Volume 186 Issue 8

A call for better data and focused research Surge capacity has been defined as the “health care system’s ability to manage a sudden or rapidly progressive influx of patients within the currently available resources at a given point in time”.1 This term entered the disaster medicine lexicon after “9/11” 2001, and has become a key feature of health disaster planning since the severe acute respiratory syndrome (SARS) outbreaks, the rising threat of pandemic influenza, and the Madrid and London bombings in 2004 and 2005. Mass trauma events in urban environments create surge problems for hospitals in the first 4 hours, with the majority of casualties likely to arrive at hospital in the first 60–90 minutes.2 Health systems need to be able to meet this increased demand after acute mass casualty events, with an influx of patients who may require “specialised evaluation or intervention”,3 including surgery for penetrating trauma. As the epidemiology of traumatic events will differ depending on the cause, any preparations to manage a mass casualty event need to be flexible enough to deal with a range of disaster types, including infectious disease emergencies and bioterrorism, with their more insidious and later-peaking surge. Meeting the challenges of surge capacity enhancement calls for cooperative action at the complex interface between clinical practice, health systems management, logistics, facilities management and personnel deployment. In this issue of the Journal, Traub et al4 quantify the physical assets needed for surge preparedness of Australasian hospitals and conclude that they do not meet United States surge preparedness benchmarks, currently the only ones available. This is the first published study to assess Australian capacity in this area and the need for further research. So what are these US benchmarks and are they an appropriate measure for Australia? The US Health Resources and Services Administration has set a minimum benchmark for surge capacity of 500 adult and paediatric patients per million population for infectious disease events and at least 50 patients per million population sustaining trauma or burns in a mass casualty event.5 These figures have not been validated and, based on other traumatic events, 100–300 patients per million may be more accurate for burns or trauma.6 The consensus from a series of articles on the “science of surge” in the November 2006 issue of Academic Emergency Medicine7 was that there is little or no quantifiable measurement for managing an acute “sudden impact” surge or the even less predictable surge resulting from infectious disease emergencies or obscure events. Even the utility of redundant infrastructure has not been formally evaluated and is known to be costly to bring into operation in response to an incident. In addition, the benchmarks don’t consider the need for management systems to be innovative in dealing with a surge situation. Establishing the exact number of required physical facilities fails to give adequate credence to what is a dynamic, non-linear and complex system. Some of the ways of increasing the capability of hospitals to respond to a local disaster include: creating temporary hospital facilities within public hospitals by having caches of medical supplies and equipment available; using alternative health facilities, such as operating theatres, beds and diagnostic resources in private hospitals and day procedure units for the “walking wounded” and “decanted” patients (patients already in hospital who can be transferred elsewhere to make way for disaster victims); using general and private health practitioners to provide a range of care for the walking wounded and decanted patients; and applying modified models of care, such as longer shifts and extended scope of practice for nurses, allied health practitioners, and medical and other health students.8 Health authorities in various Australian states and territories are putting such arrangements in place as part of their surge planning. Further research to quantify the effect these measures may have on hospitals’ response capability, and whether they can prevent critical bottlenecks, is urgently required. The sustainability of the response is also critical. Some events will overwhelm local capabilities. Having national approaches in place, such as the Australian Burn Plan (AUSBURNPLAN),9 can ensure that appropriate care is provided regardless of the site or prolonged nature of the disaster. The Australian Health Protection Committee and its predecessor the Australian Health Disaster Management Policy Committee have been working since February 2003 to put policies and plans in place to ensure that such a response will occur. A detailed review of the work of the Australian Health Disaster Management Policy Committee until the end of 2005 has been published in the MJA.10 The recent development of the Australian Health Management Plan for Pandemic Influenza;11 the National Medical Stockpile;12 the establishment and deployment to Java, Indonesia, of Australian disaster medical assistance teams in May and June 2006 (in response to the Yogyakarta earthquake); and the improved communication and coordination of health disaster responses between individual states and territories and the Australian Government have all enhanced the national capability to respond to a disaster. What remains, however, is the continuing need for further research in disaster medicine, particularly into surge capacity, so that hospitals and health services can better gauge whether the steps they are putting in place will be effective. Historically, disaster medicine has grappled with the difficulties of evidence-based research, particularly in terms of measuring outcomes through randomised controlled trials. However, without better data and focused research, we will continue to “measure the immeasurable”.

Andrew G Robertson CSC, FAFPHM, FRACMA · David M Cooper MMgt, MBA, FACEM

Inhalation is an attractive alternative to injection, but significant issues will limit its use Insulin is an excellent therapeutic agent with few significant side effects, apart from hypoglycaemia. However, it has one major disadvantage — it needs to be given by injection. Patients’ fears or reservations often mean doctors are reluctant to initiate insulin therapy, even when it is clearly indicated. Not surprisingly, many attempts have been made to find an alternate method of insulin delivery, through nasal, oral, buccal, transdermal or inhaled routes. Inhaled insulin is a theoretically attractive approach, mainly because of the large absorptive surface of the lung alveoli with relatively little proteolytic activity to cause insulin degradation, and it is the most advanced alternative method for insulin delivery in development. One such product, Exubera (Pfizer), has already been granted approval for clinical use in the United States and the European Union, and other products (AERx [Novo Nordisk] and AIR [Eli Lilly]) are in advanced stages of clinical trials. The technology required for administering inhaled insulin is more exacting than that for other inhalation medications, such as those used in treating asthma. As the dose–response curve to insulin is a near linear one, it is critical that the right amount of insulin be consistently delivered to the alveoli and blood stream. The size of the inhaled insulin particles is of paramount importance to this, as particles that are too large adhere to the upper airway and are not absorbed, while those too small will be exhaled before absorption. Manufacturers have used different technologies to address this, and clinical studies have shown that variability of inhaled insulin absorption is similar to that for subcutaneous insulin.1 Pharmacokinetic studies have shown inhaled insulin to be similar to the rapid-onset insulin analogues, such as insulin lispro or insulin aspart, but with a slightly longer duration of action. In other words, inhaled insulin is essentially a rapid-acting insulin suitable for control of postprandial hyperglycaemia. More recent clinical studies suggest that inhaled insulin also slightly lowers fasting plasma glucose levels,2 but the clinical significance of this finding has yet to be fully explored. Inhaled insulin has so far been used in combination with injections of long-acting insulin, and this practice will continue, particularly for patients with type 1 diabetes. In this context, inhaled insulin has been shown to be equivalent to subcutaneous insulin in the control of hyperglycaemia in a number of relatively long-term trials (1–2 years) encompassing all major clinical scenarios encountered in type 1 or type 2 diabetes. However, no clinical trials have shown that inhaled insulin is superior to conventional insulin administration in metabolic control. Clinical trials and theoretical considerations have raised a few potential problems with inhaled insulin, including:3 Insulin delivered via the lungs is more antigenic and results in increased development of anti-insulin antibodies, a problem largely eliminated from conventional insulin treatment by the use of purified insulin. However, these antibodies have not been found to cause major problems, such as allergy reactions or insulin resistance, and they are also unlikely to play a dominant role in the evolution of diabetic complications. Long-term delivery of a foreign substance to the respiratory tract can potentially cause damage. There is little doubt that treatment with inhaled insulin is associated with a small (about 5%) decline in lung function, when measured by parameters such as the diffusing capacity of the lung for carbon monoxide, and forced expiratory volume in 1 second.3 Fortunately, it appears that this relatively minor decline is reversible when inhaled insulin treatment is suspended.4 There is a theoretical increased risk of formation or promotion of cancer with chronic use, especially since insulin is a relatively potent growth factor. This potential danger has not eventuated after use of inhaled insulin in thousands of patients, however it is prudent to carefully monitor for this possible side effect. There are also practical issues with use of inhaled insulin. Smoking immediately before inhalation of insulin increases insulin absorption, although passive or chronic smoking, asthma, and chronic airways disease decrease absorption.4-6 Inhaled insulin therapy is not associated with increased respiratory tract infection but often causes a mild, usually transient cough. Upper respiratory tract infections have shown no significant effect on insulin absorption in clinical trials, but tangible experience is relatively limited. These confounders of insulin absorption have been carefully avoided or controlled in clinical trials, but their potential to cause hypoglycaemia or hyperglycaemia in real-world use is obvious. Some problems with inhaled insulin are highlighted by the commercially available Exubera. Due to its formulation and method of delivery, the dosage of Exubera needs to be measured in milligrams rather than the customary units. It is available in capsules of 1 mg (equal to 3 units) or 3 mg (equal to 8 units), and the product information recommends a commencing dosage based on the patient’s body weight rather than the carbohydrate content of meals. To make things even more confusing, as noted above, three 1 mg capsules are more potent than one 3 mg capsule, complicating interchange of capsules of different strengths. These limitations make minor adjustments in insulin dosage more difficult, especially in patients with labile type 1 diabetes who can develop severe hypoglycaemia with insulin dose adjustment of as little as one or two units at a time. For patients with type 2 diabetes with a high insulin requirement, multiple (more time-consuming) inhalations are needed. Most of the devices used to administer inhaled insulin are relatively large and cumbersome, and they require both time and skill to master. These factors will limit their uptake. In times past, routine tasks in the diabetic clinic included checking the glass syringe, usually carried in an alcohol solution in a special container, and inspecting the needle for sharpness. There has been an obvious improvement in the devices used to inject insulin subcutaneously, and it would be unfortunate if this trend were reversed with the use of inhaled insulin. All these practical issues with inhaled insulin will probably be overcome in time, but at the moment they remain significant. In addition, absorption of inhaled insulin is not as efficient as subcutaneous injection. Indeed, only about 10% of the dose is absorbed. This poor absorption and the high cost of development will inevitably make inhaled insulins much more expensive than their subcutaneous counterparts. This is almost certain to limit their use, particularly in Australia where government subsidy of a new pharmaceutical product largely depends on demonstration of its cost-effectiveness in comparison with existing alternative products. Although qualitative surveys have shown that patients given the choice of inhaled or injected insulin overwhelmingly prefer the former,4,7,8 this has not been shown to translate to wider or earlier acceptance of insulin treatment or a consequent lower glycated haemoglobin level for the diabetic community. The need for caution in the use of inhaled insulin in young patients and the relative lack of flexibility in adjusting dosage by a small amount make it more likely to be used as the initial insulin therapy for patients with type 2 rather than type 1 diabetes. However, for most doctors, who all too often see patients refuse insulin injections when they are needed, the availability of inhaled insulin will be a step forward, at least for some of their patients. In summary, we predict that inhaled insulin therapy will become part of our therapeutic armamentarium, albeit with a small niche market. A small proportion of patients with well controlled type 1 diabetes may use the inhaled route exclusively. Others may use inhaled insulin as a supplement, or to have a much needed “injection holiday”. Subcutaneously injected insulin will remain the major route of administration for the foreseeable future. The inhaled insulins currently available will primarily be used by adult patients with type 2 diabetes who have the financial means — and who are willing to pay — to avoid injections. We suggest that the regulatory authorities could make inhaled insulin available to individual patients for a limited time through the Pharmaceutical Benefits Scheme, to allow patients and their doctors to trial inhaled insulin therapy and assess its benefits.

Aidan McElduff MB BS, FRACP, PhD · Dennis K Yue PhD, FRACP

Objectives: To measure physical assets in Australasian hospitals required for the management of mass casualties as a result of terrorism or natural disasters.Design and setting: A cross-sectional survey of Australian and New Zealand hospitals.Participants: All emergency department directors of Australasian College for Emergency Medicine (ACEM)-accredited hospitals, as well as private and non-ACEM accredited emergency departments staffed by ACEM Fellows in metropolitan Sydney.Main outcome measures: Numbers of operating theatres, intensive care unit (ICU) beds and x-ray machines; state of preparedness using benchmarks defined by the Centers for Disease Control and Prevention in the United States.Results: We found that 61%–82% of critically injured patients would not have immediate access to operative care, 34%–70% would have delayed access to an ICU bed, and 42% of the less critically injured would have delayed access to x-ray facilities.Conclusions: Our study demonstrates that physical assets in Australasian public hospitals do not meet US hospital preparedness benchmarks for mass casualty incidents. We recommend national agreement on disaster preparedness benchmarks and periodic publication of hospital performance indicators to enhance disaster preparedness.

Matthias Traub MD, FRACS · David A Bradt MD, FACEM, FAFPHM · Anthony P Joseph MB BS, FACEM

Objective: To assess whether a diagnosis of the metabolic syndrome (MetS) improves the prediction of cardiovascular disease or total mortality beyond that already provided by conventional risk factors.Design and setting: A longitudinal cohort study conducted in Dubbo, New South Wales.Participants: 2805 men and women aged 60 years and older living in the community, first assessed in 1988–1989 and followed for 16 years.Main outcome measures: Coronary heart disease (CHD) events, ischaemic stroke events, and total mortality.Results: MetS was present in 31% of men and 34% of women. Crude CHD, ischaemic stroke, and total mortality rates were higher in the presence of MetS in men and women. In proportional hazards models that included conventional risk factors, but excluded variables used to define the presence of MetS, MetS was a significant predictor of CHD, stroke and total mortality. In men, the respective hazard ratios were 1.64 (95% CI, 1.37–1.96), 1.31 (95% CI, 0.97–1.77), and 1.53 (95% CI, 1.30–1.79). In women, the respective hazard ratios were 1.70 (95% CI, 1.43–2.02), 1.37 (95% CI, 1.04–1.82), and 1.35 (95% CI, 1.15–1.59). The use of MetS variables on an ordinal scale produced broadly similar conclusions.Conclusions: A diagnosis of MetS provides additional prediction of CHD events, stroke events, and total mortality beyond that provided by other conventional risk factors.

Leon A Simons MD, FRACP · Judith Simons MACS · Yechiel Friedlander PhD · John McCallum DPhil

Objectives: To describe the self-reported use of prostate specific antigen (PSA) tests, faecal occult blood tests (FOBTs), and whole-body skin examinations among Queensland men, reasons for use, and the personal characteristics of men undergoing the tests for cancer screening.Setting and design: Data were obtained from the Queensland Cancer Risk Study (QCRS), a population-based telephone survey conducted in 2004, which used random sampling stratified by age, sex, and geographic location.Participants: All men aged 50–75 years who participated in the QCRS (n = 2336).Main outcome measures: Use of PSA test, FOBT, or whole-body skin examination, specifically as a screening procedure; the probability of being screened; and associations with sociodemographic factors, risk behaviour, and cancer experience.Results: More than a third of men reported never having been screened for prostate, colorectal, or skin cancer. Of those who had been screened, the odds of PSA testing being reported were more than two times greater than the odds of whole-body skin examination (adjusted odds ratio [OR], 2.54; 95% CI, 1.49–4.32), and the odds of reporting an FOBT were less (adjusted OR, 0.48; 95% CI, 0.22–1.04). Men who participated in cancer screening tended to be older, white, living with a partner, and to have private health insurance. Smokers were less likely to be screened with any of the three screening tests.Conclusions: Of these three cancer screening tests, the FOBT has the best evidence for reducing mortality and yet is the least frequently used by Queensland men. There are disparities in reported screening prevalence between the specific tests and across certain population subgroups.

Philippe Carrière MD, MSc · Peter Baade PhD, AStat · Beth Newman PhD · Joanne Aitken PhD · Monika Janda PhD

Objective: To examine evolving changes in asthma and chronic obstructive pulmonary disease (COPD) in South Australia and Australia as a whole from the perspective of hospital admissions, ventilatory support and mortality data.Design: Retrospective analyses, for the period 1993–2003, of hospital separations data from the Australian Institute of Health and Welfare and the Integrated South Australian Activity Collection, and mortality data from the Australian Bureau of Statistics and South Australian hospital morbidity collection.Main outcome measures: Hospital separations, ventilatory support episodes, mortality rates, burden-of-disease rankings.Results: Between 1993 and 2003, in SA and nationally, hospital separations for asthma declined but separations for COPD increased significantly. Falling mortality rates from asthma in both men and women, and from COPD in men, contrast with increasing rates of COPD-related hospitalisation and mortality in women.Conclusions: Hospital admissions and mortality associated with asthma have fallen. Admission rates for COPD are declining for men, but there is no indication that admission rates for women have reached a peak. There is a need for higher prioritisation of COPD, including policies to reduce smoking in women, and medical practice initiatives to support primary and secondary prevention, pulmonary rehabilitation and appropriate drug therapies.

David H Wilson PhD, MPH, BEd · Graeme Tucker BSc · Peter Frith MD · Sarah Appleton BA · Richard E Ruffin MD · Robert J Adams MD

Entamoeba histolytica is one of the most common parasitic infections worldwide, infecting about 50 million people and resulting in 40 000–100 000 deaths a year. In Australia, people at risk of infection include immigrants, travellers returning from countries of high endemicity, Indigenous people, and men who have sex with men. Clinical manifestations range from asymptomatic carriage to invasive disease. Amoebic colitis and amoebic liver abscess are the most common invasive manifestations observed in Australia. Diagnosis depends on a high index of suspicion and laboratory investigations. Molecular methods (using the polymerase chain reaction) are the most sensitive for identifying and differentiating Entamoeba species. Treatment should always include a luminal agent to eradicate colonisation, prevent spread and/or reduce the risk of invasive disease. Medical therapy can successfully cure invasive disease, including amoebic liver abscesses.

Sebastiaan J van Hal MB ChB · Damien J Stark BSc, PhD · Rashmi Fotedar PhD · Debbie Marriott FRACP, FRCPA · John T Ellis PhD, DSC · Jock L Harkness FRCPA

Clinical record A 44-year-old man underwent coronary artery bypass surgery in 2004. He had been in satisfactory health as an adult apart from hypertension, for which he was receiving treatment. He ate a normal diet, including dairy products, meat and other high protein food. At age 44 years, he developed acute central chest pain while exercising at a gymnasium. An angiogram showed coronary artery occlusions. Forty-eight hours after successful coronary artery bypass surgery, during which time he received intravenous sodium but not glucose, he felt unwell, and the following day he became delirious. Computed tomography of the brain, initially reported as appearing normal, was later thought to show cerebral oedema. Plasma ammonium level was 110 μmol/L (reference range [RR], 10–50 μmol/L), but the timing of this sample was unclear. Wilson’s disease was initially considered because of elevated liver enzyme levels, but serum copper and ceruloplasmin levels were not measured. Eight days after the bypass surgery, a urine sample was sent to the NSW Biochemical Genetics Service, for metabolic screening. This showed gross elevations in glutamine and orotic acid (orotic acid, 14.8 μmol/mmol creatinine; RR, 1.23 μmol/mmol creatinine), indicating a likely diagnosis of ornithine transcarbamylase (OTC) deficiency. Plasma glutamine level was 3527 μmol/L (RR, 385–862 μmol/L). Plasma tyrosine and methionine levels were also moderately elevated, which was consistent with liver dysfunction. Plasma citrulline level was mildly elevated at 57 μmol/L (RR, 10–45 μmol/L), but this finding was difficult to interpret in the light of the other elevated amino acid levels. (Citrulline level is very low in neonatal OTC deficiency, but not necessarily low in late-onset phenotypes.) The patient was by then gravely ill, requiring maximum life support, and under consideration for liver transplantation. His condition progressively deteriorated despite introduction of intravenous sodium benzoate and l-arginine, and life support was withdrawn. He died 8 days after surgery. The diagnosis of OTC deficiency was subsequently confirmed by mutation analysis which showed hemizygosity for the c.622G>A (p. A208T) mutation in exon 6 of the OTC gene (as males have only one X chromosome, they are said to be hemizygous with respect to X-linked genes). Cascade testing of family members (Figure) showed that the patient’s mother (person II-2) was a carrier, heterozygous for the c.622G>A (p. A208T) mutation, while his asymptomatic brother (III-2) was hemizygous for the same mutation. Random biochemical testing of the brother showed normal levels of plasma glutamine, citrulline, arginine and urinary orotic acid. No information was available on causes of death of first-generation relatives. Past history revealed that, at age 7 years, the patient had an episode of acute encephalopathy following a 2-month history of intermittent nausea, vomiting and frontal headache. There was no preceding febrile illness, intercurrent infection or history of trauma. The symptoms reappeared after a 2-week period of apparent recovery. On admission to hospital, he was drowsy, disorientated and restless, but quickly settled and made a satisfactory recovery. No haematological or biochemical results were evident in his medical records. He was discharged with a diagnosis of viral meningitis/encephalitis. We report a 44-year-old man who presented with fatal hyperammonaemia after coronary artery bypass surgery. He had previously been asymptomatic, apart from a possible episode of unrecognised hyperammonaemia in childhood. The diagnosis of ornithine transcarbamylase (OTC) deficiency was made too late for successful intervention. Inborn errors of metabolism are frequently unrecognised or diagnosed late in adults. OTC deficiency is the most common disorder affecting the urea cycle. In New South Wales, the incidence is of the order of one in 70 000 births.1 It is an X-linked disorder leading to potentially lethal hyperammonaemia. The clinical severity ranges from acute neonatal hyperammonaemic coma to symptom onset at any time from infancy to adulthood, depending on environmental triggers and residual enzyme activity.2 The timing of episodes in late-onset OTC is dictated by environmental factors that increase nitrogen turnover, including dramatic increase in protein intake, medications affecting protein catabolism, viral illness or other generalised stress, rapid weight loss, and poor nutritional intake.3 Lessons from practice Inborn errors of metabolism are frequently unrecognised or diagnosed late in adults. Postoperative catabolism with insufficient calorie intake may unmask previously asymptomatic, but potentially lethal, inborn metabolic errors. Patients with acute onset of unexplained neurological or psychiatric symptoms need urgent metabolic investigation, including measurement of plasma ammonia level, to exclude metabolic causes. The first step in ureagenesis is the production of carbamyl phosphate from ammonium and bicarbonate. OTC then catalyses the biosynthesis of citrulline from ornithine and carbamyl phosphate. Thus, a deficiency of OTC leads to accumulation of ammonia and glutamine (the major extrahepatic source of ammonia for ureagenesis), and a reduction in citrulline. The accumulating carbamyl phosphate enters the pyrimidine synthetic pathway, resulting in increased excretion of orotic acid. The missense mutation p. A208T, replacing alanine with threonine at codon 208 of exon 6 in the OTC gene, has been previously reported in a late-onset OTC patient we investigated,4 and in others.5-7 This group showed an extremely wide phenotype, ranging from encephalopathy at age 4 months7 to no symptoms at age 97 years.5 Our patient developed hyperammonaemia following postoperative catabolism, caused by surgical stress and fasting with inadequate calorie supply from intravenous fluids pre- and postoperatively. In addition, a high nitrogen load from bleeding sites, tissue trauma or tissue protein breakdown could have overwhelmed urea synthesis and promoted the excessive ammonia production. OTC deficiency is not the only inborn error of metabolism that can result in fatal postoperative decompensation. Another example is medium-chain acyl-CoA dehydrogenase deficiency, which appears more prevalent, with several recorded cases (eg, Raymond et al8). Other mild fatty-acid oxidation defects and, perhaps, intermittent maple syrup urine disease could behave similarly. Death during an initial episode seems frequent in patients with late-onset OTC deficiency, as lack of familiarity with the disorder in the adult setting delays diagnosis and appropriate treatment.9-11 Treatment of hyperammonaemia is well established, and includes aggressive calorie support to counteract catabolism, early use of intravenous sodium benzoate as an ammonia sink, and intravenous arginine.2 It was unfortunate that no objective evidence was gathered on the cause of the patient’s episode of encephalopathy during childhood. While the cause could have been viral encephalitis, the clinical course, and appearance of the cerebrospinal fluid on microscopy and the brain on computed tomography did not strongly support this diagnosis. Assessment of plasma ammonia level would most likely have led to the diagnosis of OTC deficiency. Establishment of the correct diagnosis in the patient led to the finding that his mother and brother were also affected, enabling them to be advised about precautions. The genetic implications for the family’s younger generation were not critical, as the proband had no children, and his brother’s children were both male and could not have inherited their father’s X chromosome (Figure). This case illustrates the difficulty of diagnosing late-onset OTC. The X-linked inheritance may be obscured, even when more than one family member is affected, as some patients remain asymptomatic. Patients with acute onset of unexplained neurological or psychiatric symptoms need urgent metabolic investigation, including measurement of plasma ammonia level, to exclude metabolic causes. Cascade testing of family members The G-to-A base change in exon 6, which results in an amino acid substitution from alanine to threonine at position 208 (p. A208T), was identified in DNA from the blood of the patient with late-onset ornithine transcarbamylase deficiency (person III-1) and his brother (III-2). Their mother (II-2) was found to be heterozygous for the p. A208T mutation.

Mary Anne Chiong MD · Bruce H Bennetts PhD · Simone I Strasser MD, FRACP · Bridget Wilcken FRACP, FHGSA

Variation between laboratories and reference sources in the units used for reporting pathology results raises the possibility of medical error. Data submitted to the RCPA Quality Assurance Programs demonstrate wide variation in the units used for reporting therapeutic drug concentrations. This potential source of medical error needs to be addressed by all parties involved in communicating drug concentrations and providing support information.

Graham R D Jones MB BS, DPhil

A 47-year-old woman presented with a 6-month history of fever, redness and pain in her ears and nose, and hoarseness of voice. She had redness and thickening of both ears (Figure A, solid arrow) with sparing of the lobule (hollow arrow), and a collapsed nasal bridge (Figure B). A computed tomography neck scan revealed destruction of the thyroid cartilage (Figure C). Antineutrophil cytoplasmic antibodies were negative, ruling out Wegener’s granulomatosis. A diagnosis of relapsing polychondritis was made, and prednisolone and azathioprine were commenced. After 7 months of follow-up, there have been no further episodes of cartilagenous inflammation. Features of relapsing polychondritis A: Redness and thickening of the ear (solid arrow) and sparing of the lobule (hollow arrow). B: Collapsed nasal bridge. C: Computed tomography neck scan showing destruction of the thyroid cartilage.

Aman Sharma MD · Pradeep Bambery MD, FRCP(Glas) · Ajay Wanchu MD, DM · Surjit Singh MD · Naresh K Panda MS, FRCS(Edin)

A 75-year-old woman had a 2-year history of severe onycholysis and chloronychia (olive green discolouration) of most of her fingernails (Figure). Fungal infection had been excluded by microscopy and culture of nail scrapings. Some nails had superficial ridges and pits, as well as translucent orange-brown discolouration of the nail beds (the “oil drop” sign). This, along with a well defined, red, scaly plaque over her right elbow, led to a diagnosis of limited plaque psoriasis. The green discolouration was a classic sign of a secondary bacterial infection caused by Pseudomonas aeruginosa. Treatment was to consist of vinegar soaks (10 parts water and 1 part white vinegar applied topically for 5–10 minutes twice daily for 5 days) for the Pseudomonas infection, followed by a course of dexamethasone iontophoresis to treat the psoriatic onycholysis. (In the latter process, the fingers are held in contact with a dexamethasone solution and a low electrical current stimulates the transdermal migration of dexamethasone ions into the nail beds.) The “oil drop” sign is seen on the index fingernail. The fifth fingernail, without onycholysis and secondary infection, appears ridged and pitted.

Shinichiro Sakata MB BS · Anne Howard MB BS, FACD

It has long been my practice to write a letter of condolence to the next of kin of any patient of mine who has died. As a thoracic and sleep physician, I share a practice with my wife (a psychiatrist) in rooms attached to a private city hospital. I rarely see patients in their own environment, as most come from outlying suburbs or provincial towns. Nevertheless, I get to know many patients quite well, sometimes over a period of many years. However, unlike my father, who was a solo general practitioner initially in the country and later in Brisbane, I am unfortunate in being cut off from the local community in which patients live. He and my mother, who practised with him as a nurse, knew the patient’s whole family, the parish priest, and the names of their children and dogs, not to mention their full address. They saw three generations pass through their surgery.1 Over the years, I have also noted changes in the doctor–doctor relationship, which often affects our attitudes within the profession, and the fragmentation of services due to specialisation and defensive medical practice. Letters can be dangerous in the hands of lawyers, but surely not the simple letter of condolence. A recent American perspective on doctors’ letters of condolence shows that grieving has changed over the past century in US society, so that such letters from doctors are now infrequent.2 People are expected to “get over it” in a matter of weeks. This is in marked contrast with the custom in my grandparents’ time, when “full mourning” was symbolised by black clothing, followed by a period of “half mourning”, signified by grey. Certain behaviour was proscribed during the mourning period, such as going to parties and balls; any jewellery worn had to be made of jet, with no sparkly diamonds; and so it went on. As a specialist in Australia, there are several reasons why I persist in writing letters of condolence today. The first reason is that it is simply a decent, tangible thing to convey my sympathies in writing to the next of kin and show that the patient was not just a “file” or a Medicare item number. Sometimes it may be possible to explain why the patient died, particularly if relatives are confused or upset about the mode of death, but generally such details are not required. The second reason is for myself as a sort of closure on what may have been a long and pleasant relationship with not only the patient but also the family and friends. I am given the opportunity to express both grief and the positive aspects of this unique doctor–patient relationship. Some may see this as “soppy”, but I believe I have learnt over the years that we neglect our own soul at our peril. The third reason is to convey to my staff my sentiments and to give them, too, some form of closure, as the feelings of staff are often overlooked. They may type, read and send the letter and therefore are an integral part of the process. It also sets a tone for my practice. The helmsman is at the helm leading by example. The fourth reason is that a copy of this final letter is filed in the patient’s record that will go off to the archives for the statutory number of years. This indicates to anyone who recalls the chart that the patient has died. I have seen situations in which a patient’s chart is pulled out of the records and a routine appointment (eg, annual follow-up) made for someone who has long since died. This unwitting contact with relatives may lead to unnecessary distress. I recently did a lengthy medicolegal report about a man who had died of lung cancer. In a mountain of subpoenaed medical notes, including that of his family doctor, I could not find the date or situation of his death, let alone a single letter of condolence. By way of a general example, I recently wrote the following letter to the wife of an elderly ex-serviceman I had had the privilege to treat for many years. He had gone downhill very suddenly. As a former Army medical officer myself, with service in East Timor, I often shared with him reflections about world events.3 Dear [name of wife] I was sorry to hear the bad news this morning that . . . had died from lymphoma. I received the biopsy report yesterday and I was going to telephone you this morning to see how he was going. I always enjoyed seeing him and respected his long military background, which was longer than I have seen in any patient in my career and spanning nearly every conflict we have been involved in, including World War 2. The RSM* is the backbone of the Army. I salute him. I am sure he will be sadly missed by you and your family and I wish to convey my sincere sympathies at his passing. Yours sincerely . . . * RSM = Regimental Sergeant Major. I write this article for all doctors, but most of all for those who are just entering the profession. I commend this practice as we share in this mortal business, as it is also for us “for whom the bell tolls”.

Roger K A Allen MB BS(Hons), FRACP, PhD

To the Editor: We read with great interest the article by Bambrick et al relating to the potential impact of the Australia–United States Free Trade Agreement (AUSFTA) on supply of blood products in Australia.1 Our recent experience with Octagam (Octapharma Australia, Sydney, NSW), an intravenous immunoglobulin (IVIg) product produced overseas, highlights some of the quality concerns raised in their article. Routine practice in our bone marrow transplant unit is to administer IVIg weekly for 100 days after allogeneic stem cell transplantation. Until December 2004, locally produced IVIg, Intragam-P (CSL, Melbourne, Vic), was used exclusively as the IVIg product for these patients. From October 2005, because of limitations in the supply of Intragam-P, the Australian Red Cross Blood Service (ARCBS) also provided Octagam for IVIg replacement therapy in transplant recipients. It has also been routine practice within our transplant unit to repeat serological tests for a variety of transfusion-transmitted viral infections, including human T-lymphotropic virus type I and type II (HTLV-I and HTLV-II), in all transplant patients 100 days after transplantation. Until 2006, none of our patients had ever tested positive for HTLV-I or HTLV-II antibodies. After the introduction of Octagam, the first two transplant patients who received this product for IVIg replacement tested positive for HTLV-I/HTLV-II antibodies at 100 days after transplant (signal to cut-off [S/CO] ratios, 4.36 and 6.33, respectively). Subsequent investigation revealed that these results were probably secondary to passive transfer of HTLV antibodies from the IVIg product used. Both patients received Octagam from the same batch, and subsequent testing of this batch was positive for the presence of HTLV-I/HTLV-II antibodies. Of note, both patients tested negative for HTLV-I/HTLV-II antibodies before transplantation (S/CO ratio < 1.00). Their stem cell donors were also negative for HTLV on testing immediately before stem cell donation, and the only other blood product shared between the two patients (platelets from a common donor) also tested negative for HTLV. Follow-up testing for HTLV-I/HTLV-II antibodies at about 12 months after transplantation gave a negative result in both patients (S/CO ratio < 1.00). Given that Octagam is a plasma (acellular) product processed with appropriate viral inactivation steps,2 we believed it to be extremely unlikely that direct transfer of HTLV virus had occurred. The most likely explanation was the passive transfer of HTLV antibodies. It followed that Octagam must have been sourced from HTLV-positive plasma donors — a practice that is in direct conflict with current ARCBS policy, which specifies that all blood and plasma donors must be screened for HTLV-I and HTLV-II, and that any donors testing positive should be excluded from blood or plasma donation.3 The HTLV-I/HTLV-II serostatus of donors used to source plasma for Octagam are not reported on the product information sheet.2 Testing of one patient sample at the National Serology Reference Laboratory (Melbourne, Vic) suggested the positive serological results in our patients were due to the presence of HTLV-II antibodies. In collaboration with Octapharma, it was subsequently determined that plasma for Octagam was sourced from paid donors from the southern United States, an area where HTLV-I and HTLV-II seropositivity is known to be prevalent among blood donors.4 The clinical implications of our findings are unclear. Our results were reported rapidly to the ARCBS and subsequently to the Therapeutic Goods Administration. Our main concerns are that Octagam plasma is sourced from donors who would normally be excluded from plasma donation within Australia,3 and that there appears to be no current mechanism for addressing this issue. Some of the quality concerns raised by Bambrick et al appear to be not so theoretical after all.

Glen A Kennedy · Judy Cummings · Simon T Durrant

In reply: I would like to respond to the letter by Kennedy et al regarding the passive transfer of human T-lymphotropic virus (HTLV) antibodies following Octagam (intravenous immunoglobulin [IVIg]) administration. Octapharma does not routinely test Octagam for HTLV antibodies — this is in line with Australian regulatory requirements.1 Furthermore, in accordance with global regulatory requirements (including those defined by the Australian Therapeutic Goods Administration), plasma from sources in Europe and the United States is also not routinely tested for HTLV antibodies.2,3 It is important to note that the patients described by Kennedy et al tested positive for HTLV antibodies rather than the HTLV virus, and that the presence of antibodies in the finished product does not pose an infection risk. As noted by the authors, the manufacturing process for Octagam includes viral inactivation. There are two dedicated viral inactivation and removal steps: Solvent/detergent treatment, which is a highly effective step for inactivating lipid-enveloped viruses; Incubation at pH 4 and 37°C for 24 hours. The combination of these two processes ensures that lipid-enveloped and several non-lipid-enveloped viruses are inactivated. Since the initial launch of Octagam in Europe in 1993, more than 33 million grams of the product have been infused, corresponding to more than 2.5 million infusions. During this time, there have been no documented cases of viral transmission resulting from the use of Octagam. Since the launch of Octagam in Australia in 2005, Octapharma has supplied over 600 kg of IVIg, representing about 40 000 infusions to Australian patients. Therefore, taking into account the steps used to remove and/or inactivate lipid- and non-lipid-enveloped viruses, the fact that there have been no documented cases of transmission of any virus since the launch of Octagam, and that plasma sources and testing meet both Australian and global regulatory requirements, the positive HTLV antibody results reported by Kennedy et al have no bearing on either the quality or safety of Octagam.

Wolfgang Frenzel

To the Editor: Bambrick et al1 suggest that the importation of overseas plasma products or the processing of Australian plasma overseas may pose a threat to the safety and security of Australia’s supply of plasma products. The National Blood Authority (NBA) and the Therapeutic Goods Administration (TGA) would like to respond to these issues to ensure that clinicians have confidence in the safety of products currently available in Australia and a better understanding of current blood processing and supply arrangements. Australia is largely self-sufficient in plasma products, a position supported by all Australian governments.2 The governments also agree to the supply arrangements for imported plasma products when demand exceeds domestic supply (as in the case of intravenous immunoglobulin [IVIg]), and for products not supplied by Australian manufacturers (eg, fibrin sealant and other coagulation products). All these products are purchased by the NBA on behalf of Australian governments. Standards applied in Australia ensure that all products on the Australian market are derived from sources — in Europe, the United States and Australia — approved by the relevant authorities. The TGA regulates all plasma products to ensure that they meet international standards of safety, quality and efficacy, irrespective of their source.3 No centres sourcing blood from countries in the developing world, as cited by Bambrick et al, are used. National health systems around the world strive to attain degrees of self-sufficiency that suit their particular economic and policy objectives. However, as the US supplies 60%–70% of global plasma (while consuming only 40% of products sourced from this plasma4), many patients outside the US are dependent on the system of both compensated and uncompensated donors operating in the US. It is worth noting that definitions of “remuneration” vary across countries. The Commission of the European Communities reported in 2006 that “the principle of voluntary and unpaid donations does not exclude compensation for donors, if it is limited to making good the expenses and inconveniences related to the donation”.5 Examples of compensation that this article cites include tax relief of up to €70 per annum in the Czech Republic, an expense allowance of up to €25 for a whole blood donation in Germany, and up to €50 for an apheresis donation in Austria. Thus, the boundary between compensated and non-compensated donors is not distinct globally. Bambrick and colleagues’ contention that products manufactured from paid donors may be less safe is not supported by evidence. The history of blood safety clearly demonstrates that there were major safety issues with both fresh blood and blood products in the 1980s. These problems were more a result of pathogen epidemiology and governments’ blood safety policies than whether donors were paid or unpaid. For example, the Canadian and French blood systems relied entirely on volunteer donors, but the delayed implementation of safety measures and good governance measures led to pathogen risks that exceeded those of the US.6,7 In Australia, the incidence of HIV/AIDS in people with haemophilia exposed to only one type of product in the 1980s approached that of the same patient group in the US, despite the product being sourced entirely from domestic voluntary donors.8 Currently, robust plasma product safety measures in the US have proved to be effective in minimising contamination from both known and emerging pathogens, such as West Nile virus. This virus did not infect the recipients of plasma products from compensated donors, but did infect the recipients of (uncompensated) fresh blood transfusions.9 The equivalence in safety between plasma products sourced from compensated and uncompensated donors has been confirmed by the European Medicines Agency.10 Ensuring the security of supply is central to Australia’s plasma fractionation arrangements. The NBA’s contracts include provisions to ensure product supply security and product safety, including compliance with TGA requirements. Under Australia’s emergency response plans, plasma could be supplied from fresh stock, while the inventory of product in the system and the national reserve of products (funded by the NBA to cover contingencies) could provide fractionated products. A range of other supply security measures is implemented by the NBA on behalf of Australian governments, including secondary suppliers for critical products. These measures take into account the fact that Australian plasma used to manufacture fractionated products is not produced in sufficient quantities to permit storage of excess. Thus, Bambrick and colleagues’ concern that geographical factors may restrict access to Australian plasma has limited relevance in an emergency situation. In summary, Australia has a comprehensive system that draws on international best practice and national jurisdictional arrangements to ensure the supply of high-quality, safe, efficacious plasma products, irrespective of their source.

Alison Turner · Albert Farrugia

In reply: Turner and Farrugia highlight important issues about the safety and supply of Australia’s blood and plasma that constituted key findings of the recent Australian Government review of Australia’s plasma fractionation arrangements.1 These included the desirability of self-sufficiency and the need for a viable contingency plan in the event of local supply failing to meet demand, in the context of probable increased uncertainty in relevant security and supply circumstances. The authors note the importance of a rapid regulatory response to new risks, and rightly emphasise the important role of our National Blood Authority and the Therapeutic Goods Administration in relevant safety assurance processes. However, a crucial issue highlighted in the review that they fail to address is the difficulty of relying on the application and enforcement of regulatory standards if the bulk of Australia’s fractionation occurs offshore.1 After an exhaustive study of arrangements in Europe and the United States, the review concluded that overseas fractionation of Australian plasma would involve significant costs in moving away from the current local arrangements ($75 million) and, because of yield considerations, there would be the potential for an ongoing shortfall in the supply of intravenous immunoglobulin (IVIg) and other plasma-derived products. The review also found that overseas fractionation was potentially associated with major risks to the supply chain, with increased distance and handling providing more opportunities for loss and error, while a doubling or tripling of the turnaround period would have implications for continuity of supply. The review recommended that the federal government maintain the reservation exempting plasma fractionation services from the government procurement provisions of Chapter 15 of the Australia–United States Free Trade Agreement.1 The review confirmed that volunteering for blood donation should be institutionally reinforced as an important exemplar and means of sustaining Australia’s national culture, and that donor payment endorsed a very different set of values and was unlikely to assure sustainability of supply.2 It did, however, suggest initiatives such as tax relief or other institutional ways to encourage donation.1 We fully endorse the summary assessment of the review that: [T]he current structural arrangements, whereby domestically collected plasma is fractionated by CSL Bioplasma, are, subject to careful monitoring of prices, in Australia’s best interests. The present system is well entrenched in the “hearts and minds” of the Australian population and of the Australian medical community and, particularly, in the strategy, thinking and reliance of all end user groups.1

Hilary J Bambrick · Thomas A Faunce

To the Editor: Gleason et al describe the continued tendency of hospitals to allow prevocational, newly graduated doctors to flounder as they provide the cannon fodder to keep those hospitals running.1 For about 5 years, I coordinated an education program for junior medical staff. It is a model that other areas could consider. The Northern Sydney Area Health Service — as it was then — had two primary allocation centres for junior staff, and three local secondment hospitals. During their time in Sydney, all interns at those five hospitals were required to attend a 3-hour program of education provided for them, off site, each fortnight. In the first week, half of the interns attended; the program was repeated the following week to allow the other half to attend. A separate program was developed for Postgraduate Year 2 doctors that provided day-long programs centred on a particular topic (eg, renal disease or paediatrics). Each hospital put half of their Postgraduate Medical Council of New South Wales (now NSW Institute of Medical Education and Training) funding towards the cost. The teaching was provided by clinical staff within the area health service. The size of the area health service meant that there was an enormous pool of expertise to call on. The very size of that pool meant that we did not over-use the same keen teachers. The quality of the teaching was regularly rated as high by the doctors who attended. The off-site venue meant that time was truly quarantined. The provision of breakfast and morning tea meant an opportunity for junior medical staff to spend time with their cohort, compare experiences, complain, commiserate, and congratulate, in an informal atmosphere. Of course senior staff — administrative and medical — would complain. However, they gradually grew used to the format, and learned to time rounds, to make allowances and to value their happier, better educated staff. Of course there were problems of all sorts, and the system was not perfect, but it signalled to junior staff that they were valuable and valued. As Gleason et al noted: Hospitals must meet their training responsibilities and should not continue to place service demands above the training needs of doctors. Teaching time needs to be a regular, protected, paid part of every junior doctor’s day. In 2007, it is completely unacceptable to continue to see junior medical staff as “workforce”. It shames us all as senior clinicians that we have not ensured that all hospitals meet their educational and training responsibilities to these valuable members of our profession.

Lilon G Bandler

To the Editor: Medical education faces a number of significant challenges in the coming years. The most concerning problem is that there are insufficient medical educators. To compound this problem further, clinicians with an interest in teaching lack academic recognition, funding, time for medical education, and institutional support.1 While this is already creating problems for training of medical students and junior doctors, the situation will only become worse as the number of graduates increases to meet workforce shortages. A greater number of medical students and graduates will put further pressure on an already over-stretched health system to deliver adequate medical education. Therefore, we need to consider new ways of creating effective clinical teachers and learning materials that are educationally sound and supported by research. One way to address some of these deficits is to create specific training positions for doctors interested in pursuing a career in medical education, perhaps called “medical education registrars”. These positions could be jointly funded by hospital networks, universities and other interested groups, such as the postgraduate medical councils, and could be aimed at doctors with a strong interest in medical education, especially those at an early stage in their career, such as postgraduate year (PGY) 3 and PGY4. This joint funding would support hospitals and universities working collaboratively, and taking responsibility for making medical education “a priority, rather than an add-on”.2 The hospital appointment would maintain the registrars’ clinical knowledge, and the university appointment would allow opportunities for further study in medical education theory and research methods. Both environments would allow for academic mentoring. These registrars would then be ideally placed to assist in the development, implementation and evaluation of education materials using the newly released Australian Curriculum Framework for Junior Doctors as a guide. While agreeing that the Framework provides a “unique opportunity to improve the quality of medical training in Australia”, Gleason et al expressed concerns about how the Framework would be implemented.3 Medical education registrars could play a pivotal role in this process, and would be strong advocates for junior doctors. Medical education registrar positions are a simple, cost-effective way to encourage clinicians, especially those early in their careers, to pursue medical education as a serious and satisfying career path. Urgent action is required now to avert certain disaster in a few years, when hospitals are full of medical students and junior doctors, and there is no one to supervise and train them.

Andrew M Foote

In reply: A teaching registrar position is a great idea. It could foster development of knowledge and skills in interested clinicians, and perhaps higher degrees in learning and teaching, as well as serving as the start of a career pathway.1 The United Kingdom allows specialist trainees to take a year out to develop skills in education or research. Here, however, it is doomed to fail, unless there is a receptive environment for these registrars to work in or a career path that is attractive. What kind of educational environment is needed? We need leaders with vision, knowledge, new ideas and resources to lead the development of programs and assessment. We need teachers whose main focus is to teach and implement programs. And we need to support all clinicians to continue to supervise and teach. Not all these roles need to be performed by a medical practitioner.2 Without all of this, someone who is “interested in teaching” may be burdened with an unsustainable level of teaching and administration. Better integration at a local level across undergraduate and prevocational training3 (not to mention linking with other professional groups) and a culture of teaching are needed. Importantly, we need coherent and visionary leadership at national and state levels to underpin these endeavours.3,4

Fiona R Lake

Poached eggs A leading fertility expert in Israel, who removed hundreds of ova — without permission — from private patients undergoing fertility treatments in the late 1990s, will lose his medical licence for 2 and a half years as punishment, according to a news report in the BMJ. The expert had used the ova for the in-vitro fertilisation (IVF) of infertile women; in one woman, 232 ova were removed, with 155 used for IVF in 33 infertile women. Since the case came to light, the Israeli government has legislated to allow women who are not themselves undergoing fertility treatments to donate their ova altruistically. BMJ 2007; 334: 557 To scan or not to scan? At this time, screening for lung cancer by computed tomography (CT) should be considered an experimental rather than an evidence-based procedure, say US and Italian researchers. They conducted a small, preliminary study of the effect of such screening on lung cancer diagnosis and death from lung cancer in 3246 current or former smokers enrolled in one of three studies being conducted at the Mayo Clinic in Rochester, Minnesota, the Moffitt Cancer Center in Tampa, Florida, and at the Istituto Tumori in Milan, Italy. Compared with predictions of what would have occurred in the absence of screening, CT screening increased the rate of detection of small, resectable lung cancers but did not reduce the risk of either advanced lung cancer diagnosis or death from lung cancer. These findings mirror those of earlier studies of screening for lung cancer by chest x-ray. The researchers said that before CT screening is offered more widely, we should await the findings of the US National Lung Screening Trial and several other trials being conducted and planned in Europe. JAMA 2007; 297: 953-961 Bad heart days No day can be a good day for having your first heart attack; however, if you live in the United States, some days may be better than others. The Myocardial Infarction Data Acquisition System (MIDAS 10) Study Group reported that in New Jersey, from 1987 to 2002, patients with a first acute myocardial infarction (AMI) admitted on weekends had higher mortality than those admitted on weekdays. They pointed out that there was also a lower rate of invasive cardiac procedures on weekends and speculated that better access to care on weekends could improve outcomes for patients with AMI. N Engl J Med 2007; 356: 1099-1109 Twin fates Planned caesarean section may be beneficial for all twins, suggest UK researchers. They studied the records of 1377 UK twin pregnancies in 1994-2003, in which there was an intrapartum stillbirth or neonatal death but the other twin survived. They found an increased risk of death among the twins born second, rather than first, at term, which was related to delivery. Vaginally delivered second twins had a fourfold risk of death caused by intrapartum anoxia. However, an editorialist commented that it was important to obtain further evidence from randomised controlled trials before caesarean section for twin pregnancies at term becomes universal (and thus, a trial becomes impossible). BMJ Online, 2 March 2007 Resuscitate me In some witnessed out-of-hospital cardiac arrests, bystander cardiac-only resuscitation may be more effective than conventional cardiopulmonary resuscitation (CPR), according to a Japanese research group. The SOS-KANTO study group conducted an observational study of more than 4000 adult patients who had had an out-of-hospital cardiac arrest witnessed by bystanders. At 30 days after cardiac arrest, those who had received any bystander resuscitation (about 30%) had a more favourable neurological outcome than those who hadn’t (about 70%). However, in some types of patients — those with apnoea, those with a shockable rhythm, and when resuscitation began within 4 minutes of arrest — those who received cardiac compressions only from bystanders did better than those who received conventional CPR. The researchers suggested that interruption of chest compressions was the main reason why conventional CPR did not result in better neurological outcomes in these patients. Lancet 2007; 369: 920-926 The A to Z of weight loss US researchers say doctors whose patients want to try diets that are very low in carbohydrates, like the Atkins diet, can be reassured on two counts. Not only is any weight loss likely to be as large as for any other diet, but also any potential lipid effects are unlikely to be of immediate concern. Their A TO Z Weight Loss Study compared the effects of four diets representing a spectrum of carbohydrate intake — Atkins (very low), Traditional (high), Ornish (very high), and Zone (low) — in 311 premenopausal overweight and obese women. After 12 months of dieting, all four diets led to a modest weight loss of 2% to 5%, but the women assigned to the Atkins diet lost the most weight (an average of 4.7 kg) and there were no adverse effects on lipid variables. JAMA 2007; 297: 969-977

Ann Gregory