis C virus (HCV) can survive for up to 2 months in contaminated syringes. The series of experiments using a newly developed assay replicated common injecting practices among drug users, and tested two types of syringes loaded with HCV spiked blood; an insulin syringe with a permanently attached needle containing a low void volume of blood, and a tuberculin syringe with a detachable needle, with a high void volume. Syringes with a high void volume were considered more likely to transmit HCV, as a result of increased viral load and longer survival. Temperature also played a role, with lower temperatures (4°C) tending to preserve viability of the virus, especially in the low void volume syringes. HCV was found to survive for as long as 63 days in the high void volume syringes. Such findings underscore the importance of needle-exchange programs in reducing the spread of hepatitis C among injecting drug users, and suggest that exchange programs should encourage the use of low void volume syringes. J Infect Dis 2010: 202; 984990. doi:10.1086/656212

Alison Williams