Letters

To the Editor: Harper and Ko recently suggested radically changing the way we investigate patients with stable coronary artery disease (CAD), stating that they should initially undergo computed tomography coronary angiography (CTCA), with subsequent stress testing if CTCA shows atherosclerosis.1 To save costs, they recommend excluding stress nuclear myocardial perfusion imaging (MPI) because it has similar accuracy to stress echocardiography and involves ionising radiation. To suggest that the available evidence supports CTCA as the first-line investigation of stable CAD is premature, and likely to cause more therapeutic confusion than it resolves. Of particular concern is the assertion that all patients with any detectable atherosclerosis on CTCA should undergo intensive medical therapy for CAD. The clinical significance of mild atherosclerosis detected on CTCA in patients with chest pain that is unlikely to be of cardiac origin is uncertain, with no clinical studies demonstrating a benefit of medical management for CAD based on CTCA findings. Conversely, were such patients investigated first with a stress test, those with a negative test could be reassured that their chest pain is not cardiac and their intermediate-term prognosis is excellent. Importantly, the directive for unproven lifelong medical therapy for CAD is averted. It is also incorrect to imply that different stress-testing modalities may be seen as equivalent because they have comparable diagnostic accuracies, because their strengths and weaknesses remain complementary. This allows referring physicians to choose the modality that most suits the individual patient. For example, stress echocardiography is more operator-dependent, with known limitations in patients with obesity, airways disease, arrhythmias, poorly controlled hypertension, and contraindications for β-blocker cessation. It also cannot be fused with CTCA images, as myocardial perfusion images can.2 While we agree that measurement of fractional flow reserve (FFR) is an extremely important advance in interventional cardiology, it has the limitation of underestimating ischaemia in patients with small-vessel disease (eg, patients with diabetes).3 We agree that demonstrating ischaemia is the critical component in the decision to perform either percutaneous coronary intervention or coronary artery bypass graft surgery — and until now the largest body of evidence has been accumulated with MPI.4 Moreover, demonstration of regional ischaemia on MPI associated with an angiographic stenosis obviates the need for FFR. Electrocardiographic stress testing alone cannot provide these localising data. The advent of CTCA promises to reshape how we assess and manage CAD in the future, but today we should be guided by the clinical evidence, and be mindful not to put the cart before the horse.

Victor Kalff · Stephen J Duffy · Andrew J Taylor

To the Editor: Harper and Ko have ignored significant facts regarding myocardial perfusion scintigraphy and made unsubstantiated conclusions regarding computed tomography coronary angiography (CTCA) and fractional flow reserve (FFR).1 Their investigation algorithm for suspected and known stable coronary artery disease (CAD) is flawed. By definition, stable CAD patients cannot be categorised into a “suspected CAD” subgroup. CTCA is not an appropriate first-line investigation for patients presenting with atypical chest pain or equivalent syndromes. Being purely anatomical, CTCA cannot determine the functional significance of CAD, rendering it ineffective in intermediate disease. Furthermore, CTCA often overestimates stenoses due to partial voluming and blooming. Guidelines from the United Kingdom National Institute for Health and Clinical Excellence (NICE) state that “questions remain about the ability of multislice CT coronary angiography to accurately identify stenoses of functional significance ... in people with stable chest pain”.2 Harper and Ko’s algorithm will lead to more investigation, provide no differentiation over and above Step 2 (functional imaging), and increase radiation dose, contrast burden and delays between presentation and definitive treatment. The real potential of CTCA lies in its ability to non-invasively assess plaque burden and therefore help rationalise long-term pharmacotherapy in high-risk, asymptomatic patients. However, because there is a lack of evidence for this approach, it has been largely ignored and instead the CTCA community has been “sidetracked” into comparing CTCA with functional imaging. Harper and Ko claim that additional costs from the inevitable increase in CTCA scans would be “minor in comparison to the savings”. This is unproven, and not supported by NICE.2 For atypical symptoms, Harper and Ko recommend “the less expensive ECG [electrocardiogram] exercise test” to risk-stratify patients; however, stress ECG testing is poor at predicting the degree of myocardium at risk, a key factor in risk-stratifying patients. The exercise ECG lacks specificity, leading to unnecessary invasive testing, and many patients are unable to comply. NICE has abandoned exercise ECG testing completely.2 The authors state that a “wealth of data has validated the accuracy of FFR”. The original papers demonstrating a significant FFR of 0.75 were based on perfusion imaging.3,4 The review cited quotes a lower sensitivity and specificity of FFR compared with perfusion scintigraphy, and lists scintigraphy as the “clinical gold standard of ischemia”.5 This is because scintigraphy is the most sensitive technique for identifying functionally significant CAD, as it detects the first abnormality in the ischaemic cascade (reduced myocardial perfusion). Furthermore, the exact level of FFR significance (0.75 v 0.80) is debated. In practice, there is a large degree of operator “discretion” in deciding to revascularise, irrespective of the FFR value, which is often “intermediate” in itself. Harper and Ko question the cost of scintigraphy; however, it is more than $300 cheaper than FFR measurement. They overstate the cost of a standard guidewire by threefold.1 In just one study of over 5000 patients (three times the population of a recent FFR meta-analysis6), increasingly abnormal perfusion scans predicted death and myocardial infarction. There was a reduction in death following revascularisation according to defect severity.7 The nuclear substudy of the COURAGE trial confirmed this.8 No study of CTCA or FFR has been able to predict survival with revascularisation, as the technology is unable to define area at risk. Despite this, Harper and Ko conclude that the Medicare rebate for nuclear testing should be abolished. The authors state that “instituting these changes would undoubtedly result in improved outcomes and substantial savings”. This comment is opinionated, inflammatory and not based on current evidence. The only reasonable conclusion is that before a rebate for CTCA is even proposed, it should be proven diagnostically and prognostically equivalent to myocardial perfusion scintigraphy in predicting ischaemia.

William J van Gaal · Kevin Allman

To the Editor: Although Harper and Ko’s effort at proposing an algorithm1 to reduce unnecessary cardiac catheterisation and stenting is laudable, it is concerning that they advocate investigating patients with stable coronary artery disease (CAD) using computed tomography coronary angiography (CTCA), with an algorithm, furthermore, that has not been validated in terms of patient outcomes. The COURAGE trial demonstrated that percutaneous coronary intervention (PCI) provided no survival advantage or reduction in myocardial infarction over 5 years of follow-up compared with optimal medical management in patients with chronic stable angina.2 PCI provided a relative symptomatic benefit only within the first 3 years. Therefore, it is debatable whether imaging most of these stable patients — much less catheterising them — provides any enduring favourable patient outcomes. For those who do require intervention, fractional flow reserve (FFR) measurement does obviate unnecessary stenting during a cardiac catheterisation. But why perform CTCA — which cannot diagnose ischaemia — in the first place, only to then perform an expensive invasive catheterisation to compensate for it? Why not diagnose ischaemia with an initial non-invasive stress test (nuclear or echocardiography) and only proceed to catheterisation if revascularisation is warranted, thus also lessening the radiation dose? Modern gamma cameras allow nuclear stress studies with radiation doses as low as 2–3 mSv. Yes, CTCA can (as opposed to does) achieve comparable doses with prospective gating, but without the left-ventricular function analysis that nuclear or echocardiography studies provide. Advocating the elimination of the rebate for nuclear stress studies is thus premature, if not ill informed. Harper and Ko’s algorithm relies on the assumption that medical therapy is appropriate in asymptomatic patients for any degree of coronary atherosclerosis, hence justifying the use of CTCA for its detection. However, there is considerable uncertainty about the benefits of medical therapy for primary prevention in low-risk patients.3 Furthermore, the use of CTCA has not been validated as a determinant of appropriate medical management leading to favourable patient outcomes. This contention remains speculative. Indeed, contrary to the claims of CTCA supporters, the United States Centers for Medicare and Medicaid Services has declared that there is uncertainty regarding any potential health benefits or patient management alterations from including coronary CTA [computed tomography angiography] in the diagnostic workup of patients who may have CAD. No adequately powered study has established that improved health outcomes can be causally attributed to coronary CTA ... the body of evidence is of overall limited quality and limited applicability ... in community practice.4 Consequently, the technology also poses significant potential to unnecessarily increase the population radiation burden.5

Peter Karamoskos

To the Editor: Harper and Ko stress the importance of differentiating the presence of atherosclerosis from ischaemia,1 a concept that has been well understood for years. We agree that identifying myocardial ischaemia is important in the management of atherosclerosis; there are a variety of well validated non-invasive tools for this, including exercise electrocardiography, stress echocardiography, and nuclear myocardial perfusion imaging (MPI). However, we have concerns about some of the sweeping changes suggested by the authors — in particular, replacing nuclear MPI with stress echocardiography. While published studies report similar accuracy for these modalities, it is clear that each has strengths and weaknesses. Stress echocardiography is an excellent tool that will often suffice but it is operator-dependent, while MPI is more accurate for single-vessel ischaemia. It is well known that a diagnostic stress echocardiogram cannot be achieved in some patients, so an alternative modality may be required. This is particularly problematic in Australia, where there is no mechanism in place to fund the use of ultrasound contrast imaging. MPI remains a widely accepted technique globally, with a substantial evidence base to support its diagnostic and prognostic impact. For instance, a negative MPI study is generally associated with a cardiac event rate of < 1% in the following year.2 Further, its role in directing the mode of treatment on the basis of extent of ischaemia and its cost-effectiveness as the “gatekeeper” to cardiac catheterisation have been amply demonstrated.3 The use of radiation in medical procedures is always subject to the ALARA (as low as reasonably achievable) principle, and prescribed doses in Australia are typically lower than those quoted in the American literature cited by Harper and Ko. Recent advances in hardware and software with MPI have achieved additional gains in reducing radiation exposure.4 The cost of MPI is greater than stress echocardiography, but remains less than angiography and fractional flow reserve (FFR) measurement. Further, Harper and Ko’s proposed cost savings will not be realised with their call for an increased use of computed tomography coronary angiography and FFR measurement. The job of a good clinician is to choose the right test for the patient; in some cases this is a stress echocardiogram but it is often an MPI study, as supported by international guidelines.5 Thus, readers should be mindful that Harper and Ko’s views about replacing nuclear stress tests with stress echocardiography and abolishing the Medicare item number for nuclear stress tests are personal and may be contrary to sound evidence.

Nathan Better · Myles B Webb · David L Prior · Steven A Unger · George Larcos

In reply: We thank the authors of these letters for their comments on our article. In response to Kalff and colleagues, we firmly believe stress testing should be reserved for the diagnosis of ischaemia in patients with known coronary artery disease (CAD), and computed tomography coronary angiography (CTCA) is the investigation of choice for anatomical diagnosis of CAD in patients with suspected CAD. In this, we are supported by recent guidelines from the American Heart Association and others.1 Most heart attacks occur as a result of plaque rupture in coronary lesions of insufficient severity to cause ischaemia on stress testing, but that would be detected on CTCA.2 In view of the undoubted benefits of medical therapy in both primary3 and secondary prevention of CAD events, do Kalff et al seriously suggest that coronary atherosclerosis detected on CTCA should not be treated with proven anti-atherosclerotic therapy? Regarding comments made by van Gaal and Allman, and Karamoskos, we believe the upfront use of CTCA in our proposed algorithm would substantially decrease downstream referrals for both stress testing and invasive angiography in patients with suspected CAD. The current practice of referrals for invasive angiography based on clinical assessment and functional tests results in a low yield of obstructive disease.4 Furthermore, we believe the responsible use of CTCA can be promoted by restricting its use to the anatomical diagnosis of CAD as a once-only investigation. Once the diagnosis is established, further functional testing can be pursued if clinically indicated. Multicentre studies designed to determine the optimal role of CTCA are underway and will provide important insights regarding our viewpoint.5 Van Gaal and Allman place considerable emphasis on the 5000-patient study of myocardial perfusion imaging.6 We point out, however, that this was an observational study. In contrast, the FAME study on which we base our views of fractional flow reserve (FFR) was a randomised study.7 In addition, we are bemused by their statement that “The review cited quotes a lower sensitivity and specificity of FFR compared with perfusion scintigraphy, and lists scintigraphy as the ‘clinical gold standard of ischaemia’”. To quote directly from the review article: ... scintigraphy, although considered the clinical gold standard of ischemia, has limitations in identifying the hemodynamic significance of individual lesions in patients with multivessel CAD.8 Regarding comments made by Better and colleagues, we believe both stress echocardiography and nuclear stress testing are overused in clinical practice. In most circumstances, a standard electrocardiographic stress test is sufficient to exclude significant ischaemia and thus to determine who should undergo invasive angiography. In circumstances requiring stress imaging, our preference is to perform stress echocardiography, as its accuracy is comparable to nuclear stress testing, with no additional radiation. Notably, in multilesion CAD, none of the stress testing modalities are sufficiently accurate to determine which lesions require intervention.9,10 Only FFR, a proven cost-effective investigation, can decide this.9,10 Rather than putting the cart before the horse, as suggested by Kalff et al, our algorithm provides a horse with the potential to win the Melbourne Cup.

Richard W Harper · Brian S Ko

To the Editor: Handelsman and colleagues raise concerns1 about medical confidentiality with regard to legislation in New South Wales and a recent National Health and Medical Research Council (NHMRC) guideline. We wish to correct some of their statements related to the NHMRC guideline. The NHMRC recently released two documents about the use of genetic testing and genetic information in health care. The first describes the mandatory process by which clinicians can utilise new provisions in federal privacy legislation.2 This guideline was the subject of our editorial3 in the same issue of the Journal as Handelsman et al’s Viewpoint article. The second document is an information paper on medical genetic testing, including reference to disclosure of information, and directs readers to relevant resources.4 Handelsman et al cited the second document, but their concerns pertain to the first. Disclosure of genetic information without consent for the benefit of relatives was considered by the Australian Law Reform Commission (ALRC) in 2001–2003. The Commission consulted widely in preparing its two-volume report on genetics and ethics.5 The Commission recommended amending the Privacy Act 1988 (Cwlth) to allow disclosure in certain circumstances. It is incorrect to suggest that the proposal had the potential to “silently harm individuals who do not form sufficiently clamorous rights-demanding groups”.1 The federal government responded by amending the Privacy Act, and the NHMRC was charged with drafting a mandatory guideline2 for clinicians using the amendment. We were members of the working committee who, together with the NHMRC and the Office of the Australian Information Commissioner, ensured the guideline complied generally with federal privacy legislation and reflected the intent of the ALRC. The guideline was subject to public and targeted consultation — it does not represent an arbitrary standard, as suggested by Handelsman et al.1 The guideline emphasises the importance of confidentiality and the need to balance the rights of patients and genetic relatives. It carefully defines what information can be used or disclosed, and to whom. In contrast to Handelsman et al’s assertion, the legislation does not allow disclosure of a patient’s identity or genetic status. Disclosure is only permitted after extensive, documented consultation. It is incorrect to suggest that disclosure occurs at the “sole discretion” of a doctor or is the “lazy path”. Disclosure without consent represents a major change in the use of medical information, and best practice involves communicating with patients to avoid the very situation that the legislation and guideline address. But we must also recognise that, for each of us, our genetic heritage — and the right to access that heritage — is something that we share with our relatives.

Graeme K Suthers · Elizabeth A McCusker · Samantha A Wake

To the Editor: I write to clarify the National Health and Medical Research Council’s (NHMRC’s) role in supporting the dissemination of the Privacy Act 1988 (Cwlth), which regulates information privacy, and to correct some statements made by Handelsman and colleagues in their Viewpoint article.1 The process of using or disclosing genetic information to a patient’s genetic relatives in circumstances where patient consent is not given is complex. In 2006, the Privacy Act was amended to enable a health service provider to disclose the genetic information of a patient (and not specific information about an individual) to their genetic relatives. This may be done in certain circumstances and in accordance with guidelines2 developed by the NHMRC to assist health practitioners faced with the difficulty of making decisions about use or disclosure. The guidelines, approved by the Privacy Commissioner and then released in December 2009, establish when, by whom and in what manner the use or disclosure of genetic information may take place, with particular reference to the statutory test set out in the National Privacy Principles. That test provides for use and disclosure when there is: a serious threat to the life, health or safety of a genetic relative; and the use or disclosure is necessary to lessen or prevent that threat. The NHMRC agrees that such situations are likely to occur rarely, as patients in the great majority of cases agree to communicate their genetic information to relatives, particularly if there is potential for that information to benefit their relatives’ health. Furthermore, the amendments do not oblige use or disclosure. Rather, they provide the framework for this to occur in certain circumstances. Handelsman et al’s criticism of the guidelines and interpretation that “disavowal of patient confidentiality [is] at a doctor’s sole discretion” suggest they may have confused the guidelines with another NHMRC document titled Medical genetic testing: information for health professionals.3 This document discusses the legislative amendments briefly but aims, more broadly, to provide a source of information for health professionals and to assist them in communicating with patients when genetic testing is being considered. It provides support for the ordering and interpretation of results in the context of clinical decision making. The guidelines, in contrast, explore the complex issues associated with disclosure, and provide practical tools, including scenarios, key points for good practice, and flowcharts to assist in decision making. Disclosure is only permitted if it is in accordance with the guidelines, and is likely to be a lengthy and involved process.

Warwick P Anderson

To the Editor: Suthers and colleagues raise a number of significant ethical issues regarding patients’ right to privacy and the passing of genetic information to relatives.1 Although the National Health and Medical Research Council guidelines relate specifically to private practice and living patients,2 similar dilemmas arise during medicolegal autopsies. Forensic pathologists regularly encounter conditions in deceased individuals that may have implications for surviving family members. Examples include advanced coronary artery atherosclerosis in a young individual, raising the possibility of a familial dyslipidaemic syndrome; pulmonary thromboembolism, suggesting a hereditary thrombophilia;3 and haemochromatosis. However, there are no national guidelines, let alone nationally consistent laws, for the appropriate handling of such information. It would seem obvious that a deceased person’s relatives should have access to information of possible significance to their health. If a family wishes to receive such information, then mechanisms to pass it on should exist. These might range from contact with the forensic pathology service itself, to referring the family to a general practitioner or physician who has been briefed by the service. Whoever takes on this responsibility should be aware of the subtleties and pitfalls of genetic counselling. For example, the relevant condition may be an incidental finding rather than what caused the death (eg, haemochromatosis in a person killed in a road accident). The issues concerning the identification of, and appropriate medical response to, serious familial conditions at autopsy involve: accurately establishing the potential significance of postmortem findings;4 determining whether the family wish to be made aware of this information; disclosing the information to family members when a patient is deceased and not able to give consent; and deciding on the mechanisms by which this should be done. A complicating factor is the role of the coroner, as autopsies and subsequent reports are performed and written under coronial, not medical, authority. However, the authors of these reports are first and foremost doctors, and many feel a therapeutic obligation to the families of those whose autopsies they have performed. Many of these pathologists feel that the autopsy is indeed for the family — even if it is formally for the coroner and the public interest. However, it is no part of the coroner’s legislated responsibility to look after the health of surviving family members. It would therefore be a useful service to accommodate these issues in future iterations of the national guidelines.2

Roger W Byard · Stephen M Cordner

To the Editor: In Natural history of infectious disease,1 Macfarlane Burnet emphasised that poliomyelitis was the one infective disease of advanced countries that had failed to respond to improvements in the standard of living, leading to epidemics in the first half of the 20th century that increasingly involved young adults, who developed severe paralyses. There was also an abnormally high death rate among young adults. However, in tropical countries where the poliovirus was endemic, the virus spread among infants “with a minimal number of paralytic cases”. Burnet therefore defined a “safe” period in the first year of life, when infection was not likely to lead to paralysis. Burnet also described the features of the “so-called ‘virgin soil’ epidemics” that occurred in remote islands and Arctic communities. In these places, the population had not been exposed to the virus in infancy, but the “brunt of the disease was borne ... by adolescents and young adults”. In an Eskimo community, the paralysis rate was 40% and the mortality rate 14% of the whole population.1 When, in 1988, the World Health Organization decided to eliminate the poliovirus using the oral, live attenuated virus Sabin vaccine, it apparently ignored Burnet’s observations, concluding that polio was in fact a serious disease in tropical countries. As a result of the vaccination program, the wild virus was eliminated from Africa, except in Nigeria.2 However, re-infection has occurred in several African countries, leading to epidemics since 1993 in which young adults have become severely paralysed and the mortality rate has been high,3 because the population’s first contact with the wild virus has no longer occurred in infancy. Burnet predicted that young adults aged between 15 and 25 years would be the main group affected in any polio epidemic.1 He based his observations on the natural history of the disease. The current approach to the epidemiology of polio infection has, however, relied on mathematical models and molecular biology. Both of these approaches have failed to predict that epidemics would occur.4,5 In fact, Anderson and May claimed that the decision to advocate vaccination was justified — that “programmes of mass immunization against poliomyelitis in developing countries are unlikely to do harm”.4 Nearly 40 years earlier, Burnet wrote: Any escape into circulation that was not immediately dealt with could grow into the almost unimaginable catastrophe of a ‘virgin-soil’ epidemic of poliomyelitis involving all the populous regions of the world. [...] The great epidemics represented the entry of virulent strains into populations containing a high proportion of older children and adults who had escaped natural immunisation infection.6 The loss of naturally acquired immunity to the wild virus as the result of oral vaccination could have disastrous results. While it is now too late to stop vaccination, it should be extended to young adults as well as children, because current epidemics are affecting this age group, as Burnet predicted.1

Colin L Crawford

To the Editor: Grills eloquently informs readers on the dangers of the Easter bunny and chocolate, and reminds us that Easter eggs may pose a choking hazard.1 It is worth noting that the foil used to wrap Easter eggs also poses a health risk. This was demonstrated in the case of a 20-year-old woman who presented to hospital with respiratory distress after inhaling foil wrapping from an Easter egg.2 The foil wrapper had been pushed inside the woman’s salbutamol inhaler by her 2-year-old son and subsequently occluded her left lower lobe bronchus after she used her inhaler.

Mark Naunton

To the Editor: A recent study of adult bodyweights recorded in 199 countries showed an increase in mean body mass index (BMI) of 0.4 kg/m2 per decade between 1980 and 2008.1 The increase in Australia was even greater: 0.9 kg/m2 for men and 1.2 kg/m2 for women.1 This increase in BMI, sometimes referred to as an “obesity epidemic”, has also been reported in coronial cases in South Australia, with the percentage of morbidly obese individuals (BMI > 40 kg/m2) undergoing autopsy rising from 1.3% to 4.8% between 1986 and 2006.2 While attention is often focused on complications of obesity such as diabetes mellitus, hypertension and cardiac disease, there is also a link to pulmonary thromboembolism (PTE). A recent prospective forensic autopsy series found that 24 of 32 cases of fatal PTE were in individuals who were overweight or obese.3 To examine this association further, we performed two analyses. First, we recorded the numbers of sudden deaths subject to coronial autopsies in SA from 2000 to 2009 (inclusive) where the cause of death was PTE. These were divided into two periods (2000–2004 and 2005–2009) and expressed as the percentage of the total number of coronial autopsies. Second, we analysed all deaths due to PTE that were the subject of coronial autopsies in SA in 2008 (the most recent year for which complete autopsy data were available at the time that the study was initiated). Age, sex and BMI were recorded, and the group was compared with an age- and sex-matched control group taken from other coronial autopsy cases. Despite considerable yearly fluctuations, the death rate due to PTE showed a small increase over time (Box). The rate of deaths due to PTE per total number of autopsies increased from 3.59% (198/5521) in 2000–2004 to 3.71% (239/6442) in 2005–2009. In 2008, there were 44 individuals (21 men, 23 women) with PTE as cause of death. The age range of this group and the 44 matching controls was 27–93 years (mean, 67 years). The mean BMI for individuals dying of PTE (29.6 kg/m2; range, 19.0–51.8 kg/m2) was significantly higher than for the controls (25.9 kg/m2; range, 17.7–44.6 kg/m2) (Student t test, P < 0.05). If PTE is conclusively found to be associated with BMI, there may be significant effects on hospital diagnostic services, coronial forensic investigation units and public expenditure in the future. Coronial cases in South Australia where death was due to pulmonary thromboembolism, 2000–2009

Roger W Byard · Hannah Rosenfeld

To the Editor: Colchicine is an alternative to anti-inflammatory agents for the treatment of acute gout.1 It is potentially toxic and can cause multiorgan dysfunction, including hepatotoxicity, neutropenia, acute kidney injury, rhabdomyolysis and hypotension. We report three cases of toxicity from supratherapeutic dosing that occurred within 2 weeks of each other. In each case, no alternative diagnoses (in particular, sepsis) were considered likely. An 87-year-old man presented with vomiting after 3 days of profuse diarrhoea, nausea and generalised myalgia without other infective symptoms. He had been prescribed 0.5 mg colchicine daily, but he was taking 1.0 mg four times daily. Clinical features included generalised abdominal tenderness, pancytopenia, rhabdomyolysis and hepatic dysfunction. These resolved over 5 days with supportive treatment. A 67-year-old woman presented with lethargy, diaphoresis, fever, nausea, vomiting and diarrhoea. She had been prescribed 1.0 mg colchicine initially, then 0.5 mg every 6 hours “until diarrhoea develops”. She was taking colchicine every hour despite diarrhoea. Initial clinical features included tachycardia, hypotension, tachypnoea and fever. Multiorgan dysfunction developed rapidly and despite aggressive resuscitation, including an intra-aortic balloon pump and broad-spectrum antibiotics, she died within 24 hours. A 77-year-old woman presented with nausea, vomiting and profuse diarrhoea. She had been prescribed 1.0 mg colchicine three times daily “until diarrhoea develops”. On presentation, she was dehydrated but haemodynamically stable, with mild renal and hepatic dysfunction and creatine kinase level elevation. These resolved over 4 days with supportive treatment. A recent randomised controlled trial demonstrated that low-dose colchicine (1.8 mg over 1 hour) was as effective as a higher dose (4.8 mg over 6 hours). Further, the adverse effects of the lower dose were similar to placebo and significantly less than those of the higher dose.2 Adapting to the Australian formulation, 1.0 mg initially and 0.5 mg 1 hour later is now recommended.3 In contrast, repeat doses of 0.5 mg until clinical improvement or side effects (in particular, diarrhoea) was previously recommended. In the cases above, the apparent dosing regimen differed significantly from current recommendations. Ineffective patient education may have also contributed to the conditions of patients 1 and 2, who up-titrated their dose. Patients 2 and 3 persisted with dosing despite gastrointestinal symptoms. Health professionals should be aware of the new dosing recommendations. These are highlighted by the National Prescribing Service4 and the Australian medicines handbook.3 Patients with renal or hepatic impairment, or concomitant use of cytochrome P450 3A4 or P-glycoprotein inhibitors (eg, clindamycin), are at increased risk.5

Myles W H Smith · Darren M Roberts · Siobhann M Ritson · Richard O Day

To the Editor: Warfarin is an orally administered vitamin K antagonist and has many well described interactions with commonly prescribed medications.1 However, only a handful of case reports worldwide have shown that flucloxacillin, a widely used antibiotic, may affect warfarin therapy by reducing its anticoagulant effect — in one instance contributing to the development of ischaemic stroke.2,3 There are several reports describing potential mech-anisms of interaction for flucloxacillin with warfarin.4,5 We describe the case of a 64-year-old man who was diagnosed with infective endocarditis on the basis of mitral valve vegetations and blood cultures that were positive for methicillin-sensitive Staphylococcus aureus. The patient underwent mitral valve replacement with a prosthetic valve and received prolonged antibiotic therapy with intravenous flucloxacillin for 8 weeks. Initially, for nearly 4 weeks, therapeutic anticoagulation with warfarin was difficult to achieve with warfarin doses ranging between 5 and 10 mg per day (target international normalised ratio [INR], 2.5–3.5). During this time, the patient was receiving intravenous flucloxacillin at a total daily dose of 8 g (2 g every 6 h). Flucloxacillin was increased to 12 g per day (2 g every 4 h) and there was a subsequent fall in the patient’s INR, which was sustained despite increasing the dose of warfarin to 25 mg per day (Box). Completion of flucloxacillin therapy was associated with a rise in the patient’s INR and a reduction in his warfarin dose requirement. No other relevant medications were commenced during this period of observation. This case highlights the potential for a dose-related effect of intravenous high-dose flucloxacillin (up to 12 g per day) on warfarin metabolism and the need for vigilance when prescribing antibiotics in conjunction with oral vitamin K antagonists. Other penicillins such as amoxicillin can, conversely, enhance the anticoagulant effects of warfarin, so the warfarin–flucloxacillin interaction may not be intuitive for many clinicians. Doses* for oral warfarin and intravenous flucloxacillin administered to a 64-year-old man with infective endocarditis, and corresponding INR INR = international normalised ratio. * Graph shows abbreviated, not daily, data. Intervals are weekly up to Week 9, then vary according to when INR was measured.

Philip Y-I Choi · Katherine L Phillips · Ian Rae

To the Editor: We read with interest the article by Ng and colleagues on life-threatening hypokalaemia associated with ibuprofen-induced renal tubular acidosis,1 and wish to present our own experience of four patients presenting to our hospital over a year (Box). The patients all presented with biochemical signs of renal tubular acidosis with severe hypokalaemia and a normal anion gap metabolic acidosis from long-standing misuse of ibuprofen taken in combination with codeine from over-the-counter (OTC) medications. Patients 1 and 2 presented acutely with deliberate misuse that included an ibuprofen–codeine combination product. Both patients subsequently admitted to long-standing misuse of ibuprofen and codeine taken in combination. Patients 3 and 4 presented with constitutional symptoms and generalised weakness with a history of taking large amounts of an ibuprofen–codeine combination product. Both these patients required intensive care unit admission for central venous access and potassium replacement. As in the case series by Ng and colleagues, there was no history to suggest gastrointestinal loss of potassium, and medication histories were negative for drugs known to cause intracellular potassium movement or potassium wasting (eg, diuretics). Ibuprofen cessation, potassium replacement and supportive care resulted in biochemical recovery in all four patients. Opioid addiction appears to be the common thread reported by Ng et al and in our case series. Other case reports support this.2,3 Paracetamol taken in supratherapeutic doses is known to cause hepatotoxicity, and it appears that patients with opioid addiction may now be turning to ibuprofen–codeine combination products. More evidence of the danger of these products comes from a case series reporting 27 patients with ibuprofen–codeine misuse that resulted in significant morbidity, including presentations for opioid dependence, gastrointestinal haemorrhage, hypokalaemia, anaemia and/or renal failure.4 In Australia, ibuprofen–codeine combination products are available OTC, albeit in restricted amounts due to problems related to codeine misuse.1 Further restrictions may need to be considered in light of the significant morbidity related to the ibuprofen component. Baseline laboratory investigations and other characteristics of four patients with ibuprofen-induced renal tubular acidosis* RR Patient 1 Patient 2 Patient 3 Patient 4 Sex, age in years Female, 35 Male, 55 Male, 41 Female, 39 Ibuprofen dose† Unclear, years’ duration 9.0–18.0 g/day 5.0 g/day 8.0 g/day Other medications Amitriptyline 50 mg at night Esomeprazole 40 mg daily Multiple medications Nil Serum pH 7.35–7.45 7.29 7.13 7.26 7.32 Pco2, mmHg 35–45 45 42 30 28 HCO3-, mmol/L 22–32 21 13 13 14 Anion gap, mmol/L 7–17 3 9 12 11 Na+, mmol/L 136–146 137 139 142 135 Cl-, mmol/L 98–106 116 120 120 111 Urea, mmol/L 3.0–8.0 4.7 4.4 3.0 6.9 Creatinine, μmol/L 60–120 70 123 125 99 K+ on presentation, mmol/L 3.5–5.0 2.8 2.9 2.5 1.4 K+ on discharge, mmol/L 3.5–5.0 3.5 3.8 3.7 4.5 RR = reference range. Pco2 = partial pressure of carbon dioxide. HCO3- = bicarbonate ion. Na+ = sodium ion. Cl- = chloride ion. K+ = potassium ion. * Same format as used in Ng et al case series1 to allow direct comparison. † Maximum recommended: 3.2 g/day.

Colin B Page · Paul A Wilson · Aidan Foy · Michael A Downes · Ian M Whyte · Geoffrey K Isbister

To the Editor: We have previously argued that the recent rapid increase in the popularity of low-carbohydrate (“low-carb”) beers, in Australia and other countries, is more a community health risk than a healthy alternative to traditional beers.1 This contention has since been supported by a survey conducted by the Victorian Health Promotion Foundation (VicHealth), which found that “low carbohydrate beer drinkers mistakenly believe these beverages are a healthier choice than other varieties”.2 Seventy-one per cent of respondents believed that low-carbohydrate beer is healthier than full-carbohydrate beer, despite having the same alcohol content. Alarmingly, 15% of respondents indicated that they consume more beer when drinking low-carbohydrate beer because they believe it is healthier than full-carbohydrate beer. The potentially insidious marketing of health benefits for alcohol products has recently been followed by the release of lower-alcohol, lower-calorie wines such as the McWilliam’s Balance range, Cockatoo Ridge’s Low Calorie Brut Cuvée, Beringer Blass’s White Lie, and the JMB Beverages Brightlite range. These are represented as containing a “lower” rather than “low” alcohol content because, at between 6.5% and 9.5% alcohol by volume, these wines clearly contain a far higher alcohol content than the ≤ 1.15% alcohol by volume that is required by Australian food standards to be represented as a low-alcohol product.4 The health-based marketing of these wines is similar to that of low-carb beer — it implies that consuming these products is healthier than consuming traditional versions. This implied health benefit message is reinforced by endorsement of McWilliam’s Balance wines by Weight Watchers and inclusion of these wines in the Weight Watchers diet program, which is followed by over 1.8 million Australians annually.4 According to its manufacturers, “McWilliam’s Balance is destined for incredible consumer demand”.4 McWilliam’s Balance wines contain about one-third less alcohol and one-third fewer kilojoules than regular wines.4 If these wines are being consumed to replace regular wine consumption in the same quantity, they could offer a community health benefit. If they are consumed instead of soft drinks or water in the belief that they are healthier than regular wines, or consumed in larger quantities than regular wine in the belief that they are healthier, they could represent a community health threat. Presently, the Australian Government is considering making alcohol companies display nutritional information and ingredients on all beer, wine and spirits labels as a result of a submission by the Alcohol and other Drugs Council of Australia.5 Governments need to modify food regulations to help make the message more explicit: lower-alcohol, lower-calorie wines are not a licence to drink to your own health.

Stephen P McKenzie · Evie R Leslie · Peter G Miller

To the Editor: I refer to the article by Cameron and Cooke,1 and the letter from Forero and colleagues.2 Cameron and Cooke correctly identified key problems associated with using a time-based process standard as a solution to poor patient outcome linked to overcrowding in emergency departments (EDs). They highlighted the need for monitoring to ensure “patient safety and quality of care are not compromised at any stage of the emergency pathway”. Forero et al challenged their assertion that the United Kingdom has replaced the 4-hour rule, and suggested there are three new time-based measures used as indicators of performance. Unfortunately, Forero et al have not noted the critical difference in new UK policies, in which no specific time cut-off is used as the benchmark for good performance.3 While acknowledging the importance of timeliness of care, “clinical outcomes and the experience of the patient” will be the focus of emergency care in the UK under the new Department of Health initiative effective April 2011.4 Surely Andrew Lansley’s (UK Secretary of State for Health) statement that “the four hour standard should be abolished”5 must mean nothing else but complete abandonment of this standard!. Concerns regarding a time-based performance indicator (and funding) were previously illustrated by Nocera in this Journal, who showed that numerical key performance has been a primary motivation for data fraud in Victoria and New South Wales, relating to ED waiting time.5 Furthermore, there is a possibility of admitting an unprepared patient to a ward at 4 hours,6 without acceptance of care by specialist team. A patient not properly worked up and sent to ward will not have the same automatic privileged access to critical care staff ratios or fast-tracked requests for pathology or radiology investigation provided in the ED. This can contribute to the excessive length of stay and, indirectly, hospital bed block. I caution against overly enthusiastic support for the 4-hour rule being instituted in Australian hospitals as a panacea to access block in the ED, and stress the need for better discharge planning, experienced patient flow and bed management, and improved hospital staff rostering for after hours. The latter include rethinking of ED doctors’ 24-hour staffing, with ED specialists attending to patients primarily,7 and junior doctors as assistants (if a 4-hour standard is to be attained), especially when facing an imminent glut of junior doctors, who will require close supervision, rotating through the ED.8

Shyan L Goh

To the Editor: We read with interest the recent cover article by Bhonagiri and colleagues detailing increased standardised mortality rates for patients admitted to Australian intensive care units (ICUs) out of hours.1 Have the authors considered a secondary analysis controlling for night-time staffing ratios? Aside from staffing levels, human factors such as the effects of fatigue and sleepiness due to circadian rhythm misalignment and sleep deprivation should be considered and further investigated. Studies have consistently shown that human cognitive and motor performance is substantially worse at night than during the day,2 due to the influence of the endogenous circadian clock. A recent study of emergency medicine registrars at an Australian tertiary hospital showed a 21% decrease in performance at night, in clinical scenarios of fellowship examination standard.3 Sleep during the day is of poorer quality and shorter duration than sleep during the night,4 resulting in chronic sleep restriction that exacerbates the impairment occurring in night work.5 Consecutive night shifts are associated with an increasing risk of accidents and injuries.6 In doctors, including consultants, inadequate sleep (fewer than 6 hours) is associated with harmful effects on patient outcomes.7 Despite our cultural beliefs, doctors are likely to suffer from the same biological effects of sleep loss and circadian rhythm disturbance as suffered by other humans. ICUs may provide one of the best hospital environments in which to study medical error, because of defined and limited personnel compared with non-ICU ward areas. ICUs also usually have better record-keeping and error-recognition systems, because of higher staff:patient ratios. Possible risk mitigation strategies include increasing the number of night staff, implementing oversight mechanisms for specified procedures and decisions, and ensuring that rosters and shift lengths are designed according to evidence linking sleep disturbance with poor patient outcomes. Such evidence should form the basis of any regulatory frameworks.

Dev A S Kevat · Andrew R Davies · Peter A Cameron · Shantha M W Rajaratnam

To the Editor: Frei and colleagues recently drew our attention to combination analgesic misuse-related morbidity.1 The same phenomenon has also been reported in New Zealand.2 About 50 years ago, analgesic misuse was widespread in Australia and commonly involved chronic, excessive use of combination analgesics (including the aspirin–phenacetin–caffeine [APC] products, Bex and Vincent’s Powders). After many years, some people who used APC developed “analgesic nephropathy”, which made up 12%–15% of dialysis cases.3 I recently performed a retrospective chart review of patients who were referred to the Drug and Alcohol Services at the Western Hospital (Melbourne) for excessive compound analgesic use between September 2005 and September 2010. There were 32 patients (18% of all referrals; median age, 38 years; 23 were women). All had some form of chronic pain, had initiated compound analgesic use for acute pain (eg, headache) and all described progressive use of analgesics because of psychogenic effects (eg, “gave me energy”, “helped me forget”). All 32 patients were diagnosed with opioid dependence and had medical and psychiatric problems correlating with their compound analgesic misuse. One patient, a 34-year-old man, reported taking more than 70 codeine–ibuprofen tablets daily and sustained recurrent gastric ulceration, which eventually required surgery. Despite this, he continued to misuse the analgesics until he undertook opioid replacement pharmacotherapy. A 24-year-old man misusing the same analgesic, despite completing a detoxification program, also relapsed and died after bleeding from gastric ulceration.4 Overall, the patient profiles were remarkably similar to those described by Frei and colleagues.1 Combination analgesic misuse appears largely correlated with products containing drugs of dependence (eg, codeine) and the phenomenon of “rebound pain” (ie, pain that recurs after a short-acting analgesic effect wanes, or “medication overuse headache”). Most morbidity and mortality risks associated with combination analgesic misuse are a consequence of chronic overdose of the non-steroidal anti-inflammatory drug and/or paracetamol components. Paracetamol (mostly when in combination with an opioid analgesic) is reported as the commonest cause of acute liver failure in the United States and United Kingdom.5 Another long-term complication can be hearing loss.6 Two patients in my clinic group had hearing loss, and the ear, nose and throat specialist’s opinion was that it was related to analgesic misuse. Dextropropoxyphene–paracetamol combination products are still available in Australia but are no longer available in the UK. I question the need for opioids in combination analgesic products and, if used, they should be restricted to prescription.

Michael A McDonough

To the Editor: We read with interest the recent letter by He and colleagues.1 The authors described a case of ischaemic stroke in a patient with scurvy and considered whether there was a connection between the two conditions. They referred to evidence that vitamin C deficiency may be a risk factor for cerebrovascular disease, but acknowledged that a direct causal link is unlikely to be established. We propose adiponectin as the causal link between vitamin C deficiency and stroke. Adiponectin is an adipokine, secreted in multimers by adipose tissue, with insulin-sensitising, antiatherogenic and cardioprotective properties.2 A decrease in adiponectin levels — particularly the more biologically active, high molecular weight (HMW) multimers — is implicated in a number of disease states, such as obesity, type 2 diabetes, heart disease and some cancers. Adiponectin has also been reported to have cerebroprotective properties,3 and there is some evidence that levels may be reduced in patients with cerebrovascular disease.4 A recent study demonstrated that vitamin C supplementation increases the proportion of HMW adiponectin secreted from human adipocytes.5 Vitamin C levels are lower in obese and diabetic patients, and the patient reported by He et al had also been recently diagnosed with type 2 diabetes. While scurvy is admittedly rare, He and colleagues noted that subclinical vitamin C deficiency is not uncommon in the general population (about 10%). We suggest that it would be worthwhile to assay plasma vitamin C levels and total and HMW adiponectin levels in patients presenting with cerebrovascular events, and to undertake prospective studies to determine whether vitamin C supplementation improves patients’ adiponectin levels. Increased HMW adiponectin levels could be explained, at least in part, by the role of vitamin C in the multimerisation of adiponectin. Adiponectin levels are also reduced in other vascular conditions, such as ischaemic heart disease and peripheral vascular disease.4 As plasma vitamin C levels of patients with these diseases are also reduced, it is tempting to posit that vitamin C supplementation could have a role as a treatment, or even a prophylactic, in populations at risk of a range of adiponectin-related vascular diseases.

Felicity J Rose · Jonathan P Whitehead

To the Editor: The juxtaposition of articles by Braithwaite and colleagues1 and Ho and Maddern2 (a description of a trial of physician assistants [PAs] in Adelaide) provides an interesting contrast between what should be happening in health care and what is happening. While Braithwaite and colleagues argue for a value-driven change, the type of change that is occurring at the moment is that of role substitution. This is exemplified by Ho and Maddern’s assertion that “PAs ... would ... improve the quality and quantity of medical service”. The intent is to provide a substitute doctor. This is referred to as “task” substitution or transference, with the inference that those under discussion, whether PAs or nurse practitioners (NPs), will only perform a limited role — some of the tasks of a doctor. In South Australia, PAs are able to order investigations and prescribe. The former editor of this journal, Martin Van Der Weyden, discussed this issue in 2008,3 labelling it “doctor displacement”, a term that Brooks and Mitchell disputed.4 They said “[PAs] will never be able to practise independently; they will always have to practise under supervision and within the scope of practice of their supervisor”. The Health Legislation Amendment (Midwives and Nurse Practitioners) Act 2010 (Cwlth), however, provided NPs with access to the Pharmaceutical Benefits Scheme and the ability to prescribe. Any supervision by doctors seems nominal, with independent practices being set up, some attached to pharmacies. There is less supervision of NPs by the medical profession than that given to trainees in teaching hospitals, and they are effectively independent practitioners. It could be argued that United States-trained PAs are even better trained than Australian-trained NPs, and Brooks and Mitchell’s statement that they will “never” practise alone is hardly reassuring. Proponents of NPs and PAs justify their use4 by reference to a shortage of health professionals. It isn’t reasonable to take from one area of need (nursing) to bolster another area of need (medicine), and the rapid increase in medical undergraduates and increase in immigration of overseas-trained doctors renders medical workforce projections unreliable. Role substitution by PAs and NPs will only perpetuate the problems we now face, while introducing new problems related to limited training and skills and fragmentation of care. Australian GPs, traditionally well trained and highly skilled, and with the ability to manage complex problems, are the core of our health system. Proposed changes devalue their contribution by making the inevitably fatal assumption that the sick patient is easy to recognise and triage.

Patrick S Hanrahan

To the Editor: Eight Australian medical deans recently called for academic health science centres (AHSCs) — where a leading university joins with a major tertiary health care provider in a tripartite mission of excellence in clinical service, research and education.1 However, there are obstacles to their proposal due to the misalignment in purpose of universities and teaching hospitals. Why do they call for change, given that many of our public hospitals are already affiliated with universities, have academic units of the university, and are funded by governments to educate doctors? Many private hospitals are also involved in university and college education. The Garling report was commissioned in response to a crisis in the quality of acute care offered in New South Wales public hospitals.2 However, none of its 139 recommendations refer to striving to achieve excellence. While excellence must be the aim of our public hospitals, their immediate response should be patient-centred, ensuring an acceptable minimum standard of care. Other targets, such as research and concentrating limited resources on aiming for excellence in specified areas, could prove to be a distraction from this core purpose. Education was mentioned in the Garling report but in no sense was achieving excellence in clinical service or research implied to be its purpose, unless excellence is to be understood to mean the minimally acceptable level of care. Medical education and research need to fit within a public hospital system that has funding limits and is struggling to deliver its core goal of clinical care. Given the current misalignment of purpose between hospitals and universities, it is no surprise that the combined deans have expressed a tone of dissatisfaction. Mayo Clinic in the United States has a logo of three interlocking shields,3 representing a “tripartite mission of excellence in clinical service, research and education”. Mayo Clinic’s key to achieving excellence — the aspirational aim of the deans’ “world-class AHSCs” — is alignment of purpose of the hospital and the academic institution. Similarly, in Australia, the barriers highlighted by the deans1 were dealt with in the planning of Macquarie University’s medical school and hospital. A common purpose and a greenfield site, where a new culture could be introduced, were a core part of Macquarie’s approach.4 As Macquarie’s Vice Chancellor, Professor Schwartz, was reported to say in 2010: “By owning and running the hospital we can ensure that we put medical teaching and research at the forefront”.5 Macquarie knows that simply owning and running its own hospital does not necessarily assure success, but it is an important step.

Michael K Morgan · Janet D Greeley

To the Editor: The editorial on academic health science centres (AHSCs) in Australia by the deans of medicine in all eight research-intensive universities1 is an important platform for discussing fundamental issues to be considered in developing these centres. The essential basis of an AHSC is a combination of two major initiatives: Seamlessly integrating research done in laboratories and clinical trials with multidisciplinary patient care. Building a centre of innovation where questions are asked based on actual clinical problems and can be resolved by the collective efforts of on-site “knowledge workers”. It is timely and necessary to investigate this option of health care delivery, as politicians negotiate at state and federal levels on appropriate models, and is in keeping with the National Health and Medical Research Council’s objectives in its latest strategic plan for broadening and building Australia’s capacity for research.2 It is crucial that any AHSCs created justify their status. They could be national centres of excellence, funded by an agreed federal and state government contribution model, together with endowments and competitive research funds. They should not merely be cosmetic makeovers for current university hospitals, with their idiosyncratic selection of staff and clinical practice models. Their governance needs to be standardised, based on agreed principles. A competitive national selection process should include identifying submissions that define employment of specific personnel and implementation of interventions to address the challenges of health care delivery in a timely and efficient manner.3 A suggested list of required attributes of staff employed in AHSCs is shown in the Box. A single academic chief executive officer who is a proven educator and researcher and cognisant of local and regional issues of health care delivery can facilitate the development of the centre, with the assistance of similar professionals on the board of management. The knowledge workers in an AHSC should include knowledge leaders (who incorporate knowledge into clinical protocols and management plans), creators (who generate new knowledge from research data), users (clinicians and researchers who incorporate therapies and skills into the clinical arena) and learners (students training to be health professionals and researchers). Direct interaction between the financial regulators and leading researchers and clinicians in an AHSC can lead to implementation of state-of the-art clinical practice, with patient care consistently being evaluated so that clinical and financial inefficiencies can be eliminated. The interaction of knowledge workers within the AHSC can result in sharing innovative and ethical translational research with bedside clinical practice. Required attributes of academic health science centre (AHSC) staff Staff employed at AHSCs should have all these attributes: Ability — high-quality knowledge and skills Accountability — responsibility, being up to date with current evidence-based practice Availability — to the community of practice within the AHSC Accessibility — to patients and colleagues Affability — teamwork, cooperation, collaboration Affordability — social responsibility to the community and country Adaptability — dedicated to continuous professional development and ongoing research

David A Kandiah

To the Editor: A middle-aged male night-shift driver experienced dizziness, vomiting, sudden onset of limb weakness, and then collapsed and became delirious. He had no past history of renal impairment or hypertension. At presentation in an emergency department, he was hypertensive (blood pressure, 220/140 mmHg) and hyperglycaemic. An upper motor neurone pattern of persistent dysarthria and left hemiparesis was evident. An electrocardiograph and transthoracic echocardiogram confirmed sinus rhythm and left ventricular hypertrophy. Magnetic resonance angiography of the brain showed subacute basal ganglia haemorrhage (Box) and an anterior communicating artery aneurysm. Renin levels were elevated at 124.7 mU/L (reference range, 4.5–46.0 mU/L) but renal artery stenosis was absent on ultrasound. Evaluation of an elevated initial normetadrenaline level of 1610 pmol/L (reference range, < 900 pmol/L) with a diagnostic algorithm1 suggested that phaeochromocytoma was unlikely. Hypertension, proteinuria, suboptimal renal function, inactive urine sediment and normal fasting blood glucose suggested hypertensive nephropathy. Antihypertensives were initiated to mitigate the risk of further stroke, nephropathy and cardiomyopathy. Results of an electroencephalogram were normal. Further history revealed that the patient had nicotine dependence with occasional amphetamine and cannabis exposure. Before admission, he had an estimated intake of 1–2.5 bottles of cough mixture episodically (600–1500 mg of pseudoephedrine). A pattern of ingesting excessive amounts of various brands of over-the-counter combination pseudoephedrine and codeine cough mixture developed over 20 years, and he suffered cravings when abstaining. He visited various pharmacies, provided identification, was repeatedly cautioned, but rarely consulted doctors. A Naranjo score2 of 7 indicated a probable adverse drug reaction to pseudoephedrine, resulting in a disabling hypertensive haemorrhagic stroke. During 2.5 months of inpatient rehabilitation, he regained full independence in self-care and mobility. His unstable employment was interdependent with pseudoephedrine misuse. He accepted counselling regarding constructive, non-pharmacological coping strategies for social stressors. Financial, legal and housing difficulties complicated his discharge plan. He planned to return to work as a driver. Pseudoephedrine is associated with a modest elevation in blood pressure3 and hypertensive crises.4 Intracerebral haemorrhage associated with pseudoephedrine usage is anecdotally reported.5 Dependence and chronic psychoses have been reported6,7 but their prevalence is unknown. Night-shift workers, students, recreational or self-medicating users, and athletes are potentially vulnerable to misuse. Project STOP is an initiative of the Pharmacy Guild of Australia (http://www.projectstop.com.au/index.html) to curb diversion of pseudoephedrine-containing medications from retail pharmacies into illegal amphetamines. However, it is not mandatory and the National Drugs and Poisons Schedule Committee has rejected the Pharmacy Guild’s proposal to enhance Project STOP’s capabilities using an electronic messaging tool for pharmacists (NotifyRx, RelayHealth, Atlanta, Ga, USA) to regulate codeine sales (excluding cough mixtures). In contrast, New Zealand authorities now restrict the supply of pseudoephedrine to prescription only. A concerted effort from all relevant parties, including tighter regulation or legislation, is essential for harm minimisation. We urge regulatory bodies to review the effectiveness of current practice. Magnetic resonance angiogram of the patient’s brain showing likely hypertensive right basal ganglia haemorrhage

Ajay Bharatula · Peter W New

To the Editor: Laparoscopic adjustable gastric banding (LAGB) is the most common bariatric surgery performed in Australia, and diabetic patients represent an important target population.1 In one of few trials in type 2 diabetes, 73% of 30 LAGB-treated obese patients reverted to normal glucose tolerance by 2 years, compared with 13% of 30 obese patients randomly assigned to treatment with changes to their lifestyle.2 These Australian data precipitated publicity implying that many people with type 2 diabetes were cured by LAGB.3,4 However, only recently diagnosed patients were recruited. The benefits of LAGB in those with diabetes of long duration (and usually with a substantial loss of pancreatic beta-cell function5) may not be as dramatic. We reviewed the records of eight patients with insulin-treated type 2 diabetes (aged 40–65 years, with a diabetes duration of 4–39 years) referred for LAGB by one of us (T D) to two established bariatric surgical units over 4 years. All eight patients had a diabetes-specific review within 2 months of surgery and at 3 and 6 months postoperatively, and six patients had a further review at 12 months. Perioperative diabetes management was coordinated by the anaesthetist and ward staff, but patients were encouraged to contact their doctor or regional diabetes education unit for advice at any time if necessary. Patients attended the bariatric unit for saline insertion into the band once or twice before their diabetes-specific review at 3 months. No patients developed postoperative complications. The median weight loss was 14% at 6 months and 20% at 12 months (Box). By 3 months, insulin doses had been reduced in all patients and two had stopped using insulin. However, most had increased their insulin doses at 6 months (and the two who had discontinued insulin started using it again) because of unsatisfactory glycaemic control. During the 12 months’ follow-up, only one patient stopped using insulin and remained without its use without an unacceptable glycated haemoglobin (HbA1c) increase. Although our eight patients achieved the same median percentage weight loss at 1 year as trial patients with diabetes of relatively short duration achieved over 2 years,2 insulin use was still required in most cases. When asked, the patients suggested that inappropriate initial insulin reductions were not because they experienced hypoglycaemia, but rather because they expected substantial LAGB-associated metabolic improvement. Perhaps this was encouraged by media coverage3,4 and/or concerns of bariatric unit staff that continuing their insulin therapy might retard their LAGB-associated weight loss. In some cases, problems with insulin adjustment and glycaemia overshadowed the psychological benefits of progressive weight loss. These preliminary observations suggest a need for more data on diabetes management after LAGB, to inform acute treatment modification and to provide a realistic idea of the effects of LAGB on longer term disease outcomes. Changes in body weight, insulin dose and HbA1c after LAGB in eight patients with insulin-treated type 2 diabetes HbA1c = glycated haemoglobin. LAGB = laparoscopic adjustable gastric banding. Open-circle line = patients who discontinued insulin during the first 3 months of follow-up.

Timothy M E Davis · Catherine Coleman

To the Editor: Two key paradigm shifts are occurring with regard to diagnostic imaging services in Australia that may have far-reaching medicolegal and professional consequences for all doctors, have an impact on patient care and, ironically, increase costs to the health budget. First, the federal government announced a 3-year cost-saving trial in the 2009–10 Budget. Since 1 November 2010, the trial allows doctors to request certain “image-only” radiological investigations without an accompanying radiologist report.1 The investigations attract a lower rebate to cover only the technical component of image acquisition. The changes should not alter the cost to patients as most imaging studies, which currently include radiologist reports, are bulk billed. The trial currently involves chest x-rays and certain facial examinations. The government’s presumption is that for these investigations a radiologist’s assessment is not required and referring practitioners will assume full responsibility for the correct identification and interpretation of all primary, secondary and incidental findings. In doing so, referring doctors will also accept all liability for errors. While the incentive for referrers to request these new item numbers is unclear, the increased risk and responsibility borne by them is obvious. Second, in a measure to mimic the National Health System in the United Kingdom, there is advocacy from some levels of health administration to implement role extension for allied health professionals within diagnostic imaging. This would involve the provision of descriptive reports from non-radiologists with no conclusive interpretation, leaving referring doctors to draw their own clinical conclusions. Error rates for these types of reports are not insignificant when compared with a radiologist’s report.2,3 Disposing of a specialist’s evaluation may delay referral or management, and will again transfer responsibility for errors to the referring doctor.4 The false inference underpinning this model is that general medical and specialist training can be fast-tracked or circumvented without any detriment. The most concerning ramifications from these proposed changes are those for standards of patient care. Reduced radiologist input will result in higher error rates and inaccurate diagnoses;5 will remove the option of clinical discussion regarding difficult cases; and may compromise elements of patient care. Ironically, such medical errors often increase downstream burdens for health systems, negating any short-term cost savings. From a medicolegal perspective, undertaking independent interpretation of images and acting on descriptive reports is a high-risk activity and insurance premiums for doctors may ultimately reflect this.

Nicholas I Brown · Lawrence B Josey