Volume 181 Issue 8

There is yet no convincing evidence that testosterone therapy is safe or effective in counteracting any effects of ageing One of the most challenging and time-consuming aspects of medical practice is discussing with patients what they have gleaned from the Internet or popular lay publications. One topic of such discussion is the use of testosterone in male ageing, with public interest fuelled by persuasive publications like Maximising manhood,1 The testosterone revolution2 and Male menopause.3 Much of the information gained from these books may seem to be very convincing, but is, at best, “ahead of the evidence”. It is often difficult for doctors not to appear old-fashioned, ignorant or downright contrary when showing scepticism or advising caution in the face of such conviction. . . . beware when entrepreneurial business ventures overtake evidence-based medicine. In this issue of the Journal, Handelsman (page 419) has documented temporal and regional trends in testosterone prescribing in Australia.4 He has shown increases in prescribing corresponding to popular promotion of androgen therapy in ageing. This rise has been particularly marked in Western Australia, coincident with the opening in that state of entrepreneurial clinics aimed at the ageing-male market. There is no doubt that testosterone therapy benefits patients with documented hypogonadism associated with conditions such as Klinefelter’s syndrome and hypopituitarism. Current prescribing guidelines allow for this, regardless of age. However, for the use of testosterone in men of middle age and older with borderline low serum levels of testosterone, the evidence for both efficacy and safety is yet to be established. The original description of the efficacy of testosterone therapy for the vicissitudes of age has been shown to be no more than a powerful placebo effect.5 Nevertheless, it would seem that those who espouse testosterone therapy in this setting have identified an expanding market. The Australian population is ageing: 28% of men are over the age of 50 and 12% over the age of 65 years. It is projected that by 2021, 4.0 million Australians will be over the age of 65, an increase of 1.7 million from today’s estimate.6 Together with a documented fall in testosterone level with age, where it is estimated that 20% of men over the age of 60 will have a testosterone level below the reference range, this predicts a large population of men as potential customers for androgen therapy.7,8 Although most studies have shown a gradual decline in testosterone level with age, the clinical consequences of this are not known. Conditions such as muscular frailty, loss of bone mass, cognitive decline and erectile dysfunction are also age related, but whether testosterone deficiency has a causal role is not clear. More importantly, it has not yet been established that testosterone therapy has any role in correcting these conditions. It has been a widely held belief that androgen deficiency is a common and correctable cause of erectile dysfunction. However, in a cohort of 1455 men presenting with erectile dysfunction, we found that testosterone deficiency constituted a correctable cause in only 3%.9 All of those who responded to testosterone therapy for erectile dysfunction had a testosterone level < 7 nmol/L, well below the reference range of 11–37 nmol/L. The use of testosterone therapy in patients with normal or slightly depressed androgen levels seems to be of little benefit.10 A similar finding has been documented in the response of bone density to testosterone therapy. The largest and longest-term study in this setting examined testosterone treatment in 108 men aged over 65 years with a baseline testosterone level more than 1 standard deviation below the young adult mean (ie, < 16.5 nmol/L).11 Lumbar spine bone density rose significantly only in those with a baseline testosterone level below the reference range (ie, below 10 nmol/L). Another, smaller study has shown a significant increase in lumbar spine density with androgen therapy in men with a baseline testosterone level < 12.1 nmol/L.12 Muscle mass and strength decrease with age, and it is tempting to link this with the decline of testosterone. In hypogonadal men, testosterone therapy improves muscle strength. In men aged over 65 years, testosterone treatment leads to a fall in fat mass and a rise in lean body mass, together with a perception of increased strength. However, although a perception of improved muscle function was reported, no objective change in measured physical function, such as walking or stair climbing, knee flexion or in muscle strength as measured by dynamometer, was demonstrated.13,14 The safety of sex hormone therapy in menopausal women, used widely for the last 60 years, has recently been scrutinised and reassessed.15 Knowledge about the safety of sex hormone replacement in men is, by contrast, in its infancy. No studies have been conducted for long enough to identify the long term risks of androgen therapy in ageing. Because of lack of statistical power in studies, it is not known whether testosterone use will increase the incidence of androgen-dependent disease such as prostate cancer, although a sustained rise in prostate-specific antigen level has been observed.13 Two other side-effects of testosterone treatment have been identified in short term studies, namely an increase in obstructive sleep apnoea and increased haematocrit and polycythaemia, both of which may have implications for cardiovascular risk.13,16 Handelsman has documented the effect on testosterone prescribing of entrepreneurial clinics aimed at the ageing-male market.4 Daily, we see advertisements for similar commercial enterprises which deal in unregistered and/or unproven prescribing, including DHEA (dehydroepiandrosterone) or testosterone to men and women, progesterone cream, and lozenges of variable mixes of oestrogens, progestins and androgens. We should beware when entrepreneurial business ventures overtake evidence-based medicine. What is needed now regarding testosterone therapy in ageing are large, long-term, prospective, randomised, placebo-controlled studies to establish if there is, indeed, benefit for specific symptoms and to identify potential risks.17

Bronwyn GA Stuckey BA, FRACP

Many maternal deaths in Australia are still preventable The recently released report Maternal deaths in Australia 1997–1999 is the twelfth in a series of triennial reports, spanning over 35 years.1 The maternal mortality ratio (MMR) for this triennium was 8.2 per 100 000 confinements — one of the lowest national MMRs in the world. The MMR of a country reflects, in part, the overall health of women, the quality of the maternity services, and women’s access to those services. For comparison, the MMR per 100 000 confinements for the United Kingdom over this period was 11.4 and for the United States (for the year 2000), 17.0; Sweden had the lowest rate (2.0).2,3 There are huge differences in MMRs between developed and developing countries, with about 99% of global maternal mortality occurring in Africa, Asia and Latin America. The World Health Organization estimated that, in the year 2000, the MMR for Sierra Leone was 2000 per 100 000, or one death for every 50 confinements, compared with Australia’s one death for every 8500.3 While it is easy to see the magnitude and urgency of the problem of maternal mortality in developing countries, what, if anything, can be learned from analysing the rare maternal deaths in countries such as Australia? The answer lies in the fact that mortality represents the “tip of the iceberg” of severe morbidity; for every case of mortality directly or indirectly caused by pregnancy or its management, there are probably 50 women who experience a life-threatening complication but survive with varying degrees of short- and long-term sequelae. Improvements in obstetric care, which result from considering maternal mortality, can therefore be expected to apply to a broad population of childbearing women and their infants.4 Australia has witnessed a one-third reduction in the MMR over the past 30 years, from 12.7 per 100 000 confinements in 1973–1975 to 8.2 per 100 000 in 1997–1999. Higher rates of maternal mortality are seen for older women and for Indigenous women (Box). The highest risk of death was among women aged 40 years and older, who form an increasing proportion of the childbearing population, and who are more likely to use assisted reproduction techniques for conception, with higher multiple pregnancy rates. The inequity seen in all-cause mortality among Aboriginal and Torres Strait Islander women is reflected in the higher MMR for this population, who, in this triennium, represented about 3% of confinements, but 8% of deaths. There were no deaths in the triennium from terminations of pregnancy. If estimates are correct that over 80 000 termination procedures are performed in Australia annually, then these procedures carry exceptionally low risks of serious perioperative complications. Several conditions warrant special mention. Obstetric haemorrhage is still the major direct cause of maternal mortality (deaths directly caused by complications of pregnancy or its management). A recent report from Victoria supports the clinical impression of an increasing incidence of emergency hysterectomy for severe obstetric haemorrhage, possibly related to the increasing proportion of women embarking on pregnancy after one or more caesarean sections, with an increased risk of the more severe forms of abnormal placental implantation — placenta praevia, accreta and percreta.5 This is another change in the characteristics of the obstetric population which warrants close observation through systematic and comprehensive maternal mortality analysis. The Australian and UK reports encourage maternity institutions to rehearse the management of simulated cases of obstetric haemorrhage, which is happening in some, but not all, Australian institutions. Amniotic fluid embolism was responsible for seven deaths in Australia in the 1997–1999 triennium. Although small numbers are open to overinterpretation, it is noted that, in five of these cases, labour was induced. There is community and clinical concern about adverse consequences of induction of labour, especially when undertaken for non-clinical indications. We reiterate the need for maternity units (of all sizes) to have disaster plans for the immediate management of the rare and unpredictable cases of amniotic fluid embolism, which usually present with acute syncope and haemorrhage. Thromboembolism and severe hypertension continue to be major contributors to maternal mortality, with six deaths in each category. Some of these deaths were considered preventable, and, with the trend towards more obesity, higher caesarean rates and women delaying childbirth into their 40s, awareness of the indications for perioperative thromboprophylaxis is increasingly important. Adherence to clinical practice guidelines for severe hypertensive disorders, including earlier specialist referral, would have prevented at least some of the six deaths from hypertensive disorders. Deaths related to cardiac disease and psychiatric illness dominate the category of indirect maternal deaths (deaths in which the pregnancy was complicated by pre-existing disease). Eight deaths due to psychiatric causes mirrored what has been found in the UK, where increasing numbers of maternal deaths are reportedly due to psychiatric conditions. Clinicians and the community need to be reminded about not ignoring or dismissing symptoms of depression, especially suicidal intent, during and following pregnancy.2 There is an emerging group of women at risk of maternal death — those who have had surgery for major congenital heart disease as infants or children, and who are now attaining reproductive age. These women are at risk of severe cardiac complications during and after pregnancy, and require sophisticated and multidisciplinary management. Despite a very low risk of death associated with pregnancy, no-one should claim that, in a country like Australia, mere survival rates could serve as an indicator of the quality of maternity services. However, because of the implications for the broader population of women experiencing pregnancy and childbirth, maternal mortality warrants continuing systematic ascertainment, and high-level investigation and reporting, with reasonable timeliness, so that recommendations are relevant to current practice. Key findings of Maternal deaths in Australia 1997–19991 The 1997–1999 maternal mortality ratio (MMR) was 8.2 deaths per 100 000 confinements, compared with 9.1 per 100 000 in 1994–1996. There were 90 maternal deaths: 34 direct, 28 indirect, and 28 incidental deaths. The main causes of death were obstetric haemorrhage (8 deaths), psychiatric disease (8 deaths), amniotic fluid embolism (7 deaths), and cardiac disease (7 deaths). The highest risk of death was seen in women aged 40 years and older (MMR, 23.2 deaths per 100 000 confinements, compared with 4.0 deaths per 100 000 confinements for those aged 20–24 years). The MMR for Aboriginal and Torres Strait Islander women remains three times higher than the MMR for non-Indigenous women.

James F King MPH, FRANZCOG, FRCSC · Emma K Slaytor MPH, BMedSci(Hons) · Elizabeth A Sullivan MPH, MMed, FAFPHM

Establishing a causal link between HPV and certain head and neck cancers would have implications for prognosis, prevention and therapy More than half a million cases of squamous cell carcinoma (SCC) of head and neck mucosa are diagnosed each year, making it the eighth most common malignancy globally.1 The highest incidence has been observed in developed countries like Australia, where there were 853 new cases in 2001, accounting for 3.8% of all cancers in men and 1.9% of all cancers in women.2 Tobacco and alcohol consumption are accepted as the two major risk factors for head and neck cancer,3 and micronutrient deficiencies, genetic factors and poor oral hygiene are also suspected of playing a role.4 Over recent years, both epidemiological and experimental evidence has implicated oncogenic human papillomavirus (HPV) in the development of a subset of head and neck cancers.5-7 This evidence has been strongly supported by a recent large case–control study coordinated by the International Agency for Research on Cancer (IARC).8 The study, involving centres in nine countries (including the Royal Prince Alfred Hospital in Sydney), concluded that HPV plays an aetiological role in many oropharyngeal cancers and possibly in a small group of cancers of the oral cavity. In all studies, the association has been strongest in the tonsil, where HPV DNA positivity rates frequently exceed 50%. HPV type 16, predominant in cervical cancers, is overwhelmingly the most common type.8-10 The critical issue now is to confirm that the association of HPV with these cancers is causal, as is clearly the case with cervical cancer. This may have implications for prognosis, treatment and prevention. Both the mutagens in tobacco smoke and alcohol and the HPV oncogenes E6 and E7 are known to disrupt the two major pathways controlling the cell cycle — those involving the tumour suppressors p53 and pRb (the retinoblastoma protein). Experimental evidence supporting a causal role has come from reports that the HPV genome is found localised to only the primary tumour cells and their metastases, at similar copy numbers to those reported in other HPV-associated malignancies, including those of the cervix.9,10 Demonstration (by reverse transcription/polymerase chain reaction assays) that E6 and E7 are active in the cancer cell has also been of key importance in establishing an aetiological role.7 Integration of HPV DNA into the host chromosome, which frequently heralds malignant conversion in cervical lesions, has also been observed, although there are indications that the proportion of head and neck cancers carrying the virus in non-integrated form in the cell nucleus may be higher than in the cervix.10 Seroreactivity against the viral capsid proteins and oncogenes has also been associated with an increased risk of cancer.8 Molecular analyses of p53 and of cellular proteins in the pRb pathway (including cyclin D1, a positive regulator, and p16, a negative regulator of the cell cycle) have suggested that HPV-positive head and neck cancers are a biologically distinct group. HPV-positive cancers are generally characterised by loss of expression of pRb and cyclin D1 and overexpression of p16 following inactivation of pRb by the viral E7 oncoprotein.6,11 In contrast, HPV-negative tumours consistently show overexpression of pRb and cyclin D1 and loss of p16. Mutations in the p53 gene appear to be much more common in HPV-negative cancers than in HPV-positive cancers. In HPV-positive cancers, p53 inactivation is achieved by interaction with the viral oncoprotein E6. Nonetheless, HPV and p53 mutations appear to coexist in at least a proportion of head and neck cancers. The recent IARC study,8 like other, smaller surveys, shows that HPV-positive head and neck cancers are more common in patients who do not smoke or chew tobacco,8,12 but HPV infection and tobacco consumption together are associated with an increased risk (whether the effect is additive or synergistic is not known). Several studies have reported a clustering of HPV-positive cancers in younger age groups,13 but the association with age remains controversial. HPV has statistical associations with a history of multiple sexual partners, the practice of giving oral sex and a history of genital warts, suggesting sexual transmission, although conclusive data are lacking.8,12 A lower rate of oral exposure to HPV because of cultural or social attitudes relating to sexual practices (notably orogenital contact) may help explain the lack of HPV DNA in a recent small series of tonsillar cancers from Chinese patients.14 What are the practical implications of an aetiological role for HPV in head and neck cancer in the clinical setting? The prognosis for head and neck cancer has not improved appreciably over the past two decades, despite a better understanding of the biology of invasive tumours, advances in surgery and the increased use of combined-modality treatment. Clinical staging is the primary guide to treatment modality, but provides a limited guide to outcome. There have been general expectations that genetic analysis is “the way of the future” in terms of identifying markers for prevention, therapy and prognosis. Attempts to find prognostic cellular markers in head and neck cancers have been disappointing, but including HPV status as a variable in the data analysis may yield more rewarding results. There are real prospects that HPV itself may provide a new generation of prognostic markers. Surveys have consistently linked HPV-positive cancers with a more favourable clinical outcome than HPV-negative cancers. The risk of recurrence or death from the disease associated with HPV status is independent of established prognostic factors, including tumour stage and lymph-node metastasis.6,10,11 The apparently better prognosis for patients with HPV-positive cancers may reflect the inability of the viral E6 protein to fully inactivate p53 and/or the adverse effect of radiation and other toxic agents on the ability of E6 to interfere with p53 and other host proteins, so that the cancers are more susceptible to therapy. Screening of head and neck cancers for HPV 16 and/or raised p16 as a marker of HPV-related transformation (Box) may identify a new group of patients (those with HPV-negative tumours) who would benefit from closer-than-usual follow-up or additional therapy. However, large-scale prospective studies will be needed to establish the cost-effectiveness of this approach. Internationally, HPV DNA testing is gaining acceptance as an adjunct to cytology in cervical cancer screening programs. Early diagnosis is still the best predictor of good outcome in head and neck cancer. The possible public health benefits of cytological screening and/or detection of HPV DNA in cells collected from the oral cavity by dentists or general practitioners await evaluation. Such studies will need to include details of how and when specimens are taken and cost–benefit analyses. Both in Australia and overseas, clinical trials are under way to test HPV vaccines (developed from viral capsids generated in the laboratory) for prevention of cervical cancer. The demonstrated efficacy of an HPV 16 vaccine in preventing persistent infection in the female genital tract15 has raised expectations that such vaccines may eventually reduce the incidence of head and neck cancer. However, the potential of vaccine programs designed for the genital tract to reduce the incidence of HPV infection in the head and neck is still to be evaluated. Specific randomised controlled trials to address the prevention of persistent oral HPV infections may need to be considered. Therapeutic vaccines based on the viral oncogenes E6 and E7 remain at the experimental stage, but may eventually prove a valuable addition to surgical, radiation and chemotherapeutic approaches to managing advanced disease. Upregulation of p16 as a marker of HPV-mediated transformation in tonsillar cancer cells Strong, diffuse staining for p16 is shown in a high proportion of tumour cells (arrow). A close-up view of tumour cells (inset, x 400) shows staining throughout the nucleus and cytoplasm. Glossary Viral oncogenes. DNA sequences in the viral genome that code for a protein capable of setting the cell on a course to malignant transformation. Tumour suppressors. Classes of proteins that protect the cell against malignant transformation (eg, by monitoring the integrity of the cell or by preventing unscheduled cell cycling).

Barbara R Rose PhD, MASM, BSc · Christopher J O’Brien FRACS · Wei Li MMed, MM

Medical research in the UK is gripped by an anti-intellectual winter During the last British winter, a moderately well-off elderly couple in London somehow failed to pay their gas bill. After the routine warnings, the gas company turned off their supply. Interpreting the Data Protection Act of 1998 as prohibiting the divulgence of personal information to a third party without consent, the gas company failed to notify the local social services department that its former customers were now at risk. Eventually, one of the pensioners died quietly of hypothermia and the other of a cardiac event. Every piece of research . . . requires approval using a nationally standardised ethics form that potentially runs to 73 pages This true story highlights a national community that collectively has lost sight of reasonable balances between risk and benefit in regard to information and, by extension, the same equation as it applies to the systematic collection of data as it occurs in academic research. Somehow, the paediatric cardiac surgery scandal at Bristol Royal Infirmary and the murders committed by Shipman have, in the minds of tabloid editors and health bureaucrats, become conflated with the Alder Hey matter. The latter involved paediatric pathologists retaining tissue samples and body parts of deceased children “for research”, without obtaining the consent of their parents. A further ingredient is several infamous cases (mostly outside the United Kingdom) of research investigators being caught fabricating data. Throw in a typically prescriptive directive from the European Union regarding clinical trials of new drugs, and the results are a burgeoning “industry” regulating research, and paperwork that is multiplying faster than maggots in a dunghill. The UK Government has only added to the momentum. Its Research Management and Governance Framework obliges every National Health Service Trust, be it in primary care, a general hospital or the mental health services, to develop systems to ensure that it is aware of all research taking place on its “patch”, and that each project meets all relevant standards of ethical, intellectual, scientific and fiscal propriety. At first glance, this appears entirely reasonable, but the reality is that investigators now have to prepare and submit multiple copies not only of grant applications, and then ethics applications, but also of new forms for research governance. Every piece of research involving users or staff of health and social care services, or the data such organisations hold, requires approval using a nationally standardised ethics form that potentially runs to 73 pages. This also applies to research projects undertaken in these settings by students, including undergraduates. The disincentives to initiating research are clearly massive, and the burdens are becoming unsustainable, especially those on supervisors of trainees who need skills and experience in doing and interpreting research to practise their professions properly. In primary care, in particular, it may well be impossible for the responsible Trust to know which of its dozens of affiliated general practitioners are trying a new drug on a few patients at the behest of some larger or smaller pharmaceutical company, perhaps based at the other end of the country or even overseas. But if something goes wrong, it is the CEO of that Trust whose neck will be on the block. Should these individuals be lying awake worrying about that possibility? If we can disentangle Bristol and Shipman from research, because hopefully both were exceptional episodes of exceptionally bad clinical practice, then the truth is probably that the Alder Heys of this world are also rare and are likely becoming progressively rarer. Attitudes do change, and the kind of paternalism represented by the actions of the Alder Hey pathologists is generally on the wane. Indeed, several Australian jurisdictions revised their coroners acts about a decade before Alder Hey “broke”, because of a rising tide of concern that bodies had been released to grieving relatives with no indication that significant elements or even complete organs had been retained. The transition was not without pain, and pathology staff were obliged to brush up their skills for dealing with the living, but we fairly quickly reached a new level of respect for all of the parties in the forensic triangle. Certainly, Australian researchers were spared the chore faced by all British medical academics post-Alder Hey of having to account for every last histology slide up and down the country. In a highly competitive, publish-or-perish environment, the temptation to bend or break the rules of scientific and intellectual integrity may never be eliminated. And peer review is demonstrably not a foolproof defence against fraudsters. While instances of ingenious invention of interesting findings clearly do come along more frequently than transits of Venus, and therefore are more common than collisions of asteroids with the Earth, any scientifically literate member of an ethics committee will tell you that we face far greater problems in the shape of poorly designed and ethically and scientifically inadequate research protocols. At the very least, it takes the average medical student around 4 years to lose completely the capacity to describe any piece of research in terms that a potential lay participant can understand. Added to this is the problem posed by powerful computers that readily reanalyse any given set of data until at least one “statistically significant” (and therefore publishable) result emerges. Perhaps worst of all is the tendency for authoritative individuals and organisations to recommend clinical and public health policies that go well beyond the available evidence. In the UK, for example, the major diabetes charity has run a “missing million” campaign, advocating screening the population for undiagnosed diabetes. And all because rigorous controlled trials have shown that tight control of blood glucose level and blood pressure in people with clinically evident diabetes reduces vascular and renal complications.1,2 Research and its reputation are not without their problems, but breaches of the Data Protection Act, pathological paternalism, and the imminence of a cosmic collision are not chief among them. Greater clarity in thinking and improved scientific and ethical literacy in the population at large, in the community of investigators, and among policymakers and politicians would shorten the anti-intellectual winter now gripping much of medical research in the UK.

Konrad Jamrozik DPhil, FAFPHM, MFPH · Richard F Heller MD, FRCP, FRACP, FAFPHM · David P Weller MPH, PhD, FRACGP, FAFPHM

Objectives: To analyse temporal trends and geographical variations in testosterone prescribing in Australia.Design and setting: An analysis of testosterone prescribing over the past 11 years according to products and region, determined by Pharmaceutical Benefits Scheme (PBS) expenditure in Australian states and territories.Main outcome measure: Patterns of monthly PBS expenditure on injectable, oral and implantable testosterone products from 1 January 1991 to 30 December 2001, classified by state or territory.Results: There were two periods (1993–1994 and 1998–1999) of striking upsurge followed by declines in national total prescribing of testosterone. These changes were more prominent for oral than injectable testosterone products, and patterns were similar in all regions, apart from a disproportionately higher peak in Western Australia in 1998. On a per-capita basis, Western Australia showed a dramatic increase in prescribing of oral and implantable, but not injectable, testosterone coinciding with the opening of a franchised men’s sexual health clinic in Perth.Conclusion: The two striking upsurges in testosterone prescribing despite no convincing new evidence to justify them appear to reflect promotional activity to prescribe testosterone for older men, rather than overcoming the underdiagnosis of androgen deficiency related to pituitary or testicular disease in younger men. The curtailments after the introduced restrictions to PBS prescribing for older men without overt androgen deficiency were partial and temporary, suggesting that such regulatory barriers are only partly successful in counteracting the commercial and populist pressure driving excessive testosterone prescribing. Professional and community education is needed for appropriate diagnosis of genuine androgen deficiency in younger men, while discouraging unproven testosterone treatment for ageing men.

David J Handelsman FRACP, PhD

Objectives: To determine the response to colorectal cancer (CRC) screening by colonoscopy, through direct invitation or through invitation by general practitioners.Design and setting: Two-way comparison of randomised population sampling versus cluster sampling of a representative general practice population in the Australian Capital Territory, May 2002 to January 2004.Intervention: Invitation to screen, assessment for eligibility, interview, and colonoscopy.Subjects: 881 subjects aged 55–74 years were invited to screen: 520 from the electoral roll (ER) sample and 361 from the general practice (GP) cluster sample.Main outcome measures: Response rate, participation rate, and rate of adenomatous polyps in the screened group.Results: Participation was similar in the ER arm (35.1%; 95% CI, 30.2%–40.3%) and the GP arm (40.1%; 95% CI, 29.2%–51.0%) after correcting for ineligibility, which was higher in the ER arm. Superior eligibility in the GP arm was offset by the labour of manual record review. Response rates after two invitations were similar for the two groups (ER arm: 78.8%; 95% CI, 75.1%–82.1%; GP arm: 81.7%; 95% CI, 73.8%–89.6%). Overall, 53.4% ineligibility arose from having a colonoscopy in the past 10 years (ER arm, 98/178; GP arm, 42/84). Of 231 colonoscopies performed, 229 were complete, with 32% of subjects screened having adenomatous polyps.Conclusions: Colonoscopy-based CRC screening yields similar response and participation rates with either random population sampling or general practice cluster sampling, with population sampling through the electoral roll providing greater ease of recruitment.

Mike Corbett FRACP · Sharon L Chambers RN, GradDipNurs · Bruce Shadbolt PhD · Doug Taupin PhD · Lybus C Hillman MD, FRACP

Objectives: To determine whether the quality of hospital inpatient care can be improved by using checklists and reminders in clinical pathways.Design: Comparison of key indicators before and after the introduction of clinical pathways incorporating daily checklists and reminders of best practice integrated into patient medical records.Setting and participants: The study, at Wimmera Base Hospital in Horsham, Victoria, included patients admitted between 1 January 1999 and 31 December 2002 with ST-elevation acute myocardial infarction (AMI) and patients admitted between 31 July 1999 and 31 December 2002 with stroke.Main outcome measures: Compliance with key process measures determined as best practice for each clinical pathway.Results: 116 patients with AMI and 123 patients with stroke were included in the study. ST-elevation AMI. After introducing the clinical pathway program, percentage-point increases for treatment compliance were 21.4% (95% CI, 7.3%–32.7%) for patients receiving aspirin in the emergency department; 42.7% (95% CI, 26.3%–59.0%) for eligible patients receiving β-blockers within 24 h of admission; 48.1% (95% CI, 31.4%–64.8%) for eligible patients being prescribed β-blockers on discharge; 43.7% (95% CI, 28.4%–59.1%) for patients having fasting lipid levels measured; and 41.2% (95% CI, 19.0%–63.5%) for eligible patients having lipid therapy. Stroke. After introducing the clinical pathway program, percentage-point increases for treatment compliance were 40.7% (95% CI, 21.0%–60.2%) for dysphagia screening within 24 h of admission; 55.4% (95% CI, 32.9%–77.9%) for patients with ischaemic stroke receiving aspirin or clopidogrel within 24 h of admission; and 52.4% (95% CI, 33.8%–70.9%) for patients having regular neurological observations during the first 48 h after a stroke. There was a fall of 1.0 percentage point (ie, a difference of –1% [95% CI, –4.7% to 10.0%]) in the proportion of patients having a computed tomography brain scan within 24 h of admission.Conclusion: Significant improvements in the quality of patient care can be achieved by incorporating checklists and reminders into clinical pathways.

Alan M Wolff FRACGP, MBA · Sally A Taylor BN, RM, MAM (Health) · Janette F McCabe BN, CritCareCert, MAM (Health)

Melagatran is a synthetic, small-peptide direct thrombin inhibitor with anticoagulant activity. Ximelagatran, an oral prodrug, undergoes rapid enzymatic conversion to melagatran. Melagatran has rapid onset of action, fixed twice-daily dosing, stable absorption, apparent low potential for medication interactions, and no requirement for monitoring drug levels or dose adjustment. There is no specific antidote, but the drug has a short plasma elimination half-life (about 4 hours). In clinical studies, melagatran/ximelagatran is not inferior to warfarin for stroke prevention in patients with non-valvular atrial fibrillation, to heparin–warfarin for acute treatment and extended secondary prevention of deep vein thrombosis, and superior to warfarin for prevention of venous thromboembolism after major orthopaedic surgery. Major bleeding with melagatran/ximelagatran occurred at rates similar to those in patients treated with warfarin. 6%–12% of patients taking ximelagatran develop asymptomatic elevated liver enzyme levels (predominantly alanine aminotransferase) after 1–6 months of therapy; this usually resolves with cessation of therapy. Less than 1% of patients develop abnormal liver function while taking ximelagatran; this rarely persists or develops into clinical illness.

Timothy A Brighton FRACP, FRCPA, MD

Australia is now one of the safest countries in the world in which to be born. This is largely a result of the many advances in obstetric and neonatal medicine of the past 50 years. However, the “medicalisation” of birth has tended to diminish women’s satisfaction with their experience of childbirth. It has been shown that women are most satisfied by care from a single practitioner, and when they themselves have input into decision-making. Although maternal satisfaction is important, it should not be promoted at the expense of the health of mothers and babies. More realistic antenatal education and preparation should be available for all pregnant women so that both maternal satisfaction and good perinatal outcomes can be achieved.

Caroline M de Costa FRANZCOG, FRCOG · Stephen Robson MM, FRANZCOG, MRCOG

Perhaps a week of intensive training in critical thinking would be the best preparation for members of research ethics committees In a recent lecture at Monash University, the philosopher Raimond Gaita, Professor of Moral Philosophy at King’s College, University of London, and Professor of Philosophy at the Australian Catholic University, told the story of a woman facing a significant turning point in her life. It was a Friday, and a decision was needed by Monday, but she had unavoidable obligations over the weekend. She had a dear friend, a psychoanalyst and philosopher who had known her all her life. He knew her circumstances, her preferences and even her secret wishes. She contacted him and prevailed upon him to make the decision for her. In our private lives most of us would find it at least odd, and probably uncomfortable, to hand over responsibility for significant decisions to others. Yet, the prevailing paradigm for human research ethics committees has institutionalised this approach. Researchers themselves often do not consider the ethical implications of their work until it is time to fill out the various forms required by committees. Even then, the main concern is “getting through ethics” with minimal scarring of their proposal. The Nuremberg Code,1 the Helsinki Declaration,2 and even the National Health and Medical Research Council’s National statement on ethical conduct in research involving humans (the Statement),3 are not documents with which many researchers can claim significant familiarity. The reasons for their existence are faintly recalled, and current debates are only of interest if they impede research with which the researcher has a personal concern. Once an ethics committee has made its decision, there is no need to consider “ethics” again unless there is a significant adverse event. Although they undoubtedly provide a “safety net” to detect and prevent grossly unethical research, ethics committees must not and cannot be seen as the repositories for moral decision-making. Consider an imaginary (but highly plausible) ethics committee. It meets the Statement’s requirements for membership. Some members have attended the occasional seminar sponsored by the Australian Health Ethics Committee (AHEC), and some diligently read the AHEC Bulletin sent to registered committees. Some, though not all, of the members have actually read the Statement all the way through. One committee member doesn’t really agree with some of the content. The committee faces regular criticism by researchers for the amount of paperwork that must be submitted to it, and significant anger when it wishes to alter an aspect of a proposal for a multicentre trial. Although the committee’s deliberations are thorough, most of its recommendations consist of minor changes to the plain language and consent statement. It faces considerable (and understandable) pressure to reach rapid consensus. Rarely does a member ever register his or her dissent concerning a decision about which all other members of the committee feel comfortable. The committee is proud that it has never ultimately rejected any proposal. Some members of the committee are aware that they have acquiesced in decisions about which they had some misgivings. One or two of the most senior members know they can nearly always sway the committee to their point of view. In a recent editorial discussing clinical ethics committees, Margaret Somerville noted that: Committee decisions, as compared with individual ones, can spread the responsibility. A committee can make a decision that no one person — in particular, no committee member — acting alone would make.4 She uses the real-life example of decisions to shorten life by withholding treatment, or aborting a fetus, and the physicians doing this being morally reassured by the involvement of an Acute Clinical Ethics Service. She asks Might [this involvement] have allowed the caring team to implement decisions that their moral intuitions were indicating were unethical? While these decisions may have been ethical, we must always be aware that we ignore such intuitions at our ethical peril.4 Although clinical ethics committees perform a somewhat different function to human research ethics committees, there are significant similarities as far as the points made by Somerville are concerned. There is indeed comfort in allowing ourselves to be relieved of having to think about the implications of our actions, especially when the research dollar is concerned. It is problematic when the ethics review process is seen as a test of how much we are able to get away with. It cannot be persuasively argued that actions must be ethical because they have been approved by another person thought to be morally wiser. Nor is it helpful to allow oneself to be influenced to reach a decision in a group setting because of reasons of time, or because others have already reached consensus. If researchers have not thought through the ethical implications of their proposals, but instead leave that to the committee, and the committee makes decisions about which some of its members would be individually uncomfortable, this cannot be regarded as a satisfactory process. Perhaps the most essential preparation for members of research ethics committees is not studying the content of the Statement or the relevant law, but undertaking a week of intensive training in critical thinking. Perhaps we all must consider how best to deal with situations about which not all agree, and about which objections are morally relevant. Furthermore, there are many issues that are not well addressed by guidelines or law. What research should be done in the first place? How should communities from which participants are drawn be involved in the planning, implementation, monitoring and evaluation of research? What are the human rights implications of a study (particularly in populations significantly deprived of rights)? What responsibilities do researchers have to the larger community from which their subjects are drawn, and what do they owe to subjects after their research is completed? These questions and many more have important ethical dimensions. Many researchers are unaccustomed to thinking through the broader implications of their work. However, they are capable of doing what is necessary in order to fill out a form. In a seminal article in the New England Journal of Medicine, Henry Beecher stated The ethical approach to experimentation in man has several components; two are more important than the others, the first being informed consent . . . Secondly there is the more reliable safeguard provided by the presence of an intelligent, informed, conscientious, compassionate, responsible investigator.5 In concluding his lecture, Gaita rejected the notion of a “moral expert”, and called for us all to identify and rigorously analyse morally important issues without sentimentality. It is very difficult to improve on this.

Bebe Loff MA, LLB · Jim Black MB BS, PhD, FAFPHM

The Victorian Supreme Court has decided that artificial nutrition and hydration provided through a percutaneous gastrostomy tube to a woman in a persistent vegetative state may be withdrawn. The judge ruled, in line with a substantial body of international medical, ethical and legal opinion, that any form of artificial nutrition and hydration is a medical procedure, not part of palliative care, and that it is a procedure to sustain life, not to manage the dying process. Thus, the law does not impose a rigid obligation to administer artificial nutrition or hydration to people who are dying, without due regard to their clinical condition. The definition of key terms such as “medical treatment”, “palliative care”, and “reasonable provision of food and water” in this case will serve as guidance for end-of-life decisions in other states and territories. The case also reiterates the right of patients, and, when incompetent, their validly appointed agents or guardians, to refuse medical treatment. Where an incompetent patient has not executed a binding advance directive and no agent or guardian has been appointed, physicians, in consultation with the family, may decide to withdraw medical treatment, including artificial nutrition or hydration, on the basis that continuation of treatment is inappropriate and not in the patient’s best interests. However, Victoria and other jurisdictions would benefit from clarification of this area of the law.

Michael A Ashby MD, FRACP · Danuta Mendelson MA, LLM, PhD

Although cutaneous leishmaniasis is occasionally seen in Australia in overseas travellers and migrants, visceral leishmaniasis has been reported rarely and only in people who were immunocompromised. We describe an 18-year-old immunocompetent man who presented with pancytopenia and a 2-week history of fever and lethargy a year after visiting the Greek Islands. Visceral leishmaniasis was diagnosed after a bone marrow biopsy showed protozoa, and the patient responded well to treatment with liposomal amphotericin. To our knowledge, this is the first case of visceral leishmaniasis in an immunocompetent patient in Australia. The leishmaniases are a group of infections caused by protozoa belonging to the genus Leishmania. We report an immunocompetent patient with visceral leishmaniasis, probably acquired in Greece. As far as we are aware, this is the first report of visceral leishmaniasis in an immunocompetent person in Australia. Clinical recordAn 18-year-old man of Greek parentage presented with a 2-week history of dry cough, diarrhoea, lethargy, anorexia and fever. He had a past history of cutaneous vasculitis of uncertain aetiology and acne vulgaris. He was not taking any regular medication. On examination, the patient’s temperature was 37.5°C, but no other abnormalities were evident. Investigations (Box 1) showed that he had pancytopenia, raised erythrocyte sedimentation rate and hypergammaglobulinaemia. The raised rheumatoid factor and anti-dsDNA levels were thought to be most likely related to inflammation. Treatment was begun with intravenous ceftazidime and gentamicin because of the febrile neutropenia. A bone marrow biopsy performed on Day 2 to investigate the cause of the pancytopenia showed cellular bone marrow with relative erythroid hyperplasia. Multiple blood cultures gave negative results. As the fever persisted, roxithromycin was added to the regimen on Day 4, and vancomycin on Day 8. Serological testing was negative for multiple infections, including HIV infection, but positive for parvovirus B19 IgM, although negative for IgG, on Days 1 and 9. The provisional diagnosis was parvovirus-related pancytopenia. The patient was discharged without antibiotic therapy 12 days after admission, as he felt well, although he still had a fever. A month after discharge, the pancytopenia persisted. Repeated parvovirus B19 serological testing was negative for both IgM and IgG. A second bone marrow biopsy at this time revealed occasional macrophages containing amastigotes (the resting intracellular stage of leishmaniae, formerly known as Leishman–Donovan bodies; Box 2). Review of the first bone marrow biopsy specimen failed to show any protozoa. Further questioning revealed that the patient had visited Greece for a 5-week holiday a year before presentation. A diagnosis was made of visceral leishmaniasis, probably acquired through sandfly bites. The patient was treated with liposomal amphotericin B, 250 mg daily for 4 consecutive days and then weekly for 2 weeks. The diagnosis was confirmed when promastigotes (the motile form of the organism) were successfully cultured. Leishmaniasis is not notifiable in Australia. The patient’s blood count returned to normal 2 months after starting treatment with amphotericin. He remained well 2 years after ceasing treatment. DiscussionThis case illustrates an unusual cause of pancytopenia in Australia. It is important to consider leishmaniasis as a differential diagnosis when the patient has a relevant travel history, as in this case. The diagnosis was not made on the first bone marrow biopsy, as smears from bone marrow usually contain few organisms, and have a sensitivity of 80%–85%.1 The major clinical syndromes caused by the genus Leishmania are cutaneous, mucosal and visceral leishmaniasis. Leishmaniasis is transmitted by Phlebotomus or Lutzomyia sandflies and infects dogs and foxes as well as humans. Numerous cases of cutaneous leishmaniasis have been reported in Australia,2,3 but to our knowledge only three cases of visceral leishmaniasis — one in a renal transplant patient4 and two in patients with HIV infection.5 Visceral leishmaniasis occurs in the Mediterranean region, northern Africa, the Middle East and central Asia, China and parts of South America.6 The differentiation of species causing leishmaniasis is complex and is performed only in reference laboratories. Visceral leishmaniasis is caused by Leishmania donovani, L. infantum or L. chagasi. L. infantum is the common species in the Mediterranean littoral and probably the cause of our patient’s infection. The incubation period of visceral leishmaniasis is usually 2–8 months. Many infections are subclinical, but the classic presentation is with fever, weight loss, hepatosplenomegaly, pancytopenia and hypergammaglobulinaemia. The definitive diagnosis depends on demonstrating either amastigotes in tissue or promastigotes in culture. Splenic puncture is the most sensitive means of obtaining a diagnosis, but biopsy of the bone marrow and liver is almost as good. In immunosuppressed individuals, promastigotes can sometimes be cultured from the buffy layer of centrifuged blood, and amastigotes can be seen in macrophages in biopsy specimens from various tissues. Testing for antileishmanial antibodies is not generally available and may be difficult to interpret. Pentavalent antimony compounds have been used to treat leishmaniasis for decades, but often have severe side effects, and resistance is developing.7 The aminoglycoside aminosidine (also known as paromomycin) is also effective when used either alone or in combination with an antimonial compound.7 Amphotericin can achieve 98% long-term cure in both antimonial-unresponsive and previously untreated patients.7 Lipid formulations of amphotericin are the most active antileishmanial agents.8 Although leishmaniasis is rare in Australia, it needs to be considered as a possible differential diagnosis of pancytopenia, remembering that the disease can present up to several years after return from an affected area. If untreated, visceral leishmaniasis is potentially fatal. 1 Investigations in a patient with visceral leishmaniasis Investigation Result Reference range Haemoglobin level (g/L) 82 135–175 Total white cell count (cells/L) 2.2 × 109 4–11 × 109 Neutrophil count (cells/L) 0.8 × 109 2.0–7.5 × 109 Platelet count (cells/L) 106 × 109 150–400 × 109 Erythrocyte sedimentation rate (mm/h) 120 0–15 Antinuclear antibody titre 1/160 < 1/160 Anti-dsDNA level (IU/mL) 10 < 8 Rheumatoid factor level (kIU/L) 240 < 20 IgG level (g/L) 19.1 6.5–16.0 IgA level (g/L) 4.3 0.6–4.0 2 Bone marrow biopsy in visceral leishmaniasis Light micrograph of bone marrow, showing macrophages containing amastigotes characteristic of Leishmania spp. Amastigotes are the non-flagellated intracellular stage of the protozoan, formerly known as Leishman–Donovan bodies. They appear as spherical or oval organisms 2–6 μm in length that contain two darker-staining organelles — a nucleus and a kinetoplast (arrow). The latter is a distinct region of the mitochondria containing mitochondrial DNA. (Giemsa stain; scale bar = 10 μm.) Transmission electron micrograph of bone marrow, showing an amastigote with a kinetoplast characteristic of Leishmania spp. (arrow). (Scale bar = 1 μm.)

Oui Ju MB BS · David I Grove MD, FRACP, FRCPA · Wilfrid J Jaksic FRACP · Geoffrey W Dart FRACP, FRCPA

QuestionIn patients without diabetes, but with impaired renal function and proteinuria, does combination treatment with the angiotensin-converting enzyme (ACE) inhibitor trandolapril and the angiotensin-receptor blocker (ARB) losartan prevent deterioration of renal function more effectively than treatment with either agent used alone? Trial details Design: Randomised controlled trial. Setting: Outpatient clinics in Kisarazu, Kimitsu and Futtsu (Japan). Participants: 263 patients (mean age about 45 years; 54% men) with non-diabetic biopsy-proven renal disease, impaired renal function (creatinine clearance rate, 20–70 mL/min), persistent proteinuria > 0.3 g/24 h, and hypertension. Patients with proteinuria > 10 g/24 h, renovascular or malignant hypertension, urinary tract infection, heart failure or myocardial infarction, connective tissue disease, chronic hepatic or pulmonary disease, cancer and those pregnant or breastfeeding were excluded. Duration: 36 months. Main outcome measures: Combined endpoint of end-stage kidney disease (ESRD) or doubling of serum creatinine level, death and proteinuria. Main results: At 3 years’ follow-up, 10 (11%) of 85 patients receiving combination treatment reached the combined primary endpoint, compared with 20 (23%) of 85 taking trandolapril alone (hazard ratio, 0.38; 95% CI, 0.18–0.63) and 20 (23%) of 86 taking losartan alone (hazard ratio, 0.40; 95% CI, 0.17–0.69). Proteinuria decreased significantly in all treatment groups, but the effect was greatest with combination teatment; the maximum median change in daily urinary protein excretion was –42.1% with losartan, –44.3% with trandolapril and –75.6% with the combination treatment (P = 0.01). There was no significant difference in the mean fall in systolic blood pressure (losartan, – 5.1 mmHg [standard deviation, 1.6]; trandolapril, – 5.2 mmHg [SD, 1.3] mmHg; combination, – 5.3 mmHg [SD, 1.4]) and diastolic blood pressure (losartan, – 2.9 mmHg [SD, 0.9]; trandolapril, 22.9 mmHg [SD, 0.8]; combination, 23.0 mmHg [SD, 0.7]) among the three groups. Conclusion: Combining ACE inhibitors and ARBs reduces the incidence of the combined end-point of ESRD or doubling of serum creatinine in non-diabetic chronic kidney disease, with moderate reduction of renal function and moderate proteinuria by at least 30% compared with treatment with either drug alone. CommentaryRationale for the trialThe renin–angiotensin system has been implicated in the progression of non-diabetic renal disease, and ACE inhibitors and ARBs have independently been shown in randomised controlled trials to reduce proteinuria and the risk of end-stage kidney disease (ESRD) and doubling of the serum creatinine level. Proteinuria lies in the causal pathway to ESRD; therefore, the antiproteinuric effect of these agents is a component of renoprotection.1-6 What needed further exploration was whether the combination of ACE inhibitor and ARB was better than treatment with either agent alone. The COOPERATE triallists postulated an advantage of complete inhibition of the renin–angiotensin system with combination treatment of ACE inhibitor and ARB at maximum dosage.7 Trial methodsAfter an 18-week active run-in period, participants were randomly allocated to ACE inhibitor (trandolapril, 1–3 mg/day plus a lactose placebo), ARB (losartan, 12.5–100 mg/day plus a lactose placebo) or ACE inhibitor (trandolapril, 0.5–3 mg/day) plus ARB (losartan, 12.5–100 mg/day). Run-in was performed to establish safety, adherence and the maximum tolerable dose of ACE inhibitor. There was no run-in period for the ARB, and the maximum tolerable dose was chosen on the basis of a previous study.8 Information essential for assessing the quality of the trial was missing from the methods section. The triallists detailed their randomisation technique, but the methods by which the randomised intervention was allocated to the patients — such as use of sequentially labelled, sealed, opaque envelopes, or a central or pharmacy randomisation, which would guarantee concealment — was not specified. Patients and investigators were blinded by the drugs being dispensed in identical containers, but there was no mention of whether the capsules themselves were similar. There was also no indication of blinding of the outcome assessors. Analysis was reported as intention-to-treat, but six of 263 patients (who could have had outcomes measured) were actually excluded from the trial because of “protocol invalidation” or “discontinuation”, and so analysis was “per-protocol” and not intention-to-treat. Overall, seven of 301 (2.3%) patients were lost to follow-up, a small number which was unlikely to cause any differences in the final results of the trial. The authors performed a subgroup analysis to assess whether the effect of combination treatment compared with single therapy with ACE inhibitor or ARB was affected by the degree of baseline proteinuria. Their methods for this subgroup analysis and reporting of results were incorrect and misleading, although common. They analysed the three groups separately, and showed no statistically significant benefit of combination treatment in patients with low grade (< 1 g/24 h) proteinuria (hazard ratio, 0.69; 95% CI, 0.22–2.28), and statistically significant benefit in patients with moderate (1–3 g/24 h) proteinuria (hazard ratio, 0.33; 95% CI, 0.19–0.74) and heavy (> 3 g/24 h) proteinuria (hazard ratio, 0.40; 95% CI, 0.21–0.84). What they should have done was a formal test of interaction (a statistical assessment of differences in the proportion of an outcome [eg, ESRD] in the three treatment groups across patients with low-grade, moderate or heavy proteinuria).9 Given the overlapping confidence intervals in their separate analyses, it is very likely that the test of interaction would have been negative. The absence of demonstrable effect in the lowest-risk stratum is explainable on the basis of lower event rates in this group, and therefore lack of power to detect an effect in this group separately. Urinary protein excretion on a continuous scale in grams per 24 hours (rather than groups) could also have been used as the explanatory variable. This study may also have been relatively underpowered to detect the harms of treatments because of the relatively small sample size (263) and because of the active run-in phase, which would tend to exclude patients at risk of adverse effects from the intervention. Run-in periods are common and very useful for improving the efficiency of trials by selecting patients most likely to comply with treatment and fulfil follow-up requirements. However, if there is an active run-in when patients with adverse effects are not randomised, there may be a systematic underestimation of the true harms of an intervention. Also, only the dose of ACE inhibitor, but not ARB, was established by an active run-in, and the dose of ARB in the combination treatment was not reported clearly. The potential variability of doses makes it possible that outcome differences are related to dose rather than greater efficacy. However, blood pressure — probably the most important confounder — was equalised across the two groups. Finally, in this study, no account was taken of how the effect of ACE inhibitors, ARB, or their combination, may be influenced by histological types of renal disease with different rates of progression. New informationIt is likely that there is an advantage of complete inhibition of the renin–angiotensin system with combined ACE inhibitor and ARB treatment compared with therapy with either drug alone, at maximum doses, in patients without diabetes, but with moderate (1–3 g/24 h) and heavy (> 3 g/24 h) proteinuria and moderate to severe renal impairment (creatinine clearance rate, 20–70 mL/min). These findings of the COOPERATE trial are promising, but need to be further explored in light of some of the study’s limitations, and the fact that comparative findings in patients with cardiovascular disease are conflicting. In chronic heart failure, combined treatment has proven to be advantageous compared with individual treatment with ACE inhibitors or ARBs.10,11 This has not been confirmed in studies of post-myocardial infarction.12 Implications for clinical practiceIn patients without diabetes who have renal impairment (creatinine clearance rate, 20–70 mL/min; proteinuria, > 0.3 g/24 h), combination treatment with ACE inhibitors and ARBs at their maximum dose appears to be effective and well tolerated. Further studies are awaited to confirm the findings of the COOPERATE trial. In the meantime, combination therapy may prove a useful strategy in patients who continue to have high-grade proteinuria despite maximal doses of ACE inhibitor or ARB.

Giovanni FM Strippoli MD, MPH, MM(Epi) · Craig C Jonathan MB ChB, PhD

Clinical trials typically require the collection of many data to describe the participants and for measuring their response to an intervention. In addition to the primary analysis of treatment effect, investigators can use these data to perform multiple analyses, but there are important pitfalls with their use.1,2 Here, we discuss three common types of secondary analyses: analyses of multiple outcome variables; analyses of trial outcomes that account for prognostic factors (adjusted analyses); and using trial data to answer secondary research questions (see definitions in Box 1). The use of trial data for population subgroup analyses has been discussed earlier in this series.3,4 What are the problems?The two main problems introduced by multiple analyses are, firstly, the increased probability of detecting intervention effects where none exist (“false positives” owing to multiple comparisons — type I errors), and secondly, the limited capability (“power”) of trials to detect a true treatment effect in secondary outcomes if not enough participants are enrolled to show a statistically significant difference in these outcomes (“false negatives” — type II errors). One study compared trial protocols with their subsequent publications, and provided empirical evidence of the selective reporting of positive trial results.5 The use of multiple analyses is therefore of particular concern when these are conducted post-hoc as a “fishing expedition”, and undue emphasis is given to positive findings. Item 18 of the CONSORT checklist (Box 2) recommends that investigators report on all multiple analyses and declare which were prespecified and which were conducted as exploratory activities after investigators were unblinded to the treatment allocation of participants.6 Analyses of multiple outcomesAdvantagesInvestigators may choose multiple outcomes to measure the effect of treatment. This is an advantage when different parameters provide information about different aspects of the treatment response.1 Secondary analyses may also assist the interpretation of the primary analysis. For example, in a recently reported trial comparing chemotherapy regimens in the treatment of patients with metastatic breast cancer, investigators selected two primary outcomes as the most important measures of treatment effect: the overall tumour response rate (measured as complete and partial response) and time to treatment failure. Secondary outcomes, including overall survival, toxicity and quality of life, provided additional information about the treatment effect.7 Including a set of supplementary outcomes may also be a practical solution when different investigators value outcomes differently. Multiple-outcomes analysis is particularly useful when a statistically significant benefit of treatment on the primary outcome can be confirmed or strengthened by a consistent effect on other relevant outcomes. The Long-Term Intervention with Pravastatin in Ischaemic Disease (LIPID) trial evaluated the effectiveness of pravastatin for preventing cardiovascular events in patients with diabetes or impaired fasting glucose and a history of coronary heart disease.8 The finding of a statistically significant reduction in the risk of a major coronary event was supported by a similar reduction in the risk of a revascularisation procedure or stroke. Such findings may also advance the understanding of the relationships between outcomes. PitfallsThe type and number of analyses performed should be reported so that readers can assess the probability of detecting a treatment effect by chance alone. This is often poorly documented in trial reports.5 Additionally, major discrepancies have been observed between the primary outcome specified in the trial protocol and that reported in the published article.5 Chan et al reported that of 76 trials that prespecified a primary outcome in the trial protocol, 20 (26%) did not report on this outcome in the published article, and of 63 trials that specified a primary outcome in the published article, in 11 (17%) it was not mentioned in the trial protocol.5 An example of the latter was a study reporting on the percentage of patients with graft occlusion as the primary outcome, even though the study was not originally designed to measure a difference in this outcome.5 Caution is needed when unexpected results from multiple analyses are interpreted. Inconsistent results are more credible if the outcome variables are restricted to those that were prespecified in the trial protocol, are clinically relevant, and are based on plausible biological mechanisms. A variety of statistical corrections can be performed to take into account the increased probability of a chance finding with multiple testing. A statistically significant treatment effect for one outcome and not other clinically related outcomes may also indicate that the sample is too small — that is, the study lacks power. Interpreting the results of analyses that are underpowered is difficult. This is a common problem, for example, in the mandatory reporting of adverse events in drug trials. A trial may report on a large set of adverse events, but it will commonly not have been powered to detect a statistically significant difference in these outcomes between the trial’s study groups.1 Composite endpointsTo overcome the problem of insufficient power, investigators may combine data from clinically related outcomes to form one or more composite endpoints. This approach reduces the number of analyses required while retaining all the potentially valuable information. The TAXUS IV trial was a double-blind randomised controlled trial to determine the safety and effectiveness in coronary artery disease of paclitaxel-eluting stents compared with bare metal stents.9 The primary outcome of the trial was the incidence of revascularisation procedures due to reocclusion of the target vessel at 9 months. A composite endpoint, “major adverse cardiac events” (defined as death from cardiac causes, myocardial infarction, or revascularisation procedures), was a secondary outcome. At one year after the procedure, the rates of cardiac death and myocardial infarction were similar between the study groups, while the rate of target-vessel revascularisation was 62% lower (P < 0.0001) in the patients receiving paclitaxel-eluting stents than in those receiving bare metal stents.9 The treatment effect on revascularisation rates appeared to drive the results for the composite endpoint, resulting in a reported 49% reduction in major adverse cardiac events (P < 0.0001) at 12 months. Combining disparate events can lead to an overestimate of the clinical importance if a positive finding is largely driven by the less important events. Overall, as a result of the potential to overinterpret or misinterpret the results of analyses of multiple outcomes, readers should seek information in the methods section of the report about the primary purpose and outcomes that the trial was designed to address and interpret any additional findings in this context. Adjusted analysesClinical trials use a concealed randomisation process, with or without stratification by key prognostic factors, such as age and sex, to help to ensure the baseline similarity of the study groups.10 However, even well-conducted random allocation may still result in chance imbalances.11 If an imbalance in an important prognostic factor occurs, statistical methods can control for this imbalance by including the factor as a “covariate”. This is referred to as adjusted analysis, or a multivariate analysis if more than one covariate is included. While adjusted analyses can statistically accommodate imbalances between study groups in non-randomised studies, in randomised studies they should usually be considered supplementary to the unadjusted analysis of the primary outcome. If the adjusted effect estimate differs from the unadjusted estimate, interpretation may be a problem. For example, if some covariate data are missing and these participants are excluded from the adjusted analysis, it will not be clear whether observed differences result from controlling for this factor or another, unknown effect of these exclusions. Adjusted analysis may be indicated when a factor is known to strongly predict the outcome (for example, age and survival), even when the imbalance observed between study groups does not reach statistical significance.12 In general, adjusted analyses frequently improve the precision of the estimate of treatment effect, even when the correlation of the covariates with the study outcome is not strong.13 Another recent review of 50 consecutive published trials showed that the methods and reporting of adjusted analyses vary widely in clinical trials.14 Of the 36 trials with an adjusted analysis of the primary outcome, 42% did not report on the methods used to select the covariates.14 Using inappropriate methods for adjusted analyses may cause inaccurate and misleading results. Ideally, investigators should prespecify any prognostic factors that, if unbalanced, may affect the study outcomes and should plan for adjusted analyses accordingly. However, some strong predictors of the outcome may only become apparent at data analysis on formal testing (so-called exploratory analysis). In this situation, investigators should clearly describe when and how covariates were selected for the adjusted analysis. In any case, the primary emphasis should be on the unadjusted results, because investigators are able to conduct multiple adjusted analyses using different sets of covariates, which may lead to overinterpretation or selective reporting of significant findings. The findings of the primary unadjusted analysis are strengthened if the results of the adjusted analysis are consistent with them. Other ancillary analysesA clinical trial may seek to address ancillary questions unrelated to the primary question so as to optimise the use of resources required for a large clinical trial. Ancillary questions may relate to the treatment effect on other conditions of interest, such as the association between hormone replacement therapy and dementia in women recruited to a large trial investigating hormone replacement therapy and cardiovascular disease.15 Substudies may also use trial data to investigate epidemiological questions about the natural history of disease, the biological mechanisms of the disease16 or the treatment response.17 Ideally, these ancillary studies should be designed before the trial starts. However, important new information or scientific debate may arise during or after the trial to justify the use of trial data to investigate new hypotheses. Their results are more convincing if the decision to conduct the analysis has been made before unblinding. The same potential for overinterpretation and selective reporting of the results of multiple comparisons and reduced power apply to exploratory analysis, and any new findings should be regarded as new hypotheses for validation in future studies. The principles of planning, reporting, analysing and interpreting multiple analyses are shown in Box 3. These are not intended to discourage investigators from conducting potentially important exploratory analyses of plausible new hypotheses. Rather, they encourage the balanced reporting of all analyses to prevent unsound manipulation of data or undue emphasis on particular findings that may misdirect future research or compromise the interpretation of results for clinical practice. 1 Definitions Primary outcome: The health parameter measured in all study participants to detect a response to treatment. Conclusions about the effectiveness of treatment should focus on this measurement. Primary analysis: The statistical test performed to determine whether there is a difference in the primary outcome between participants allocated to receive the treatment and those allocated to the control arm. Secondary outcomes: Other parameters that are measured in all study participants to help describe the effect of treatment. Baseline variables: The characteristics of each participant measured at the time of random allocation. This information is documented to allow the trial results to be generalised to the appropriate population/s. Specific characteristics associated with the patient’s response to treatment (such as age and sex) are known as prognostic factors. Multiple analyses: Comparisons between the study groups for more than one outcome. They increase the likelihood of detecting a difference between the treatment and control group owing to chance alone (false positive). Common examples of multiplicity in trials include the use of: multiple outcomes, including surrogate endpoints; multiple treatment comparisons (in a multiarm trial); subgroup analyses to detect differences in the treatment effect in one or more subsets of trial participants; adjusted analyses to control for imbalances in prognostic factors between the study groups; repeated measures over time of the same outcome; and interim analyses of the treatment effect at different stages in the trial. Exploratory analyses: Analyses that were not specified before the trial or, for blinded studies, analyses planned after the investigators were unblinded to the treatment allocation of participants. These analyses may be driven by the results of the primary analysis. 2 CONSORT checklist of items to include when reporting a trial6 Selection and topic Item no. Descriptor Ancillary analyses 18 Address multiplicity by reporting any other analyses performed, including subgroup analyses and adjusted analyses, indicating those prespecified and those exploratory. 3 Checklist for multiple analyses Design and methods Were the primary and secondary outcomes for the detection of treatment response prespecified? Was the trial designed to have adequate power for the analyses of all outcomes? Were the covariates for the adjusted analyses and/or the method used to select these covariates prespecified? Were the substudies based on an existing trial or biological data? Were the substudies planned prior to unblinding of data? Analysis Have corrections for multiple-significance testing been performed? Was the combination of data into a composite outcome appropriate? Was the interpretation of the composite endpoint results appropriate? Reporting Are the total number of analyses performed reported? Was the power calculation reported for the primary outcome? Secondary outcomes? Are the rationale and methods of any adjusted analyses reported? Are the number and type of covariates in the adjusted analyses reported? Are the unadjusted and adjusted results reported? Are the prespecified analyses clearly distinguished from the exploratory analyses? Interpretation Is appropriate emphasis given to the primary outcome? Have the relationships between interrelated outcomes been explored with equal interest? Are the findings of the multiple analyses discussed in the context of current biological knowledge and current research?

Sarah J Lord MB BS, MScEpid · Val J Gebski BA, MStat · Anthony C Keech MScEpid, FRACP

A 74-year-old woman presented to hospital with dyspnoea of sudden onset. She was in cardiogenic shock, with blood pressure of 90/50 mmHg, and pulse rate of 115 bpm. The electrocardiogram (ECG) was difficult to interpret because of severe, persistent and inexplicable artefacts (Box 1A). Chest x-ray showed acute pulmonary oedema thought secondary to left ventricular failure. Echocardiography showed a left ventricular ejection fraction of 40% and anterior akinesia. Coronary angiography confirmed a proximal occlusion of the left anterior descending coronary artery, which was successfully treated by direct angioplasty and stenting. Review of the patient’s previous medical records revealed, on a thoracoabdominal image derived from computed tomography data, a foreign body in the central abdominal region (Box 2). Questioning of the patient after the angioplasty revealed that this was an electrostimulator (Itrel II, Medtronic, Minneapolis, USA) implanted a few years earlier as part of a dynamic graciloplasty to treat faecal incontinence. An ECG performed 12 hours after the angioplasty with the device deactivated (by the patient’s control programmer) produced an ECG tracing free of artefacts (Box 1B). Persistent and inexplicable ECG artefacts should raise the suspicion of interference from a device generating high frequency electrical impulses. Such devices may include minute ventilation rate-responsive pacemakers1 and central or transcutaneous neurostimulators (eg, deep brain neurostimulators used in Parkinson’s disease,2 and spinal neurostimulators used in chronic back pain3). Occasionally, when the vector of impulses is perpendicular to an ECG lead, the tracing from that lead may be free of artefacts (eg,Box 1A, lead I). Intrathoracic devices are usually discovered on physical examination or standard chest x-ray, but devices in other parts of the body may not be obvious. In dynamic graciloplasty to treat anal incontinence, the gracilis muscle is transposed around the anal canal and electronically stimulated by a device placed in a subcutaneous pocket in the abdominal wall.4 In our case, this device led to ECG artefacts potentially interfering with the ability to diagnose the acute myocardial infarction. Implanted electrostimulation devices are being used increasingly in medicine and should be considered as a possible source of ECG artefacts, especially by frontline clinicians confronted daily with such life-threatening conditions as acute coronary syndromes. Electrocardiograms in a patient with an electrostimulation device A. On presentation, showing artefacts. B. After temporary deactivation of the device. 2 Thoracoabdominal image calculated from computed tomography data, showing the device (arrow)

Robert F Bonvini MD · Edoardo Camenzind MD

Correction Re: “Prevalence and nature of connexin 26 mutations in children with non-syndromic deafness”, by Hans-Henrik M Dahl, Kerryn Saunders, Therese M Kelly, Amelia H Osborn, Stephen Wilcox, Barbara Cone-Wesson, Julia L Wunderlich, Desiree Du Sart, Maria Kamarinos, Robert J McKinlay Gardner, Shirley Dennehy, Robert Williamson, Neil Vallance, Patricia Mutton (Med J Aust 2001; 175: 191-194). A recent audit of our use of Guthrie cards has shown that the 1000 anonymous Guthrie blood spots used to estimate the connexin 26 carrier incidence in our study were collected in a week selected at random from 1984, and not May 1986 as stated on page 193. (Also, in the Abstract [page 191], the collection year, given as 1986, should read 1984.) This does not affect the study results or our conclusions.

Hans-Henrik M Dahl

John Young, Professor of Physiology, former Dean of Medicine and Pro-Vice Chancellor at the University of Sydney, died on 10 February 2004 of acute myeloid leukaemia. He was an exceptional scientist and a humanist with a passionate and informed interest in classical culture, Reformation theology and music. John was born on 18 April 1936 in Brisbane. He graduated from the University of Queensland with a BSc (Hons I) in pathology in 1956 and with an MB BS (Hons I) in 1960, earning the University Medal. In 1962, after an internship at Royal Brisbane Hospital, John joined the Kanematsu Memorial Institute in Sydney, where his studies on renal amino acid transport earned him an MD from the University of Queensland (1965) and a C J Martin Fellowship from the National Health and Medical Research Council. He then joined Professor Karl Ullrich at the Physiologisches Institut of the Free University in Berlin. It was Ullrich who persuaded him to work on salivary physiology, the field that John was to dominate until his death. John took up a Senior Lectureship in Physiology at the University of Sydney in late 1966, rising to become Professor of Physiology in 1976. His contributions to science were recognised by the award of a DSc from the University of Queensland (1975) and the Research Professorship of the Alexander von Humboldt Stiftung (1998). He was elected to Fellowship of the Royal Australasian College of Physicians (1976) and of the Australian Academy of Sciences (1986), in which he served as Vice President and Secretary (Biological). After he became Dean of Medicine at the University of Sydney in 1989, he oversaw the introduction of the 4-year graduate program, the development of independent clinical schools, and the foundation of the Clinical School at Canberra. Between 1994 and his retirement in 2003, he was Pro-Vice Chancellor (Health Sciences). Although he was active on many boards, including those of the Central Sydney Area Health Service and the Children’s Hospital at Westmead, he was particularly proud of his role in fostering the Menzies School of Health Research (in Darwin) and (in Sydney) the Kolling Institute at the Royal North Shore Hospital and the Anzac Institute at Concord Hospital. John was always active outside his own profession. He co-edited the University of Sydney’s Centenary book of the Faculty of Medicine and worked unstintingly for the Australian Physiological and Pharmacological Society and for Musica Viva. He also found time to write several biographies on important figures in Australian medicine for the Australian dictionary of biography. Most importantly, his passion for classical art and archaeology led him to play key roles in the University of Sydney’s Nicholson Museum concerts and in the creation and development of the Australian Institute of Archaeology at Athens. David I Cook

David I Cook

Jean Edwards was an eminent radiologist who was also well renowned for her knowledge of general medicine. Jean was born in Woollahra, Sydney, on 25 November 1913. She attended Frensham School, and graduated in medicine from the University of Sydney in 1938. After working as a junior Resident Medical Officer at Sydney Hospital, she married George Read, a former fellow medical student, in 1940. On completing her studies, Jean intended to become a paediatrician. However, when the war intervened, her father asked her to join his radiology practice in Macquarie Street. As there was no formal training in radiology at the time, Jean became a “radiological apprentice”, learning to take, process and report on x-ray films. Soon afterwards, she was invited to join the honorary staff at the Rachel Foster Hospital (RFH) as a radiologist. During her time at RFH, she also held honorary positions at Royal Prince Alfred Hospital (RPAH) and St Vincent’s Hospital. Jean’s sons, Andrew and David, were born in 1941 and 1944, respectively. She continued her career while raising her children. However, her family commitments led her to decline a later offer to join RPAH as Director of the Radiology Department. In 1955, Jean joined a private practice at Campsie, where she remained until 1980. She continued to work as an honorary medical officer at three public hospitals (RPAH, RFH and Canterbury Hospital). These were very happy years. The patients were diverse, including many rugby league players and a pet python! Jean had developed a love of art, history and poetry while at Frensham School. She maintained a long association with the school, acting as both a Governor and Chair of the Board for many years. Jean was fond of a passage by Walter Savage Landor ending with the words “whatever may become of the fruit, make sure of the flowers and leaves”. This summed up her philosophy on life. Jean was a keen skier and a member of the Sydney University Ski Club. She was a talented artist, pursuing her interest in art as a member of the Medical Art Group. After her retirement, she worked as a volunteer at the Genealogical Society and enjoyed afternoons in Centennial Park with her grandchildren. Jean died on 30 May 2004, at the age of 90, after a prolonged illness. Rebecca Read

Rebecca L Read

Christine Y Lu,* Kenneth M Williams,† Lyn March,‡ James V Bertouch,§ Richard O Day¶ * PhD Candidate, † Deputy Director, ¶ Director, Therapeutics Centre, St Vincent’s Hospital, University of New South Wales, Victoria St, Darlinghurst, Sydney, NSW 2010. ‡ Rheumatologist, Royal North Shore Hospital, The University of Sydney, St Leonards, NSW. § Head, Department of Rheumatology, Prince of Wales Hospital, Randwick, NSW. christine.luATstudent.unsw.edu.au To the Editor: The tumour necrosis factor (TNFα) inhibitors etanercept, infliximab and adalimumab are new treatments for rheumatoid arthritis (RA) subsidised by the Pharmaceutical Benefits Scheme (PBS) under strict criteria. These are based on data supplied by the sponsors and protracted discussions between the stakeholders: the Pharmaceutical Benefits Advisory Committee (PBAC), the sponsors, rheumatologists and consumer representatives.1 One eligibility criterion is that adult patients must be or have been rheumatoid-factor positive. This requirement raises particular concerns. Rheumatoid factor is not exclusively associated with RA. It is found in a number of other autoimmune and infectious diseases, and has been detected in healthy people. Rheumatoid factor is a serological criterion in the American College of Rheumatology classification for RA. However, its presence is not definitive for a clinical diagnosis of RA, with rheumatoid factor being absent in about 30% of patients with RA. Studies suggest that patients who test positive for rheumatoid factor at baseline are more susceptible to relatively severe expression of RA, with development of erosions and functional disability.2 As the PBS already limits access to TNFα inhibitors to patients with severe active RA, the predictive value of rheumatoid factor for severity is redundant. We systematically reviewed 23 clinical trials of TNFα inhibitors available in the public domain (references available on request). The review showed that only the Early Rheumatoid Arthritis trial restricted recruitment to patients who tested positive for rheumatoid factor.3 The aim of this study was to examine the effect of etanercept on the rate of development of erosions in patients with early RA, not whether rheumatoid-factor status affected response. We conclude that the evidence for an association between rheumatoid-factor status and response to TNFα inhibitors in patients with severe RA is inadequate to justify rheumatoid-factor status as a criterion for PBS subsidisation. Similarly, no evidence for an association between rheumatoid factor and response to disease-modifying anti-rheumatic drugs has been reported.4 Nevertheless, potential predictors of response to TNFα inhibitors and such drugs are emerging (eg, polymorphisms of HLA-DRB1, TNFα, and interleukin-10).5 Further analysis of the unpublished individual patient data held by the sponsors is needed to shed light on this question. Clinical studies specifically designed to evaluate the influence of rheumatoid-factor status on response would be helpful. The PBAC may have been provided with these data, but patients and clinicians operating in the public domain have not. This gives rise to an ethical dilemma whereby prescribers cannot provide a plausible explanation for why rheumatoid-factor status is an access criterion. For example, if rheumatoid-factor-positive status was selected to exclude patients with rheumatoid-factor-negative psoriatic arthritis, then the rationale would not appear to be ethically sound. Greater transparency with respect to the rationale and the evidence for the criteria selected for targeting subsidised access to important high-cost pharmaceuticals will enhance confidence in the PBS process.

Christine Y Lu · Kenneth M Williams · Lyn March · James V Bertouch · Richard O Day

Lloyd N Sansom Chair, Australian Pharmaceutical Benefits Advisory Committee, School of Pharmacy, University of South Australia, Adelaide, SA 5000. Lloyd. SansomATunisa.edu.au Comment: I am grateful for the opportunity to put the perspective of the Pharmaceutical Benefits Advisory Committee (PBAC) on this important issue for patients and prescribers using TNFα inhibitors for rheumatoid arthritis. I can confirm that the PBAC based its recommendation to exclude patients who test rheumatoid-factor negative on a meta-analysis of the data presented to it in the sponsor’s reports of two published randomised trials.1,2 The sponsor of etanercept requested a restriction that limited use to patients with rheumatoid-factor-positive status — a proposition that was queried by the PBAC in its evaluation. The sponsor indicated in its response that rheumatoid-factor status was a treatment-effect modifier. As anticipated by Lu and colleagues, the reason for the recommendation was neither to confirm the diagnosis of rheumatoid arthritis nor an attempt to identify patients with more severe disease. The main reason is that there was some statistical support for the contention that, in patients with rheumatoid arthritis, being rheumatoid-factor positive is associated with a better response to treatment with these drugs. Unfortunately, my request to the sponsor to be able to release in this letter the information on which the decision was based was refused. The statistical analysis relied on by the PBAC was based on individual patient data, stratifying trial participants into two groups (rheumatoid-factor positive and rheumatoid-factor negative), and then formally applying a test for interaction against the reported treatment effect compared with placebo. Although a post-hoc analysis, this approach has the advantage of relying on the entire trial dataset rather than increasing the likelihood of detecting spurious differences by conducting a series of post-hoc sub-group analyses. It also differs from the systematic review reported by Lu and colleagues, which sought to examine the question based on the reported eligibility criteria of the trials, and therefore could only be conducted at the level of the overall trial, not the level of the individual patient in the trials. The interpretation of the data available to the PBAC at the time was supported by rheumatologists advising the PBAC about actual restrictions for the subsidy of these expensive agents. It is perhaps noteworthy that many decisions that guide treatment are made purely on the basis of biological plausibility. On this occasion, there was some statistical support, albeit from a post-hoc analysis, to support any biological arguments. There is inevitably an element of judgement about these decisions, as is clear in the letter from Lu et al, who describe “accumulating evidence” in relation to whether gene polymorphisms “may be predictive of clinical responsiveness”. The current situation, as exemplified in this letter, in which the PBAC is unable to give detailed reasons for its decisions, clearly demonstrates the urgent and critical need for greater transparency of the PBAC processes, and underscores the fundamental right of Australian consumers and prescribers to information relevant to decisions about the subsidy of medicines in this country. The PBAC is willing to examine any new evidence relating to the influence of rheumatoid-factor status on health outcomes and the cost-effectiveness of these agents. If there is additional evidence supporting a review of the current listing, the PBAC would encourage the sponsors of these drugs to prepare appropriate submissions to enable us to evaluate these data. As Chair of the PBAC, I intend to continue to work towards the goal of greater transparency, whereby the outcome of any such new evaluation, and its basis, could be made available to Australian consumers and prescribers.

Lloyd N Sansom

Tim Woodruff President, Doctors Reform Society, Suite 207, 320 Victoria Parade, East Melbourne, VIC 3002 twoodruffATbigpond.com To the Editor: It is encouraging to see the Journal continuing its tradition of taking on contentious issues in publishing articles about the negative influence of the pharmaceutical industry.1 With respect to the practical suggestions to address this issue, I suggest that reliance on our profession to substantially improve the situation, although laudable and appropriate, is too optimistic given the gross denial by our colleagues that there is an issue. Our professional bodies simply do not have the support to enforce codes of conduct. The Royal Australasian College of Physicians (RACP) has published guidelines on this issue,2 but I doubt most of its members have read them. I fully support Breen’s comments relating to funding of educational activities, but I suggest that the pharmaceutical pseudo-educational dollar be bypassed by a major expansion in government funding.3 The provision of regularly updated, easily accessible treatment guidelines integrated into prescribing software (which most general practitioners use daily) would go a long way to decreasing our reliance on the drug dollar for information on appropriate treatment. This requires government investment and professional college cooperation, but would lead to recurrent savings to the Pharmaceutical Benefits Scheme and better treatment. Currently, the federal government spends $21 million on drug information to doctors,4 while the drug industry spends $1 billion on marketing.5 To partially redress this imbalance would, however, require both political will and pressure from the profession.

Tim Woodruff

Linda V Graudins Senior Pharmacist (projects), Prince of Wales Hospital, High Street, Randwick, NSW 2031 graudinslvATsesahs.nsw.gov.au To the Editor: Thank you for publishing the Viewpoint by Breen regarding pharmaceutical industry relationships with the medical profession1 — but please do not coin the word “pharmaproof”. This word unintentionally casts aspersions on fellow clinicians — pharmacists — who share the dilemma of aggressive pharmaceutical marketing influencing professional decisions. The relationship between the industry and pharmacists is actually more complicated, as we are not only advocates for patients and advisers to doctors on the safe and evidence-based use of medicines, but also the buyers of the pharmaceutical products. This last function means that most pharmacists cannot be removed from the business side of medicine supply and must work with the manufacturers to obtain supplies in a timely fashion and at the best price. In my work as a hospital pharmacist, the pharmaceutical industry helps in planning financial aspects of medicine supply, sponsoring various activities that the public health system and universities are unable to, and providing specific product information — be it for marketed, unregistered or trial medicines. Most pharmacists cannot choose to not see industry representatives. The formation of such relationships can indeed insidiously affect our clinical decision-making, as outlined by Breen. Unfortunately, the issue of industry’s influence on pharmacists’ decision-making has only rarely been discussed in the pharmacy literature.2-4 Both medical and pharmacy clinicians must be aware of this influence and act accordingly. Peter Mansfield’s Healthy Skepticism (www.healthyskepticism.org) is a good starting point to increase this awareness. However, the movement must include physicians and pharmacists on the same side. I have often been told that we must have the latest new drug on the hospital’s formulary because . . . and have been given a hefty manufacturer-prepared dossier as the sole reason for the request. I urge physicians to work with pharmacists and be “pharma(cist)friendly”, yet also “industryaware”.

Linda V Graudins

Scott Masters General practitioner, Musculoskeletal Medicine, Caloundra Spinal and Sports Medicine Centre, 39 Minchinton St, Caloundra, QLD 4551 cfmpATozemail.com.au To the Editor: The Journal recently published three interesting articles on the relationship between the medical profession and the pharmaceutical industry.1-3 With Medicines Australia (the pharmaceutical manufacturers’ association) setting up a strict code of conduct (tighter than for any other industry I know), a better balance seems to be on the horizon. Breen reminded us of our responsibilities to protect ourselves and our patients from slick marketing by pharmaceutical companies.1 One technique our surgery has found useful is to have a personal code of conduct. Our surgery has a guide for pharmaceutical reps (copies available from the author). It advises reps that we are not interested in seeing their promotional material, especially those useless coloured graphs. However, we are very happy to look at published trials regarding their product and associated diseases. Personally, I have found the resources available from many reps useful and time-saving. Breen is concerned about our professional leaders being in denial about the influence of the pharmaceutical industry on doctors’ prescribing habits. If this is so, then the same leaders have complete amnesia and catatonia about another influence that potentially threatens to engulf us. The sale of supplements and complementary medicines in Australia is a billion-dollar business now. Every month I receive more requests to use supplements for conditions varying from heart disease, cancer and fatigue to non-specific therapies such as detoxification, immune support, metabolic enhancer and anti-ageing. I can sell all these products directly to consumers (patients) at a mark-up I consider reasonable. Alternatively, I can recruit patients to become sellers in a multilevel marketing scheme (similar to pyramid selling). One doctor who practises nutritional medicine full-time has told me he buys $10 000 worth of vitamin E at the start of the year and manages to sell it over the ensuing 12 months for $100 000. I imagine most of that doctor’s patients are recommended vitamin E for their health complaints or health maintenance. This is entirely legal, although there are major ethical concerns about conflict of interest. To date, there has been little debate among our leaders regarding proper guidelines and regulation of this behaviour. With the enormous potential of the complementary industry to be a useful partner in health management, this needs to be sorted out sooner rather than later.

Scott Masters

Rosanna Capolingua Chair, Ethics and Medico-Legal Subcommittee, Australian Medical Association, PO Box 6090, Kingston, ACT 2604 Comment: I commend Masters on his personal code of conduct in dealing with the pharmaceutical industry. I also share his concerns about the ethical minefield that lurks in the interface between complementary and conventional medicine. Some of the issues involved have recently been explored in the Journal’s series on Complementary and Alternative Medicine.1 However, his scenario of a doctor selling vitamin E to patients raises specific concerns, and the Australian Medical Association’s Code of Ethics2 provides some ethical principles in this regard. Specifically, it states that a doctor should: “make sure that you do not exploit your patient for any reason” “exercise caution in publicly endorsing any particular commercial product or service not covered by the Therapeutic Goods Advertising Code”3 and “when referring your patient to institutions or services in which you have a direct financial interest, provide full disclosure of such interests”. Translating these principles into daily professional conduct means the practitioner must at all times declare pecuniary interest in the sale of products and be aware of the evidence base of the commercial recommendation so as to not mislead the patient. This involves clear communication of potential benefit, adverse effects, and possible drug–drug interactions of whatever product is promoted. But, above all, the whole thrust of the AMA Code of Ethics is to ensure that perverse incentives remain foreign to the patient–doctor relationship. Ultimately, it is an individual doctor’s choice.

Rosanna Capolingua

Hugh G Dickson Chair, Human Research Ethics Committee, South Western Sydney Area Health Service, Locked Bag 7103, Liverpool BC, NSW 1871. Hugh. DicksonATswsahs.nsw.gov.au To the Editor: Roberts et al1 state that the National Health and Medical Research Council’s National statement on ethical conduct in research involving humans2 clearly outlines that, once approval has been gained from one human research ethics committee (HREC), other sites should accept that approval. Nowhere does the national statement state this. What Roberts et al appear not to appreciate is that research projects may consume resources within a health service. The HREC might also have the responsibility of ensuring that the tasks required for successful completion of research can be performed without compromising the standard of care of the health facilities in which the research is to be performed. The project in question required the extraction of medical records — someone has to do this and be paid for the task. While centralisation of human research ethics approval might make the task of approval of research faster, each administrative unit has to be able to decide whether to allow participation in a project given its current level of service demand. Approval for multicentre research will therefore still take time whether or not central ethics approval is obtained, and researchers will not find an easy solution to this.

Hugh G Dickson

Kerry J Breen Chair, Australian Health Ethics Committee, National Health and Medical Research Council, PO Box 9848, Canberra, ACT 2601. kerrybreenATaccess.net.au Comment: It is correct that the National Health and Medical Research Council’s National statement on ethical conduct in research involving humans1 (paragraph 3.4) does not say that other sites should accept ethics review of a research proposal from another human research ethics committee (HREC). However, the national statement is written to permit and encourage this, but not to enforce it. Dickson’s letter also raises the issue of the terms of reference of HRECs. He alludes to an additional role — as gatekeeper of an institution’s resources — which some may feel is beyond the domain of HRECs. The national statement provides the framework for the composition and role of HRECs in their work of ethical review, but does not prevent an institution from asking its committee to also be the steward of access to resources regularly used (eg, medical records, pathology laboratories). The national statement requires an institution to set out the terms of reference for its HREC, “including the scope of its responsibilities” (paragraph 2.2). The prime function of HRECs is to make recommendations about whether a research proposal is ethical, but the final approval about whether the research can proceed rests with each individual institution. Thus, there is more to getting approval for a research proposal than just the ethical review.

Kerry J Breen

David J Maxwell,* Karen I Kaye† * Project Officer, † Executive Officer, New South Wales Therapeutic Advisory Group, PO Box 766, Darlinghurst, NSW 2010. tagprojectATstvincents.com.au To the Editor: We are currently coordinating the New South Wales arm of a national quality assurance (QA) project to improve the management of community-acquired pneumonia in Australian hospital emergency departments (CAPTION project: Community-Acquired Pneumonia: Towards Improving Outcomes Nationally, funded by the National Prescribing Service). In the project drug usage evaluation methods are used, and it involves an audit (retrospective review of patient medical records) as well as feedback and targeted education of healthcare professionals about national management guidelines for community-acquired pneumonia.1 There is no direct patient contact at any stage. Because this is a QA project, we did not request that participating NSW hospitals seek human research ethics committee (HREC) review. However, we did request that the project proposal be supported by HRECs as a QA activity. Further endorsement from the Chief Executive Officer, Director of Emergency, Director of Pharmacy and other key people and groups was also required. According to the National Health and Medical Research Council (NHMRC),2 an appropriately planned QA activity can proceed without HREC review and patient consent if: there is consistency with National Privacy Principle 2.1(a); and all people involved in the activity are unlikely to suffer burden or harm. To assist hospitals reviewing this QA project proposal, we developed a number of tools that explicitly outlined the nature of the project and provided answers to the questions in the NHMRC document. 2 These tools were provided to emphasise that this QA project was consistent with NHMRC requirements. The NHMRC further describes the establishment of hospital policies to allow efficient review of QA proposals that involve minimal risk, burden, alteration of care or invasion of privacy. It is recommended that a member of the HREC be appointed to review such proposals. Despite the NHMRC recommendations and the provision of tools as described above, full HREC submissions have still been required by four out of eight hospitals to date (recruitment still ongoing). After receipt of an initial expression of interest, completing the HREC submission and obtaining approval in three of these hospitals has taken an average of 5 months. The HREC review has not yet been completed in the fourth hospital. As the project is only funded for a 2-year period, this has resulted in a significant delay in the progress of the project. We do not challenge the vital role of HRECs, and we value the importance of ensuring the safety and privacy of all involved in human research. If QA coordinators and HRECs work together within the NHMRC recommendations, valuable time and resources could be saved, for both parties. In doing so, the Australian Research Ethics Committees’ belief that “quality assurance activities are an essential and integral part of health care delivery that should be encouraged and facilitated” would be one step closer to becoming a reality.

David J Maxwell · Karen I Kaye

Martin B Van Der Weyden Editor, The Medical Journal of Australia, Locked Bag 3030, Strawberry Hills, NSW 2012. medjaustATampco.com.au Comment: The distress conveyed by Maxwell and Kaye is the latest in a litany of concerns and complaints to the Journal over recent years about the workings of ethics committees.1 At the centre of the protests is the perceived tardiness and inefficiency of human research ethics committees (HRECs) in Australia.2,3 But the past decade has also seen reform. The National Health and Medical Research Council (NHMRC) has clarified the relationship between research and quality assurance and when quality assurance proposals require ethical scrutiny.4 Furthermore, the NHMRC’s National statement on ethical conduct in research involving humans is clear in its advice that HRECs must establish working procedures that include “timely consideration and review of research protocols”, and it also outlines procedures for expediting review of minimal-risk research.5 On multicentre research, the statement is also clear, stressing the need to minimise unnecessary duplication in review of such research, and calling for “prompt and efficient consideration of multi-centre research protocols” and adoption of “other administrative procedures to accelerate timely consideration and avoid unnecessary duplication”.5 With such clear enunciations, what could go wrong? But, as Maxwell and Kaye attest, things do go wrong. It seems that either the NHMRC recommendations have yet to be translated into reality or the “silo” mentality of HRECs is deeply ingrained. The time may well have arrived to consider performance indicators for HRECs, or even an accreditation process. After all, clinical research and quality assurance should not be at the mercy of inefficient HRECs.

Martin B Van Der Weyden

Gordon R W Davies Psychiatrist, 33 Smith St, Wollongong, NSW 2500. alienistATihug.com.au To the Editor: The recent article on obesity by Proietto and Baur in the Journal1 coincided with another by Campos in New Scientist,2 in which he criticised the conventional view of the risks of obesity and the norms usually accepted. Campos quoted evidence suggesting that, in fact, the group in the overweight range (body mass index [BMI], 25–30 kg/m2) are healthier than those with a BMI below 25 kg/m2. He also noted that between 1990 and 2002, despite a further increase in the prevalence of obesity in the United States, the incidence of type 2 diabetes hardly changed, while cardiovascular death rates fell. According to Campos, similar claims about the risks of obesity have been repeated over the past 50 years and relate more to cultural and political factors than to reliable scientific evidence. Clearly, this view is inconsistent with that articulated by Proietto and Baur. While Campos’ view obviously does not apply to the grossly obese, there is a strong suggestion that the overall evidence base is inconsistent. This may be because the assumption of a linear relationship between excess weight and illness is false. It is further likely that there is confounding of variables, with weight a proxy for lack of exercise. As Campos points out, large-scale observational studies are inevitably poorly controlled. If this is so, then it may well be more useful for the medical profession to emphasise exercise and lifestyle rather than weight loss. It may be much easier to obtain and reinforce behavioural change in these areas, and avoid the common feeling of hopelessness (“why bother”) expressed by people who find it hard to diet and to lose weight.

Gordon R W Davies

Ray C McHenry,* Richard W Gilhome,* Chris Hensman* * General Surgeon, Eastern Surgical, Suite 7529, Police Road, Mulgrave, VIC 3170. To the Editor: We take issue with the recommendations on treatment of morbid obesity in the otherwise excellent article on obesity management by Proietto and Baur.1 Like most non-surgical clinicians involved in the management of obesity, they fail to differentiate between the treatment of obesity (body mass index [BMI], 30–35 kg/m2) and morbid obesity (BMI > 35 kg/m2). The literature is crystal clear — non-surgical treatments are unsuccessful in achieving and maintaining weight loss in morbid obesity.2,3 We are unaware of any branch of medicine, other than morbid obesity management, where respected clinicians routinely recommend treatments (drugs, diet and lifestyle modification) which have been proven not to be effective. We challenge all clinicians to accept what the evidence clearly shows, that: the only known effective treatment for morbid obesity is surgery;2,3 and laparoscopic adjustable gastric banding is much safer than bypass/diversion surgery4 and just as effective; it is the treatment of choice for morbid obesity.

Ray C McHenry · Richard W Gilhome · Chris Hensman

Huy A Tran Director of Clinical Chemistry, John Hunter Hospital, Hunter Region Mail Centre, Locked Bag No 1, New Lambton Heights, NSW 2310. huy.tranAThunter.health.nsw.gov.au To the Editor: I read with interest the recent article on obesity in Australia by Proietto and Baur1 and would like to comment on the issue of proteinuria and measurement of insulin level in obese patients. Proteinuria in obesity, commonly referred to as obesity-related glomerulopathy, is a clinical syndrome with an estimated incidence of about 2% in obese subjects.2 With a fifth of the population being obese,1 the sheer number suspected to have this condition will create an enormous management and cost burden. Furthermore, the incidence of this condition appears to have increased disproportionately to the incidence of obesity.2 The syndrome of obesity-related glomerulopathy comprises the triad of morbid obesity, marked proteinuria without oedema, and normal serum albumin concentration. It can occur in any degree of obesity but is more common in the morbidly obese group (body mass index > 40 kg/m2; Class III obesity). It often presents as proteinuria on urinary dipstick testing, with marked proteinuria seen on confirmatory testing (up to 32 g/day).2 Other features of the nephrotic syndrome do not occur, and the cholesterol level is often lower than that in patients with nephrotic syndrome. However, glomerular filtration rate is raised, and glomerulosclerosis is seen on biopsy. The pathogenesis is unknown. Obesity-related glomerulopathy is a diagnosis of exclusion: secondary causes of proteinuria should be fully eliminated, including hypertensive renal disease and undetected type 2 diabetic renal disease. More often than not, biopsy will be required to guide management, with cost implications. Although the condition is said to be benign, in a small proportion of patients it progresses to end-stage renal failure requiring replacement therapy, further adding to the cost of management. Fortunately, the condition is readily reversible with weight loss, which is an important emphasis in management. 3 My second comment relates to the case of the overweight adolescent described by Proietto and Baur. In this patient, measurement of insulin level is not indicated.4 There is no standardised insulin immunoassay, the sample has to be collected and processed correctly to produce a valid result, and the result would not add to or alter the management of the condition. It is doubtful if normative data exist for adolescents, but the clinical picture suggests the insulin resistance syndrome. As the primary goal would be to detect disordered glucose metabolism, appropriate testing of glucose level is all that is required.

Huy A Tran

Joseph Proietto,* Louise A Baur† * Endocrinologist, Department of Medicine, Repatriation Hospital, Heidelberg, VIC 3081; † Paediatrician, Children's Hospital at Westmead Clinical School, Sydney, NSW. j.proiettoATunimelb.edu.au In reply: We agree with McHenry and colleagues that, until very recently, surgery was the only effective treatment for morbid obesity. However, the development of effective pharmacotherapy that targets the underlying cause — increased hunger — may well change this situation, as evidenced by the effect of leptin treatment in leptin-deficient children.1 In the not-too-distant future, a medical alternative may be possible. The issue of the relative merits of surgery to insert a foreign body (gastric band) or permanently alter the anatomy of the gastrointestinal tract versus lifelong pharmacotherapy will need to be considered. However, economic as well as health issues may be important, and, as McHenry and colleagues suggest, may still favour surgery as the preferred therapy. We thank Davies for bringing to our attention Campos’ book The obesity myth, in which he claims that overweight individuals are in fact healthier than those of normal weight.2 While many of the book’s other claims can be challenged (such as the statements that the prevalence of type 2 diabetes is not rising in the United States, and that bald men have higher testosterone levels), the fact that there is no simple linear relationship between body mass index (BMI) and illness is correct. Sex, race and fat distribution can all influence the relationship. Moderately overweight women with gynoid (hip and thigh) fat are not at increased risk of illness.3 In contrast, South Asian people have an increased risk of developing diabetes at lower BMI values than people of European background. We agree with Davies that maintaining fitness through regular exercise is very important in minimising the health consequences of obesity. Tran raises the issue of whether it is useful to measure insulin levels in children. It is not unreasonable to assume that insulin levels are raised in most obese children, but this is not always the case. The National Health and Medical Research Council clinical practice guidelines for management of overweight and obesity in children and adolescents state that: “Fasting insulin and glucose should be considered in obese children or adolescents, particularly those with a family history of type 2 diabetes, those with acanthosis nigricans and those from certain ethnic backgrounds”.4 In the presence of insulin resistance, serum glucose level remains normal because of high insulin levels. Thus, glycaemia cannot be used to monitor improvement in insulin sensitivity.

Joseph Proietto · Louise A Baur

Richard X Davey Chemical Pathologist, Melbourne Health Shared Pathology Service, Western Hospital, Gordon Street, Footscray, VIC 3011. richard.daveyATwh.org.au To the Editor: In 1996, the Journal published my assessment of the scientific validity of a 1994 decision by the Australian Health Insurance Commission (HIC) to limit the Medicare rebate payment for assessment of thyroid function to thyroid-stimulating hormone level (TSH), except in certain more complex clinical conditions.1 For these, levels of TSH and of other indicators of thyroid function, such as thyroxine, are tested simultaneously and a rebate given for the group of tests. Although discussion concerning this diagnostic strategy persists,2 nearly a decade later it is appropriate to assess the outcome of this initiative. Publicly accessible HIC data3 on privately ordered, then publicly refunded, thyroid function tests in Australia for the fiscal years 1994–2002 were retrieved and are presented graphically (Box). Figure A shows the change in thyroid function test ordering sought by the HIC occurring in 1997 through to 2000 and probably now consolidating. Overall, before the initiative, there were 1.55 TSH tests ordered for each thyroid function group test, and in 2002 this increased to 2.65. The outcome in the elderly is similar, but the change is even more noticeable for thyroid testing in young women and men (with ratios of about 5 to 1 and 7 to 1, respectively). Clinicians have presumably come to accept the high negative predictive value of a normal TSH result for ruling out primary thyroid disease as both necessary and sufficient to finalise thyroid diagnoses in the young. By contrast, thyroid disorders are more common among the elderly, who thus more readily satisfy the HIC requirements for thyroid function group tests. The approximately fourfold increase in absolute terms in the number of single TSH tests performed over 9 years (from 2731 per 100 000 persons per annum in 1994 to 10 763 in 2002) can also be seen as vindicating the HIC’s decision to run with a “TSH first” testing protocol. The same pattern is seen in the age and sex groups illustrated (15–24 years and 75–84 years). The present speed, ease and relative economy of obtaining a TSH test, and the reliability, particularly at low TSH levels, of using this measure for thyroid disease case finding, make ordering a TSH test no longer an indulgence,4 but a clinical necessity. The national cost-of-living index for 30 June of each fiscal year5 was used to standardise the annual expenditure on thyroid function tests to 1994 dollar values, thus allowing comparison across the decade (Figure B). From 1995 to 2002, the HIC has contained annual expenditure to under $20 million for thyroid function group testing. This is both desirable and appropriate. That this has been sustained for 7 years in the face of increasing numbers of first-line TSH tests is both astonishing and commendable. Most of the increase in costs of TSH testing is probably explained by the acceptance of its use as a first-line test. Between the 1996 and 2001 censuses, the population grew from 17.9 to 18.8 million, and the proportion over 65 years also increased (from 12% to 12.5%). Both these trends are continuing,5 and both also explain some of the increase in ordering of TSH tests. It is unclear if any of this change is also due to testing moving from the totally public, hospital sector, not funded by the HIC, to the HIC-funded sector. Effect of changes to the Health Insurance Commission rebate for thyroid function testing A: Ratio of the number of single tests ordered (thyroid-stimulating hormone [TSH]) to the number of thyroid function group tests ordered (TSH and thyroid hormones). Data are tests per 100 000 persons per annum. B: Annual expenditure on thyroid function testing in Australia.

Richard X Davey

Jan R Stockigt Senior Endocrinologist, Alfred Hospital, Commercial Road, Prahran, VIC 3181. jrsATnetspace.net.au Comment: In his timely review of changing patterns of thyroid function testing, Davey suggests that Australian Health Insurance Commission (HIC) policy is responsible for the increased emphasis on a “TSH-first” strategy, with consequent containment of costs for other thyroid function tests. While this may in part be true, the trend towards initial TSH testing has been advocated worldwide1 following the development of TSH assays sufficiently sensitive to distinguish the typical suppressed TSH levels of thyrotoxicosis from normal levels. The developments documented by Davey are a consequence of technological development, perhaps enhanced by selective rebating as a result of HIC policy. It is unfortunate that current HIC policy is sometimes described as prohibiting more complete thyroid function testing, unless TSH level is abnormal. Rebate policy does not prohibit any line of testing and it is because the “TSH-first” approach has some serious, well-documented deficiencies.2 Measurement of levels of thyroid hormone in addition to TSH is clearly sanctioned in HIC regulations when TSH level alone can be misleading, for example in suspected pituitary dysfunction, or in monitoring the treatment of thyroid dysfunction. The adverse consequences, both human and financial, of relying on TSH measurement alone in such situations can be serious and may outweigh the savings achieved by restrictive testing. It must be noted again that a normal concentration of immunoreactive TSH has no predictive value in ruling out potentially life-threatening hypopituitarism,3 which may present with prominent hypothyroid features. The effective integration of clinical and laboratory investigation of potential thyroid dysfunction requires an active laboratory–clinical interface. There are over a dozen patterns of thyroid function — some trivial or inconvenient, some quite serious — that can be misdiagnosed or incorrectly managed if communication across this interface is inadequate.4 Effective communication requires relevant information from the clinician and a response to this information within the laboratory. It is a reality that current patterns of investigation in Australia frequently fall short of this ideal. If, as a result of automation and effective competition, the unit cost of assays can eventually be reduced in relation to the total cost of medical care, it may become appropriate to revert to a more complete panel of initial testing that integrates tropic hormone and target gland secretion, a strategy that remains the cornerstone of definitive endocrine investigation.

Jan R Stockigt

Peter J Gilbar,* Christine V Carrington† *† Co-Chair, Committee of Specialty Practice in Oncology, The Society of Hospital Pharmacists of Australia, Suite 3, 27-33 Raglan Street, South Melbourne, VIC 3205. peter_gilbarAThealth.qld.gov.au To the Editor: The national media recently highlighted the tragic consequences of the inadvertent spinal administration of the antineoplastic drug vincristine. The 7.30 Report (ABC Television) detailed the case of a young man erroneously administered vincristine intrathecally instead of, as intended, intravenously, resulting in progressive neurotoxicity, paralysis and death.1 Since the first report in 1968 of unintentional intrathecal administration of vincristine,2 many, invariably fatal, cases have been described. These have involved a combination of human and system errors affecting the medical, nursing and pharmacy professions. On behalf of the Society of Hospital Pharmacists of Australia, we recommend the following strategies to reduce the opportunity for error: Only specifically trained and designated oncology staff should prepare, dispense and administer cytotoxic medication. Intrathecal chemotherapy should only be administered during normal working hours, and in an area where no other cytotoxic drugs are given or stored. Medical staff must use a formal checking procedure, involving an oncology-trained nurse, to ensure the right drug is given at the right dose, by the right route, to the right patient. Intrathecal drugs must be packaged separately and clearly labelled both on the syringe and outer container “For intrathecal use”. Specifically designated containers should be used for transportation of intrathecal drugs from the pharmacy and for storage on the ward. Intrathecal doses should be delivered separately and preferably administered after drugs to be given by other routes have been supplied to the ward and administered. Vincristine should be prepared in a small-volume intravenous bag rather than a syringe.3 For adults, prepare vincristine in an intravenous bag in 50 mL of sodium chloride 0.9% and administer it as a short intravenous bolus over 5–10 minutes. Smaller volumes and a slower administration rate are suggested for children. While this method has been criticised as potentially increasing the risk of extravasation injury, this has not been reported as a problem. Vincristine should be clearly labelled both on the intravenous bag and outer container “For intravenous use only — fatal if given by other routes”. Negative labels, such as “Not for intrathecal use”, must never be used. Many hospitals currently use syringes for vincristine and increase the diluent volume in the syringe as a deterrent to intrathecal administration; however, fatalities have occurred after the administration of vincristine supplied in 10 mL4 and 20 mL5 syringes. The safest method of eliminating the potential for spinal instillation of vincristine remains the abolition of the syringe as a means of administration.

Peter J Gilbar · Christine V Carrington

Project management in health and community services Judith M Dwyer, Pauline Stanton, Valerie Thiessen. Sydney: Allen and Unwin, 2004 (xxi + 200 pp) ISBN 1 74114 055 2. Would I be interested, the polite letter from the Journal asked, in reviewing a book about project management? I turned the letter over gloomily — I couldn’t think of a less interesting title. The sub-text, however, Getting good ideas to work, was a bit of a pull, as I have a big project just starting. One of the frequent deficits of doctors when they assume levels of seniority is a thorough incompetence at management. We are trained with quite different skills in mind, like caring for individual people. Project management is not always well done, even for medicos involved in research — projects often fizzle out not because of problems with design, or messing up the sample, but because of pragmatic failings. So this book might have relevance to all of us involved with projects, and — look! — written for us health folk, and by true-blue Aussies. The book comes in two parts: some of the research and thinking behind projects, and where they fit in organisations and their strategies. There is quite a bit of theory, but few pragmatic data on what works. The theory is littered with technical stuff: it was good to get my head around things I had come across but wasn’t really sure about, like PERT (Project Evaluation and Review Technique), Gantt Charts, WBS (Work Breakdown Structure) and PMBOK (Project Management Body of Knowledge). There isn’t anything here that is intellectually tricky. Don’t worry, they say, about all the technical stuff. Forget the acronyms, all you really need is buckets of common sense. We are introduced to tools (mostly software) that help manage projects, but the finger is wagged in our faces about not relying on them, and the plans they generate. “Project management is an art not an algorithm” (p63). So does any this stuff work? Empirical research is weak, but suggests the usual suspects for failure: insufficient planning, buy-ins from the organisation, planning and so on… The second part of the book is more of a workshop manual. What to do, and, more pertinently, what not to forget. It suggests telltale signs of things going wrong; how to spot the folk who might cause you problems (like the white-ant); and what to do about them. This is the sort of book in which you might pause, smile in recognition, and read on again. There are examples — some real — from Australian hospitals, community health settings, and also Divisions of General Practice. A good book. I am the better for having read it. I think. Christopher B Del MarDean Health Science and Medicine Bond University Gold Coast, QLD

Christopher B Del Mar

Atherosclerosis and heart disease. Andrew M Tonkin (editor). London: Martin Dunitz, 2003 (xi + 235 pp). ISBN 1 84184 123 4. The Internet is great for reviewing literature rapidly, but the more I use it, the more I realise that the classic “Review article” has a role. PubMed can be very efficient for accessing publications, but drawing accurate conclusions can be hard work and reliance on abstracts can stifle critical analysis. Independent critical comment is difficult to find electronically. Journals realise they have to hold back some full-text editorial material for their loyal subscribers, and websites may be “free” because they are heavily sponsored. Systematic Cochrane-type reviews are readily accessible on the Internet, and are great if one needs to quantify evidence, but they can be a dull read. Tonkin’s book very effectively uses the classic “Review article” format in an ambitious attempt to review current knowledge on atherosclerosis and heart disease. The book consists of 15 chapters, covering the pathophysiology of atherosclerosis, each of the major coronary risk factors, the scientific basis of management of the main clinical coronary syndromes and the complications of coronary heart disease, including cardiac failure and arrhythmia. Tonkin is Director of Health and Scientific Affairs at the National Heart Foundation of Australia and he has selected some of the world’s most respected and influential practitioners in cardiovascular research to present authoritative and balanced overviews. Each is a very useful summary of current knowledge. There is some inconsistency in the format, with some chapters omitting detailed referencing and recommending “further reading”, while others are completely referenced, but this does not detract from the information presented. Clinicians and clinical researchers who want rapid access to balanced and reliable information on the commonest cause of premature death in this country will appreciate this book. For a compact and highly readable review of all the current work in atherosclerotic heart disease between two covers, it achieves its aim. Peter L ThompsonCardiologist and Clinical Professor of Medicine and Population Health University of Western Australia Perth, WA

Peter L Thompson

Tarnished gold Colonoscopy, as the gold standard for detecting colorectal neoplasia, has lost some of its glitter, according to a US expert.1 Lieberman was commenting on a multicentre trial in the US, which found that, compared with same-day, state-of-the-art 3-D virtual colonoscopy, optical colonoscopy missed 55 of 511 polyps detected in a series of 1233 asymptomatic patients; 21 of the polyps measured at least 6mm, one being an adenocarcinoma.2 The adenomas missed were usually on the proximal side of a colonic fold or near the anal verge. Virtual colonoscopy also missed adenomas detected by optical colonoscopy, underscoring the complementary nature of the two techniques. Nevertheless, Lieberman says optical colonoscopy remains the pre-eminent method for diagnosing and treating colonic neoplasia. 1 Ann Intern Med 2004; 141: 401-403 2 Ann Intern Med 2004; 141: 352-359 Human versus H. pylori Most people infected with Helicobacter pylori stay symptom-free because of the natural antibiotic activity of a glycoprotein in human mucin, according to a team of Japanese and American researchers. In-vitro studies by Kawakubo and colleagues showed that mucin-type glycoproteins, with unique α1,4-N-acetylglucosamine residues, suppressed the growth and mobility of H. pylori by interfering with the organism’s ability to form a key cell-wall component. The researchers hope their work will lead to the development of novel and potentially safe therapeutic agents to prevent and treat H. pylori infection. Science 2004; 305: 1003-1006 Double trouble When managing women with chronic pelvic pain, doctors may be able to diagnose and treat concurrent irritable-bowel syndrome (IBS), say US researchers. They studied 987 women, newly referred to a pelvic pain clinic, who had all had self-reported pelvic pain for more than 6 months. About one in three of these women also had IBS, as defined by Rome I criteria. Women who were older than 40 years of age, and with muscular back pain, depression and/or a history of adult physical abuse, were more likely to have IBS as well as chronic pelvic pain, indicating that these two conditions may not be manifestations of the same disorder. Obstet Gynecol 2004; 104: 452-458 Stop yoghurt for thrush? Australian researchers have found that Lactobacillus preparations may not be effective in preventing post-antibiotic vulvovaginitis. In a placebo-controlled trial, Pirotta and colleagues randomised 278 women aged 18 to 50 years who were to commence short-term antibiotic treatment for non-gynaecological infections to also receive lactobacillus orally or vaginally, or both, or neither. Overall, 23% of the women developed post-antibiotic vulvovaginitis and lactobacillus was ineffective in prevention. BMJ 2004; 329: 548-551 Suffer the little children . . . When a child is dying, is it OK for a parent to talk to that child about death? Yes, say Swedish researchers. They surveyed almost all parents who had lost a child to a malignant disease in Sweden from 1992 to 1997, with over 400 responding.1 None of the 147 parents who had talked about death with his or her child regretted it; however, nearly a third of the 258 parents who hadn’t talked about death with their child regretted that they hadn’t had such a conversation. Parents were more likely to talk about death with their child if they were religious, if the child was older, and if they sensed the child was aware of his or her imminent death — in an accompanying editorial, a US paediatric oncologist said he believed most dying children are aware of their impending death.2 1 N Engl J Med 2004; 351: 1175-1186 2 N Engl J Med 2004; 351: 1251-1253 INTERHEART Increased consumption of fruit and vegetables, moderate activity, and smoking avoidance (together with lipid lowering) should make up the cornerstones of coronary heart disease prevention in all populations worldwide, according to the INTERHEART Study Investigators.1 This large case-control study, involving about 30 000 subjects in 52 countries from all over the world (including Australia), has determined that these three measures could lead to about an 80% lower relative risk for myocardial infarction. Among other potentially modifiable risk factors for myocardial infarction, the INTERHEART group also found that psychosocial factors, in particular stress at home and at work and financial stress, may play a more important role than commonly recognised.2 1 Lancet 2004; 264: 937-952 2 Lancet 2004; 264: 953-962