Volume 176 Issue 7

The impact of stroke remains considerable, despite a modest decline in case fatality1 and an encouraging reduction in incidence on the west coast of Australia during the past decade.2 To the optimists among us, it appears that some efforts, particularly in public health and primary care, are paying off. However, population risk-factor surveys indicate that, although the prevalence of smoking has declined, the prevalences of hypertension and hypercholesterolaemia have changed minimally and the prevalences of obesity and diabetes have increased dramatically over the past decade.1,3 The role of the general practitioner is pivotal in identifying and managing these risk factors. The Avoid Stroke as Soon as Possible (ASAP) general practice stroke audit, published in this issue of the Journal (page 312),4 provides further evidence that modifiable risk factors for stroke and vascular disease are highly prevalent in the Australian community. Importantly, the ASAP audit has measured the prevalence of atrial fibrillation and transient ischaemic attack (TIA). These factors are unique to stroke pathophysiology and are associated with high absolute risk, but have received limited attention in previous risk-factor surveys. Sturm et al reviewed 16 148 general practice consultations surveyed from a random sample of 321 GPs across five States. Seventy per cent of patients aged 30 years or more had at least one major risk factor; 34% had two or more. Not surprisingly, hypertension (44%) and hypercholesterolaemia (43%) were the most prevalent modifiable factors. Importantly, the proportion of older patients at high absolute risk was considerable. For example, among the men, 14% aged 70–79 years and 16% aged over 80 years had atrial fibrillation. Reliable measurement of the occurrence of TIA is more difficult given the variability in presentation and clinical assessments. However, the figure of 4% for any history of TIA is valuable information and provides a baseline for further assessments. The high prevalence of vascular risk factors identified in the ASAP survey highlights the potential value of screening during all routine general practice consultations to identify individuals with high absolute risk of stroke. The use of absolute-risk estimates in guiding therapeutic decision making is a well-established paradigm and is being adopted increasingly often in national and international vascular disease management guidelines.5 The value of risk stratification is highlighted by the exercise of choosing the most appropriate antithrombotic agent for primary stroke prevention in atrial fibrillation. Although the risk of first stroke in atrial fibrillation averages 5% per year, risk varies from 1% or less per year up to 12%, depending on additional clinical characteristics.6 At 1% risk, any benefit from anticoagulation therapy would be nullified by the approximately 1% risk of major bleeding events. Antiplatelet therapy would be the more appropriate option in this setting. In contrast, at 12%, the reduction in risk of thromboembolic events achieved with anticoagulation far outweighs the risk of adverse bleeding events. In this setting, warfarin shows a clear net benefit and would be more appropriate. Absolute risk translates readily into clinically useful statistics, such as "number needed to treat" at the bedside, and, at the population level, the recently proposed "population impact number".7 At a practical level, however, owing to the complexity of interacting risk factors, risk stratification typically requires decision-assistance charts or computer programs (generally based on Framingham risk estimates).8,9 At present, such systems are still under evaluation or just beginning to be adopted in general practice. From the cost–benefit perspective, in the absence of a framework for risk stratification, identifying relatively low-risk populations with isolated risk factors increases the possibility that expensive drug therapies will be used as first-line options when low-cost, non-pharmacological interventions would be more appropriate. An example is a 58-year-old woman with moderate hypertension (repeated blood pressure readings 160/95 mmHg) and no other risk factors. Her absolute risk of stroke is 1.5% over five years.10 Although she is a potential candidate for antihypertensive drug therapy, the low absolute risk may suggest to the patient there is questionable value in complying with prolonged drug therapy with its attendant side-effects. It indicates to the clinician that the number needed to treat over five years in this type of patient exceeds 120 to prevent one stroke.11 Accepting the various difficulties and risks associated with screening in general practice, combining a "high risk" approach and an appropriately organised "mass" approach is likely to have a much greater impact on the overall burden of vascular disease. This combination helps minimise the problem of focusing efforts on very cost-effective treatment of small populations at high risk while ignoring the larger subgroup of the population at moderate risk, from where most vascular events emerge. Low-risk groups would also benefit from further surveillance for progression into higher-risk categories. The ASAP stroke audit provides further evidence that considerable opportunity exists to reduce the burden of stroke and other vascular diseases. Quite rightly, the authors emphasise the importance of making the most of this opportunity. However, a prerequisite for success is a framework to facilitate rational use of interventions, balancing absolute risk and potential benefit at both the individual and population levels. Although debate continues about the best decision-assistance tool to be used, any system will need to be simple and practical to use within the time and resource constraints of an already overburdened primary healthcare sector.

Christopher R Levi FRACP · Parker J Magin FRACGP · Balakrishnan R Nair FRCP, FRACP

The ascendancy of evidence-based medicine over the last decade has fostered an unprecedented growth in practice guidelines.1-3 Departments of health and associated agencies, specialty societies and other medical organisations have all embraced the development of guidelines in the belief that adherence to their recommendations translates into benefits for patients (improved outcomes), practitioners (improved quality of care), and providers (improved cost-effectiveness).4 As the practice guideline movement has matured, the interest and debate surrounding guideline developments has shifted from the what to the how.5 In fact, the National Health and Medical Research Council's Guidelines for the development and implementation of clinical practice guidelines, published in 1995,6 was a pacesetter in the how of guidelines, and at its core are the three criteria that underpin the quality of guidelines: a balance of healthcare disciplines in the guideline development group, together with consumer representation; a systematic review of relevant literature and its stratification of the data according to a hierarchy of levels of evidence; and the generation of evidence-graded recommendations that also take into account the generalisability of the evidence, its practice relevance and resource implications. Despite the availability of guidelines for developing guidelines,6,7 recent reviews of practice guidelines have shown that most fail to meet quality standards.1-3 This is hardly surprising, as guideline development is an intensive, laborious and at times uncertain task.8 In essence, it is a human affair with all its attendant nuances, complexities and biases. These human elements are illustrated in this issue of the Journal by Edmonds and colleagues (page 332)9 in their account of the workings of the Australian COX-2-Specific Inhibitor Prescribing Group. The outcome of its deliberations is also published in this issue of the Journal (page 328).10 The process pursued by the Prescribing Group followed the essentials outlined in the guidelines for guidelines.6,7 The working party was multidisciplinary, but with a difference! It had a high proportion of academics and clinicians with connections with the pharmaceutical industry, together with an undeniable first — representatives from the pharmaceutical companies involved in marketing COX-2 inhibitors. Initially, the group was embroiled in a debate on conflict of interest and its impact on the credibility of the outcomes, but this was only the beginning. A clash of minds in the interpretation of evidence led to resignations from the group, as did irreconcilable differences with its ultimate recommendations. This story suggests that the debate in guideline development should now shift even further from the how to the who . The major issues arising from the Australian COX-2-Specific Inhibitor Prescribing Group's deliberations involve aspects of conflicts of interest and the management of differing interpretations of evidence. The definition of conflict of interest is complex, but a pragmatic description is "conflicts of interests comprise those which may not be fully apparent and which may influence judgment of authors reviewers or editors. They have been described as those which, revealed later, would make a reasonable reader feel misled or deceived".11 In contrast to the recent emphasis in disclosure of conflict of interest in scientific and medical publishing, it appears that this dsclosure has a low profile in practice guidelines. In two extensive surveys on quality in more than 700 practice guidelines, a statement of conflict of interest was not among the standards surveyed.1,2 Further, in a recent survey exploring the relationships between authors of clinical practice guidelines and the pharmaceutical industry,3 financial support for authors by pharmaceutical companies was declared in only two of the 44 guidelines surveyed, and industry financial support for the creation of the guideline was reported in only 20. Relationships with the pharmaceutical industry are widely accepted as being a potent catalyst for the potentiality of conflict of interest.12 But the time has come to explore conflicts in relationships with health departments, other government agencies, various healthcare providers and specialty organisations. Full and candid disclosure of conflict, as pursued by Edmonds and colleagues and as advocated by others (Box), is the only solution. One inherent assumption of evidence-based medicine is that differences and controversies stemming from interpretation of data are inversely related to the quality of evidence. However, if the experience of the Australian COX-2-Specific Inhibitor Prescribing Group is anything to go by, this assumption would appear to be optimistic, as at the end of the exercise eight members of the group had resigned, essentially because of differing mindsets. The Journal has long encouraged authors of practice guidelines to cite the level of evidence underpinning recommendations. What remains intriguingly unknown is the range of levels of agreement on the evidence. The account by Edmonds and colleagues is a refreshing and welcome exposition of the "real world" of guidelines development. Its central message is that practice guideline development is not a black-and-white affair, and that the time has come for open disclosure of conflicts of interest and the degrees of consensus or dissent. As guidelines have a powerful influence on clinical practice, doctors and their patients deserve nothing less than complete openness and transparency. Recommendations for managing conflict of interests in practice guidelines development3 A formal process should exist to disclose potential conflict of interest before the guideline development begins. All members of the guideline group should be involved in a discussion of conflicts of interest and how significant relationships will be managed. Participants who have relationships with industry, government agencies , healthcare organisations or specialty societies need not necessarily be excluded, but the group has to decide among itself a threshold for exclusion. There must be complete disclosure to readers of the practice guidelines of financial and/or other relationships with industry, government agencies, healthcare organisations and specialty societies.

Martin B Van Der Weyden MD, FRACP, FRCPA

The widespread application of public iodine supplementation programs, which cover about 3–4 billion people worldwide,1,2 is a response to the paucity of iodine in the natural diet in many regions of the world and the severe public health consequences of iodine deficiency.3 However, in some countries, the tendency to low iodine intake has mostly been corrected by haphazard increases in the iodine content of certain parts of the diet.4 In the United Kingdom, dairy products may contain extra iodine as a result of adding iodine to cow feed to increase the animals' reproductive performance or of using iodine-containing cleansing agents in the dairy industry. As reviewed by Phillips,5 this unplanned increase in iodine intake has eliminated endemic goitre in Britain during the last 30–40 years. In the United States, London and colleagues encountered cases of unexplained very high iodine intakes (1100–1300 µg per day) in 1964. Subsequently, they discovered that bakers used iodine-containing conditioners in bread and that this caused high levels of iodine intake.6 Obviously, such unplanned variation in dietary iodine is a hazardous way of providing a population with an adequate intake of iodine, as mechanisms unrelated to disease prevention can profoundly alter iodine intake. To some extent, Australia may be a country where factors other than disease prevention have modulated the intake of iodine, and iodine intake may now be in an unplanned phase of decrease.7 A report by McElduff et al8 (page 317) in this issue of the Journal seems to support this proposition. McElduff and colleagues looked at the frequency distribution of whole-blood thyroid-stimulating hormone (TSH) concentrations in newborns in the northern Sydney area. TSH is measured as part of screening for congenital hypothyroidism. In 5%–10% of infants around 72 hours after birth, TSH values were above 5 mIU/L. The World Health Organization (WHO) recommends assessment of TSH concentrations in newborns to detect population iodine deficiency, and specifies that less than 3% of newborns should have a whole-blood TSH concentration over 5 mIU/L. In a subsample of neonates, McElduff et al found that, during pregnancy, their mothers had a median urinary iodine concentration of 109 µg/L, indicating borderline mild iodine deficiency. WHO specifies that the median urinary iodine concentration in adults should be over 100 µg/L, and an extra iodine intake of 50 µg/day in pregnant and lactating women.1 Thus, the corresponding median urinary iodine concentration of pregnant women would be around 130 µg/L. McElduff et al warn that Sydney may be an area of iodine deficiency, and suggest that iodine intake and risk of disease should be investigated in more detail. Their concern is well founded. Even if there is no documentation that these borderline iodine values are harmful to a mother and child, the margin of safety is small. Furthermore, in Australia, there is apparently no regular surveillance of population iodine status, or of the variable iodine content of dairy products and other foods. The iodine intake may well be even lower in other sections of the Australian population. Severe iodine deficiency may cause brain damage and other developmental disorders,3 and goitre and its complications may affect a significant proportion of the population at all levels of iodine deficiency.9 Any public healthcare system should evaluate iodine intake and prevent disorders caused by iodine deficiency. Several factors need to be taken into account in such an evaluation and prevention program: The relationship between iodine intake and the risk of thyroid disease is not a simple one. Even if iodine supplementation may decrease the risk of some thyroid disorders, the risk of other disturbances at a younger age, such as hypothyroidism and Graves' disease, may increase.4 Severe iodine deficiency, with a median urinary iodine excretion of less than 25 µg/24 h, is an instance where giving any type of iodine supplementation is better than doing nothing. However, at higher levels of intake, careful planning and surveillance are needed. The methods often used for evaluating iodine intake and the risk of disease are not perfect. Neonatal screening showing more than 3% of TSH values over 5 mIU/L is not, by itself, enough to indicate insufficient maternal iodine intake. Detection of neonatal hypothyroidism requires identification of relatively high TSH levels (20–25 mIU/L), and many TSH assays and screening programs are not designed to identify TSH values around 5 mIU/L with reasonable confidence. Technical aberrations can easily give an increased frequency of elevated blood TSH concentrations. One such aberration occurs with sampling of blood before TSH has fully returned to baseline levels after the early postnatal surge. As discussed by McElduff et al,8 early sampling may have contributed to their findings. Iodine deficiency is not the only pathogenetic mechanism leading to an increase in blood TSH levels in neonates. Iodine has an autoregulatory inhibitory effect on the thyroid gland, with a fall in both thyroid hormone synthesis and secretion. Possibly, this mechanism has been developed to protect against hyperthyroidism induced by a sudden iodine load. In a variety of abnormal states the thyroid gland overreacts, producing hypothyroidism. The thyroid of the fetus and infant is considerably more sensitive to iodine inhibition than the maternal thyroid. Excess iodine intake, rather than iodine deficiency, in mother or infant has been a more common cause of transient neonatal hypothyroidism in countries with a relatively low iodine intake, such as Germany, Italy and Belgium.4 In severe iodine deficiency, iodine supplementation to the mother decreases the abnormally high serum TSH in both the mother and the newborn.10 On the other hand, in pregnant women with urinary iodine concentrations around 50 µg/L, iodine supplementation decreases TSH levels in mothers, but not in cord blood. TSH levels in the newborn may even be higher after iodine supplementation.11 The finding of McElduff et al of a positive correlation between maternal urinary iodine concentrations during pregnancy and whole-blood TSH levels in neonates needs further elaboration,8 but it may be an example of iodine autoregulation of the fetal thyroid. Should pregnant women living in mild and moderately iodine-deficient areas receive iodine supplements, and does supplementation involve any risk? An increase in iodine intake will improve thyroid function in pregnant women, which is important for early brain development in their infants.12 There are still things to be learned about the influence of small amounts of iodine on neonatal thyroid function in mild iodine deficiency, and about pituitary/thyroid feedback regulation in the fetus and small infant. A tendency to a slight increase in neonatal TSH level after iodine supplementation may be of little importance, as, in these infants, the serum concentration of T4 (which may be the major thyroid hormone influencing brain development12) does not show a concomitant reduction.11 Finally, iodine supplementation imposes no risk of worsening of postpartum thyroid dysfunction in the mother.13 In conclusion, pregnant women should not be iodine deficient. To strictly follow WHO guidelines on iodine intake, pregnant women with similar urinary iodine levels to those found by McElduff et al could alter their diet towards more iodine-rich foods, or they could take a small iodine supplement as part of the vitamin and mineral supplements recommended for pregnant women in most countries. However, there is at present no evidence that a supplement will have beneficial effects. Ideally, iodine intake should be evaluated and kept optimal in the entire population, taking into account that unnecessary high iodine intakes may be associated with more hypothyroidism.4 The studies by McElduff et al8 and other researchers7 demonstrate the need for national monitoring and adjustment of iodine intake as part of a program of prevention of thyroid disorders and their complications. Such initiatives normally involve government nutrition or public health agencies in collaboration with experts in thyroid diseases, nutrition and epidemiology and prevention.1,2 It would be an added bonus if the program elucidated some of the unresolved issues in the field of population iodine supplementation in developed countries. This would continue the considerable contribution of Australian scientists to the understanding and correction of iodine-deficiency disorders.3,14

Peter Laurberg MD, DMedSci · Susanne B Nøhr

In 1974, Caffey suggested the label "whiplash shaken infant" to describe infants who had subdural haemorrhages without evidence of external injury.1 Much has been learned about inflicted head injury since that time. In September 2001 a conference was held in Sydney to share this knowledge among people from many disciplines who work with, or are affected by, children thought to have been injured in this manner. It was jointly organised by the US National Center on Shaken Baby Syndrome, the Children's Hospital, Westmead, and the Sydney Children's Hospital, Randwick. Definition and controversiesShaken baby syndrome (SBS) is a form of child abuse that occurs when someone violently shakes an infant, most often an infant younger than six months, resulting in brain, eye and skeletal injuries.2 Subdural haemorrhage on CT scan is frequently used as a marker for SBS. In a 1999 population-based study in the United Kingdom, the incidence of subdural haemorrhage due to child abuse was found to be 21 per 100 000 in children under the age of one year and 12.8 per 100 000 in children under the age of two years.3 There is wide variability in the clinical presentation, ranging from non-specific symptoms, such as vomiting, to coma or death. In its less severe forms the diagnosis is often missed, being confused with viral illness or gastrointestinal upset.4 The long-term prognosis in survivors is very poor, with a high incidence of intellectual impairment, cerebral palsy, epilepsy and cognitive/behavioural problems.5 The mechanism of injury is inferred from a possible triad of signs: severe brain swelling and/or diffuse axonal injury; subdural/subarachnoid haemorrhage; and bleeding in the retina, in the absence of a history of significant accidental injury or other medical conditions sufficient to explain the findings. There may also be other evidence of abuse, such as rib or long-bone fractures.6 SBS is a well established diagnosis in paediatrics and paediatric neurosurgery, yet some doctors and lawyers express doubts about its validity. The important question arises as to whether it is possible to differentiate this abusive form of head injury from accidental head injury or disease. Is this constellation of injuries unique to shaking?Injury to the eye. Retinal haemorrhages are present in at least 80% of cases of SBS.7 The critical importance of the role of injuries within the eye, which distinguish inflicted head injury from accidental injury and disease, was emphasised. The importance of accurate description, in terms of number and types of haemorrhages and their distribution within the retina and other parts of the eye, was noted. Only SBS (not accidental injury or disease) can result in a pattern of multiple haemorrhages distributed throughout the retina to the periphery, especially if these are in the presence of preretinal, vitreous or subhyaloid haemorrhage. Haemorrhages in the eyes may be unilateral or absent, depending on the severity of the injury.7 Bleeding in the optic nerve sheath, retinal folds, retinoschisis and retinal detachment are highly associated with severe rotational forces.7 Examination by an ophthalmologist after pupillary dilatation is critical for diagnosis. Brain injury. The mechanisms of brain injury were reviewed. It was contended that abusive head injury results from severe rotational inertia injuries (due to shearing from acceleration–deceleration forces) and secondary (mainly hypoxic) injuries. The resultant symptoms and signs reflect the severity of these forces. Shearing injuries frequently lead to subdural haemorrhage and apnoea,8 with or without diffuse axonal injury; these disturbances can lead to cessation of breathing and concussion or prolonged traumatic coma.9 In contrast, accidental injuries resulting from common household falls cause mainly contact injuries (due to direct mechanical forces) or translational inertia injuries (due to linear acceleration–deceleration), which are usually not life-threatening.10 Most short falls do not reach the rotational velocity threshold to cause even concussion, let alone more serious injury. Short falls do not cause serious injury or death, except in most unusual circumstances, such as with extradural haemorrhage, mass-effect subdural haemorrhages, secondary effects of injury, or in falls from swings, which have a significant angular velocity component. Impact injuries sufficient to cause immediate serious injury or death are commonly associated with evidence of external injury and are seen with motor vehicle accidents and long falls. Subdural haemorrhage. The most common lesion seen in SBS is subdural haemorrhage (SDH). Trauma is the cause of virtually all SDH. There are conditions in which the brain does not fill the space available within the cranial cavity, and this may cause SDH to occur with lesser degrees of trauma than would normally be expected. Benign enlargement of the CSF spaces, which is physiological, has not been shown to predispose to SDH, whereas pathological conditions that enlarge the subarachnoid or subdural space, such as old SDH, post-traumatic hydrocephalus or atrophy, may predispose to haemorrhage with lesser degrees of trauma.11 Different densities seen within an SDH on x-ray may also cause confusion about timing of injury.12 The rapidity with which symptoms develop, the presence of acute brain injury and the presence of retinal haemorrhages will assist in correct diagnosis. Excluding alternative causes for medical findingsAll known disease states simulating abuse need to be excluded, but these are few. Coagulopathy is an important one. Accidental injury, including obstetric injury, needs to be excluded. Common forms of courtroom defence in cases of alleged SBS include that the constellation of injuries seen is due to (i) reaction to vaccination, or (ii) metabolic disorders secondary to vitamin or other deficiencies. However, large-scale studies have shown that vaccines never reproduce the findings seen in SBS.13 Vitamin C deficiency has been hypothesised as a cause for the signs seen in SBS, on the basis of bench research showing that deficiency may predispose to bleeding. But, even if it did, coagulopathy due to other causes does not reproduce the signs found in SBS.7 There is no disease or condition that fully mimics the complete diagnostic picture of SBS. Is impact required for serious or fatal injury?A single study, based on experiments with biomechanical dolls implanted with accelerometers, has suggested that impact is always necessary for serious brain damage or death to occur, but this is disputed.14 The authors of the study measured impact at the end of shaking. The thresholds used to predict injury were generated from adult primates subjected to single-impulse rotational events. There are no equivalent thresholds for shaking injury in adult or immature laboratory animals. The forces generated during whiplash-shaking are different from those seen in falls or other forms of impact. Evidence from the shaking of adults,15 together with numerous articles in peer-reviewed journals and confessions by perpetrators, refute the suggestion that impact is necessary for severe or fatal injury. PreventionEpidemiological research has shown that preventive efforts should target young men and daycare providers as well as parents, particularly fathers.16 Starling et al found that biological fathers inflict the injuries in 45% of cases, and the mother's boyfriend, with no paternal relationship to the child, inflicted the injuries in 25% of cases. Female babysitters and mothers were each responsible for 15% of cases. Another study found very similar results.17 The need to educate people about positive ways of dealing with crying babies was emphasised. "Dads 101", a program for new and expectant fathers that teaches them about the dangers of shaking babies as well as educating them on how to bond with their child, was presented as an example. ChallengesIt is clear that SBS is a preventable form of abuse and that it has dire consequences for the child, the family, the perpetrator, and society at large. Its recognition is important. The recognised episode is frequently not the first episode of shaking. Training of professionals, particularly doctors, to improve recognition of the milder manifestations of SBS would perhaps prevent later manifestations of severe injury and death. Prevention has been shown to work in pilot studies done in maternity hospitals in the United States. The challenge is to reach those most at risk. There needs to be a concerted education campaign involving all those who have the responsibility of caring for very young children. More research still needs to be done, particularly on prevention of shaking and on long-term management of children who have suffered SBS.

Kieran T Moran FRACP

Objectives: To determine the prevalence of stroke risk factors in a general practice population and to identify pharmacotherapies currently used in management of stroke risk factors.Design: Multicentre, observational study by 321 randomly selected general practitioners who each collected data on 50 consecutive patients attending their surgery.Patients and setting: 16 148 patients aged 30 years or older attending general practices across Australia during 2000.Outcome measures: Prevalence of hypertension, current smoking, diabetes, hypercholesterolaemia, atrial fibrillation, recent history of stroke or TIA; extent of pharmacotherapy use in risk-factor management.Results: 70% of patients had one or more risk factors and 34% had two or more. Hypertension was the risk factor with greatest prevalence (44%), followed by hypercholesterolaemia (43%) and current smoking (17%). The prevalence of risk factors generally increased with age, except for current smoking, where a decrease with age was seen. The most common pharmacotherapies were cardiovascular agents, followed by antiplatelet agents. Two-thirds of patients with hypertension were taking cardiovascular drugs, most commonly angiotensin-converting enzyme inhibitors.Conclusions: Stroke risk factors are highly prevalent in general practice patients and GPs are ideally placed for opportunistic case-finding. There is considerable scope for improving management of stroke risk factors. The Avoid Stroke as Soon as Possible (ASAP) general practice stroke audit provides a baseline against which progress in risk-factor management can be measured.

Jonathan W Sturm MB ChB, FRACP · Geoffrey A Donnan MD, FRACP · Stephen M Davis MD, FRACP · John G O'Sullivan MB BS, FRACGP · Miriam E Vedadhaghi BSc, PostGradDipNutr

Objective: To determine whether thyroid-stimulating hormone (TSH) concentrations in a large sample of neonates meet World Health Organization criteria for an iodine-replete population (< 3% of neonates with whole-blood TSH concentrations > 5 mIU/L), and, in a small subset of neonates, to examine the correlation between maternal urinary iodine and neonatal TSH concentrations.Design: Cross-sectional study of neonatal whole-blood TSH values obtained as part of a routine newborn screening program.Setting: Royal North Shore Hospital (RNSH) in northern Sydney.Participants: Two anonymous samples of neonates born at RNSH (1316 infants born between August 1998 and April 1999 and 1457 infants born between 1 March and 31 December 2000); and 84 infants whose mothers had attended RNSH between September 1998 and August 1999 and supplied a urine sample for iodine measurement.Main outcome measures: Iodine status of neonates (proportion with whole-blood TSH values > 5 mIU/L), and urine iodine concentrations of pregnant women.Results: In the two large population samples of neonates, 8.1% (95% CI, 6.6%–9.5%) and 5.4% (95% CI, 4.3%–6.6%), respectively, had whole-blood TSH values > 5 mIU/L (prevalence range for mild thyroid deficiency, 3%–19%). Comparing the TSH values of the 1316 anonymous infants and the 84 identified infants showed no difference between the proportions with TSH values > 5 mIU/L (8.1% v 10.7%, respectively; P = 0.39). Urine iodine concentrations in the 84 pregnant women indicated borderline mild iodine deficiency. TSH values in their 84 infants were positively correlated with maternal urine iodine concentrations.Conclusions: Our results suggest that the population of northern Sydney may have mild iodine deficiency. However, the expected relationship between maternal urine iodine levels and neonatal TSH concentrations was not found.

Aidan McElduff PhD, FRACP · Patrick McElduff BMath, PhD · Jenny E Gunton MB BS, FRACP · Graham Hams MAppSc · Veronica Wiley PhD · Bridget M Wilcken MB ChB, FRACP

Objective: To determine the prevalence of colorectal neoplasia detected by rescreening people with average risk five years after initial screening by flexible sigmoidoscopy.Design: Prospective survey of results of a colorectal cancer screening program.Participants: People aged 55–64 years with no symptoms or family history of colorectal cancer who were recruited from the community for flexible sigmoidoscopy screening five years previously (July 1995 to December 1996) and had no colorectal neoplasms detected.Setting: Fremantle Hospital, Western Australia, a community-based teaching hospital, December 2000 to June 2001.Main outcome measures: Number and size of colorectal neoplasms (adenomas or cancer) compared between rescreened patients and initial screening population (all 982 people screened between July 1995 and December 1996).Results: 803 people were eligible for rescreening; 138 were no longer at the recorded address, and 361 of the remaining 665 (54%) were rescreened. Rescreening found a significantly lower prevalence of colorectal adenomas than initial screening (8% [95% CI, 5%–11%] versus 14% [95% CI, 13%–15%]; P < 0.05) and also a lower percentage of adenomatous polyps over 5 mm in diameter (32% [95% CI, 15%–49%] versus 51% [95% CI, 46%–56%]; no significant difference).Conclusion: Average-risk people who have been screened for colorectal neoplasms, with none found, have a low prevalence of neoplastic lesions five years later. Longer rescreening intervals need to be considered.

Cameron F E Platell PhD, FRACS · Gillian Philpott EN · John K Olynyk MD, FRACP

Australia has been experimenting with fundholding in primary care for nearly a decade. When the concept was floated in 1992,1-4 the debate was heavily influenced by the problems experienced in the United Kingdom with fundholding in general practice, including lack of measurement of improvement in quality of care.5 Fundholding still engenders disquiet because of its potential for a primary focus on cost savings, increased control of clinicians by management,6 and a reduction in quality of care7 and equity of access.8 Over the past five years, there have been 14 trials in Australia that included a fundholding model: nine General Coordinated Care Trials (GCCTs) (one of which comprised four subtrials) (Box 1);9-17 four Aboriginal and Torres Strait Islander Coordinated Care Trials (ATSI CCTs) (Box 2);18 and the Maitland After-Hours Care Trial (MAHT) (Box 3).19 Such trials are expected to continue, and the debate about the impact of fundholding on patient outcomes remains unresolved. It is therefore an appropriate time to review the Australian experience of fundholding in primary care. Here, we present the results, focusing on the question: Does the Australian experience of fundholding thus far provide evidence of improved patient health and well-being? Key conceptsFundholdingFor this article, fundholding is defined as a framework within which specified resources, agreed prospectively, are made available for a defined period, and from which a range of services are provided to a specific group of patients. It is a framework for funding a healthcare initiative rather than an initiative in itself. It overcomes the constraints that the existing funding structure and availability of funds impose on a healthcare initiative by introducing flexibility within an overarching healthcare system in relation to the fundholder(s), the funds pool (budget), and the economic benchmark. The fundholderThe fundholder is the organisation holding the funds. It may be established specifically for that purpose or be part of an existing regional body. Typically, the fundholder considers a broader range of patient management factors than would be possible under existing structures. The role and objectives of the fundholder are critical to the success or otherwise of fundholding. Compared with existing fundholders, a new fundholder may bring a very different perspective to the healthcare of a group of patients. The funds poolThe funds pool comprises the resources managed by the fundholder to support the services and infrastructure otherwise financed by more than one program or budget (eg, Medical Benefits Scheme [MBS], Home and Community Care [HACC], hospitals, and Pharmaceutical Benefits Scheme [PBS]). The economic benchmarkThe economic benchmark is the guide to "how much" ought to be contributed to the funds pool by the various financial stakeholders. In the Australian trials, the benchmark was usual care, or usual care plus a specified additional payment (eg, ATSI CCTs18 and MAHT19). It is also possible to base the economic benchmark on the resources required for a specified group of services (eg, in relation to a care protocol for a patient recently diagnosed with diabetes, or best-practice antenatal shared care).20 Review of the Australian experienceIn Australia, a dominant feature of fundholding discussions is that improved health and well-being of patients, rather than cost containment, should be guiding its implementation. Initiatives so far in Australia have included: care planning; care coordination; after-hours care; care for people with chronic and complex needs; hospital admission prevention or improved discharge programs; and Aboriginal community empowerment. All these initiatives were aimed at improving the care offered to the people enrolled. The GCCTs9 and ATSI CCTs18 were underpinned by care plans and evidence-based guidelines, and the MAHT by improved patient access and satisfaction.19 It is worth noting that care coordination was originally intended to be undertaken by a skilled professional, not necessarily a GP. The change in federal government in 1996 resulted in a decision by the new health minister that these trials should involve GPs in the care-coordination models, as they were the "central focus of primary health care" (see Chapter 5 — Care planning and coordination).9 ReviewIn reviewing the evidence for improved patient outcomes with fundholding, we consulted both published and unpublished reports of the trials, and also interviewed organisations and individuals involved in the trials themselves or in their evaluation as trial managers, trial sponsors, financial stakeholders or evaluators. Boxes 1, 2 and 3 summarise the trials — the healthcare initiative, the fundholding model, and the outcomes. Given the complexity of the trials, it was not possible to attribute any gains in patient outcomes directly to fundholding. Thus, our review distinguished between three types of outcomes: The outcomes of the initiative itself (eg, did improved coordination of care lead to improved patient health and well-being?); The effectiveness of fundholding's contribution to achieving these broader program initiatives (eg, did fund pooling contribute to improved coordination of care?); and The operational outcomes of fundholding (eg, was the program in operating surplus or loss at the end of an agreed period?). FindingsOur findings suggest that fundholding has a role in overcoming the constraints imposed on specific healthcare initiatives by the Australian healthcare system. However, the relationship between fundholding and patient health and well-being is largely dependent on the objectives and effectiveness of the overall initiative. Four main themes emerged with regard to evidence of the relationship between fundholding and patient outcomes: evidence of impact on patients, role of fundholders, institutional acceptance, and information systems. Evidence of impact on patientsThere was limited quantitative evidence of reduced morbidity and fewer admissions to hospital. Although it would be inappropriate to attribute health gains across each of the components of the coordinated care intervention (care plans, care planners and fundpooling), some important service provision changes were facilitated by fundholding, and, supported by the qualitative evidence, had significant implications for patient well-being. In the GCCTs, community service use increased for the trial clients in trials pooling these services. (Individuals enrolled in the trials were referred to as clients rather than patients.) There was also limited evidence of improvements in well-being, as assessed by SF36 (36 questions covering eight domains of mental and physical well-being),21 and some evidence of reduced hospital admissions and readmissions.9 In the MAHT, there was an increase in after-hours access to medical care.19 In the ATSI CCT, improved access to coordinated services,18 and a more coordinated approach for patients with complex needs, was accompanied by community empowerment and capacity building.18 Additional money (Commonwealth-equivalent MBS/PBS funding) led to significant improvements in the range and quantity of services available to the communities. Role of fundholdersThe types of organisations acting as fundholders varied across the trials. In the ATSI CCTs, community-controlled boards determined how the pooled funds would be used,18 and in MAHT the GP Division managed the funds.19 In the GCCTs, the fundholders included the Department of Human Services in one State, regional health services, and specifically constituted organisations.9 There is substantial evidence of the significant role the fundholder played in the GCCTs and ATSI CCTs. In the ATSI CCTs,18 the fundholders, mostly community representatives, replaced government organisations as key decision-makers for funds covering a broad range of services. The fundholder had three main responsibilities: the financial viability of the program (ensuring there was no operating loss); engaging stakeholders in the process (both providers and fundholders); and achieving the objectives of the overall initiative (eg, patients' health and well-being). A critical role for fundholders was balancing the trade-offs between additional services for patients and the program's financial position. The evidence from these 14 trials is that future fundholders will need training and skills to manage this difficult and critical task. The most appropriate fundholder organisation will depend on the nature of the initiative and the type of stakeholder engagement required. Institutional acceptanceStakeholders' willingness to engage in fundholding varied across the trials and stakeholders and was influenced by at least three factors. Shifts in resources — the pooling of financial stakeholders' resources in one budget leads to increased engagement across institutions in resource-allocation decisions, but shifts in resources away from a contributor's institution as a result of fundholding can reduce willingness to participate. Availability of savings — in most fundholding frameworks, providers of care are encouraged to make decisions leading to financial savings (efficiency gains). There is a greater likelihood of providers generating these efficiency gains if the savings remain available to the initiative. Workforce implications — ensuring that the workforce implications of changing the mix of patient services are considered requires broader institutional engagement than can be achieved through fundholding alone. In the ATSI CCTs, it was noted that, without additional workforce expansion and training, fund pooling alone would not bring sustainable health system change.18 The role of information systemsInformation technology (IT), in particular the timely availability of accurate data on patients' service use, care plans and associated expenditure, played a significant role in the effectiveness of fundholding in the GCCTs. These data included both historical data on service use by patients (used to estimate the size of the fund pool) and service use throughout the trials. Unless it was clear to fundholders which services were being used by the trial patients, it was difficult to monitor quality of care and ongoing financial trial viability. The data collection and information technology (IT) framework for monitoring the funds pool, establishing the economic benchmark and evaluating the trials, was time consuming and expensive. If fundholding is to successfully improve patient care, the most efficient fundholder is unlikely to reach these goals without excellent IT systems, and the second round of the GCCTs has recognised IT system development as a key requisite for effective care coordination and fundholding. Next stepsWe identified three necessary steps before fundholding can be adopted more widely: More formally evaluated trials of a greater range of initiatives focused on specific health and well-being outcomes; Improved prospective analysis of the likely effectiveness of fundholding on both a general and case-by-case basis; and Informing the broader health community. It is important that future trials, including the second round of CCTs and their evaluation, are designed to provide evidence of the impact on patient health and well-being and the role of fundholding in facilitating any such gains. Also needed are rigorous analyses of how a particular initiative could be improved by including a fundholding framework. Such analyses were completed by participants in the GCCTs and informed decisions about funding of the trial initiatives, the contributions to the funds pool, and the ongoing viability of fundholding. The simulations supporting the second round of CCTs are more accurate, as these trials have access to improved data sources and better estimates of costs per patient (capitation rates). Despite these improvements, the estimates of possible efficiency gains in coordinated care (eg, reductions in hospital admissions and MBS service use) seem too optimistic. Preliminary work on the application of such simulations has begun with Shared Care Obstetrics20 and GP fundholding.22 Evidence from other countries23,24 can inform the debate, but should be analysed from an Australian perspective. Informing the broader health community of the potential benefits and limitations of fundholding is important. Time and patience are needed for some clinicians and managers to change their perspective and embrace the concept.25 In MAHT, there was a need to foster relationships with existing providers (eg, emergency physicians) and manage the change in these relationships.19 In both the ATSI CCTs18 and the GCCTs,9 appropriate engagement of the community was crucial to develop transparent and accountable processes for allocating the funds and establishing relationships with service providers. 1: General Coordinated Care Trials (GCCTs)*9-18 The initiative The first round of GCCTs comprised nine trials between June 1997 and December 1999. The overall objective was to improve health and well-being through improved coordination of care and care planning. The intended target group was people with chronic illness and/or complex needs; however, people actually enrolled included those currently without such needs, but with historical or anticipated needs for greater care, particularly before and after hospital admissions. The trials involved 5580 control and 10 953 intervention clients (individuals enrolled in the trials were referred to as clients rather than patients) and a range of healthcare professionals providing, in addition to the usual services, care planning, care coordination and service coordination. The trials developed a range of models of care coordination, all of which involved a GP in developing the medical component of the care plan. Beyond this, the role of the GP varied across the models, with most trials employing care and service coordinators. Fundholding model The objectives of fundholding were: to remove those barriers patients and providers have in accessing appropriate services created by the program-based funding structure — referred to as "removing labels from dollars"; and to ensure savings generated by the efficiency gains of care coordination were used to fund services for clients or infrastructure for the initiative (eg, if care coordination reduced the need for a hospital admission, these savings were made available to the trial). The economic benchmark for the fund pool was usual care (ie, the dollar value of care patients were expected to access during the trial). The estimates of the funds pool for each trial were based on historical use (Medical Benefits Scheme [MBS], Pharmaceutical Benefits Scheme [PBS], community services and hospital inpatient costs), adjusted for the expected effects of ageing and increased illness severity and expected risk of hospitalisation. The range of services pooled varied, but always included MBS, PBS, and hospital inpatients. Some trials included Health and Community Care (HACC), Department of Veterans' Affairs and Royal District Nursing Service (RDNS). The fundholders varied across trials and included the Department of Human Services in one State, regional health services and specially constituted trial organisations. In two cases, Divisions of General Practice were fundholders for part of the overall budget relating to payments to GPs for care coordination services and strategies involving GPs. Outcomes Care coordination and care planning outcomes: There was general consumer satisfaction with care coordination services and care planning, and this tended to be higher for consumers who had a greater need for coordination of their services. The highest exit rate was from a trial that enrolled people with limited potential to benefit. No trial had a "meaningful" gain in health and well being as measured by the SF36, for intervention compared to control, although three subtrials showed small gains (see Chapter 11 — Client health and well-being).9 There was limited evidence of a reduction in hospital admissions, with reasonable evidence from only one trial with a randomised control (see Chapter 18 — Impact on service utilisation and expenditure).9 Trials that directed a substantial portion of their budgets to increased RDNS and HACC services to clients who could benefit from these reported that there was an improvement in client well-being. Fundholding as a contributor to the effectiveness of care planning and coordination: Fundholding allowed trials to fund strategies and infrastructure that would otherwise not have been possible. Clients' community service use increased for trials that pooled such services. At the conclusion of the GCCTs, there was limited evidence that there had been reduced use of hospitals, MBS and PBS. There was limited evidence that trials met the expectation that they could generate sufficient savings in some areas (eg, PBS and hospital admissions) to fund coordinated care activities and additional community services (see Chapter 16 — Financial and economic experience; and Chapter 18).9 (Some trials made savings in relation to specific programs [eg, PBS], but this was often because the initial contribution was an overestimate of expected use.) Fundholding as a funding model: The funds pool for each trial varied from $7.5 million to $28.4 million. This comprised infrastructure (range, 23%–42% of total) and service income. The variation across trials was a result of client numbers, their rates of leaving the trials, the range of services pooled, whether the enrolled clients were high or low users of services, and the extent to which infrastructure funding could be attracted from outside the core Commonwealth grants. Three trials did not have an operating loss, partly as a result of overestimates of usual care and the subsequent funds pool size, as well as infrastructure contributions. Funds pool estimates were generally poor (between plus or minus 25% from the best estimate of usual care). In a number of trials, it was very difficult to obtain an estimate of what the costs of care would have been in the absence of coordinated care. * Caveat: There were nine distinct trials (one of which contained four subtrials) and the generalisations made in this summary are not necessarily representative of the experience of each trial. For more details, see the local evaluations of each trial or the national evaluation.9-18 2: The Aboriginal and Torres Strait Islander Coordinated Care Trials (four trials) (ATSI CCTs)18 The initiative These trials were established to "achieve a more coordinated approach to the delivery of health care services to people with a diverse range of complex health needs".18 In contrast to the GCCTs, these trials had another series of aims related to community empowerment and capacity building, with the aim of improving the health status of Aboriginal communities. Fundholding model The four trials ran from 1997 to 1999. The actual trial sites are given below. Trial Clients Fund pool (in millions) Infrastructure and sponsorship funding (in millions) Katherine West (NT) 2012 $5.6 $1.1 Tiwi Islands (NT) 1818 $7.6 $1.4 Wilcannia (NSW) 780 $4.7 $1.5 Bunbury/Perth (WA) 1990 $5.6 $1.2 The fund pool included Commonwealth-equivalent MBS/PBS funding, existing local health services and, in some cases, hospital resources. Importantly, additional funds above usual care were provided in the form of Commonwealth-equivalent MBS/PBS funding. Each trial received $599.70 per client, either enrolled or in a specified community. This represented per capita (Australia-wide) use of MBS and PBS which was significantly greater than historical use in these regions. The fundholder varied from community-based boards to the Health Department of Western Australia and a local Health Service. The policies and strategies of the latter group were determined by a "community working group".18 Outcomes Care coordination and care planning outcomes: There was improvement in service access and flexibility (by the use of the funds pooling) and improved appropriateness of service provision (by the use of care coordination). Fundholding as a contributor to the effectiveness of care planning and coordination: There was development of the organisational structures required to implement fundholding; evidence that fundholding could be an effective mechanism for allocating "new" resources in the MBS/PBS equivalent funding; and improved individual and community empowerment. There was an opportunity to develop substantial infrastructure, including financial infrastructure and information systems. Fundholding as a funding model: All trials operated within their financial budget, and all had unexpended funds as a result of several factors, including the delay in recruiting program staff required to implement agreed strategies, and the cautious approach to expenditure adopted by the fundholders. MBS/PBS = Medical Benefits Scheme/Pharmaceutical Benefits Scheme 3: Maitland After Hours Primary Care Trial (MAHT)19 The initiative In response to an identified and quantified need for improved after-hours care in the Maitland area of the Hunter region, a comprehensive trial with four elements was established in October 1999. An after-hours triage service; A service provided by GPs (Maitland After Hours GP Service [MAGS]) in the Maitland Hospital Emergency Department A funded transport service; and A home visiting service. The two main needs met by MAGS were to ease the patient burden on the Maitland Hospital Emergency Department, and to lessen the difficulties faced by local GPs in meeting patients' after-hours care. A key element was organising the after-hours GP workforce more effectively. Fundholding model The trial involved pooling funding from two Commonwealth sources (Medical Benefits Scheme, Practice Incentive Payment [$457 000] and Commonwealth infrastructure funding [$279 000]) and the Maitland Hospital (staff, pathology, investigations, pharmaceuticals and consumables [$200 000]), making the total fund about $936 000. The economic benchmark was estimated usual after-hours care for the region across GP and Maitland Hospital Emergency Department services, with an additional contribution by the Commonwealth to fund the program's establishment. The fundholder was the Hunter Urban Division of General Practice, but the Board of Management for MAGS included Area Health Service personnel. Outcomes After-hours care service: Overall, the GPs and all stakeholders were very happy with the service and wanted it to continue. In all, 11 315 patients were seen at MAGS, 2093 telephone calls were handled and 103 home visits completed. There was a 61% decrease in Category 4 and 5 patients (the two least urgent triage categories) seen by the Emergency Department staff during the time MAGS was operating. From a survey in February – April 2000 of 81 users of the service (34% response rate), 79% rated the "overall quality of the service as excellent or very good (5% said it was fair/poor), 91% said they would return to the service for medical care, and 89% would recommend it to their family or friends". At a follow-up interview in August – September 2000 of 422 people who had used the service, 89% agreed they were able to get the care they wanted, 84% were happy with the doctor they saw and 35% felt they had to wait a long time. There were 12 formal complaints. Fundholding as a contributor to the effectiveness of an after-hours service: Fundholding provided an alternative model for paying GPs (salaried rather than fee-for-service); It provided the additional funding above usual-care costs to establish and maintain the alternative model; and It engaged stakeholders in the process of changed models of care, partly as a result of the transparency of the model in terms of its financial and patient-load impact on the various providers. Fundholding as a funding model: The model was financially viable (ie, within budget). Payments were timely and initial infrastructure funding was sufficient for startup. A key to successful implementation was that the Area Health Service remained committed to the trial and facilitated dealings with Maitland Hospital. This was important, as the reduction in patient load clearly had resource implications for Maitland, but the exact extent of the impact on Emergency Department resources was difficult to estimate, partly owing to the cost structure of emergency departments.

Justin J Beilby MB BS, MD · Brita Pekarsky BEc(Hons), DipHEc

Riding waves on a bodyboard, or boogie board (bodyboarding), is a popular seaside pastime. Abdominal injuries have not previously been reported from this activity. We report the cases of three adolescents who sustained major blunt abdominal injuries while bodyboarding. The three patients were seen between February 1998 and March 1999 at hospitals serving large beachside communities in Queensland. The circumstances of the accidents suggest that the method in which a bodyboard is usually ridden places a rider at risk of abdominal injury. Clinical recordsPatient 1A 14-year-old boy was "dumped" by a wave onto a sandbank, causing his board to be driven against his upper abdomen. Over the next few hours he experienced worsening left-side abdominal pain. On presentation, six hours after the injury, he was pale and in pain. His pulse was 120/minute and blood pressure was 110/60 mmHg. He had marked tenderness in the left upper abdominal quadrant with guarding. A computed tomography (CT) scan of the abdomen revealed a full-thickness laceration of the spleen, extending into the hilum (Box 1). The patient was managed by active observation, and did not require surgery or blood transfusion. Subsequent scans revealed healing of his injury by three months. Patient 2A 12-year-old boy presented with abdominal pain 16 hours after being hit from behind by a large wave while attempting to ride it to shore. The nose of his board was caught on the sandy sea floor, driving one of the base corners of the board into his upper abdomen. On examination, he was pale with a pulse of 130/minute and blood pressure of 100/60 mmHg. His abdomen was tender in the right upper quadrant with guarding. A CT scan of the abdomen revealed a large laceration to the right lobe of his liver, with a significant amount of free intraperitoneal fluid (Box 2). The patient was observed closely in hospital for one week without adverse sequelae. Subsequent ultrasound scans showed resolution of his injury by four months. Patient 3A 15-year-old girl was thrown off balance by a wave while riding her bodyboard. She immediately experienced abdominal pain, which became increasingly severe. She recalled the board making contact with her abdomen during the fall. On presentation, within two hours of the fall, she had severe abdominal pain, right shoulder tip pain and light-headedness when standing. On examination, her pulse was 130/minute and blood pressure was 90/50 mmHg. She had signs of generalised peritonitis and there were no bowel sounds. An ultrasound scan of the abdomen revealed a large amount of free fluid in the peritoneal cavity. No definite injury to the liver, spleen or kidneys was detected. As gastrointestinal tract perforation could not be excluded, a laparotomy was performed. This revealed 500 mL of fresh blood in the peritoneal cavity as a result of a superficial liver laceration. Haemostasis was achieved initially by packing, and then by the application of topical thrombostatic agents. No other significant injuries were noted. The patient made a very good recovery and was discharged home four days after admission. DiscussionThere are very few reports of injuries associated with bodyboarding. A literature review revealed only one report of injuries (facial) sustained during this activity.1 There are anecdotal accounts of injuries as a result of bodyboarding misadventures, including severe spinal injuries, on bodyboarding-related forums on the Internet.2 Bodyboards can also pose a hazard to other swimmers in the surf.3 A bodyboard rider lies with the upper part of the body on the board, parallel to the sea floor, placing the base of the board against the ventral trunk. Should the rider be thrown off balance and the nose of the board strike or be caught on the sea floor or sandbank, the base of the board can be driven with considerable force into the upper abdomen (Box 3). Although most boards have a concave curvature of the base, the two corners can potentially inflict significant blunt injuries, as evidenced by Patient 2. As they lie parallel to the sea floor, riders are also less able to protect themselves when thrown onto the sea floor or a sandbank. Injuries from conventional surfboarding are well documented.4-6 Lacerations make up a high proportion of surfboard-related injuries,5 reflecting surfboard design and configuration (sharp nose, fins). Blunt trauma would be expected to predominate in bodyboarding, as bodyboards have less sharp edges and have a relatively softer consistency. Bodyboarding is also more popular, especially among children and teenagers. Although many bodyboarders are very skilled (a professional circuit exists), bodyboarding is an activity in which most beachgoers can partake. The boards are readily available, and less skill and swimming ability is required to catch a wave than in conventional surfing. As there have been no previous reports of abdominal injuries associated with bodyboarding, we are unable to determine how commonly they occur. Our experience may be unique, but, given the popularity of the pastime with young people, we suspect that this association may be under-reported. Bodyboarders, as well as their medical attendants, need to be made aware of the potential for abdominal injuries during what most would consider a benign beachside activity. Altering the design of boards, such as rounding off the base corners or increasing the pliability of the core material, might help prevent such injuries occurring. However, this suggestion may not be welcomed by bodyboarders themselves, as it might be impossible to make such alterations without adversely affecting the performance of the boards. 1: Computed tomography scan of the abdomen of Patient 1, a 14-year-old boy A full-thickness laceration of the spleen, extending into the hilum, is visible. 2: Computed tomography scan of the abdomen of Patient 2, a 12-year-old boy A large laceration of the liver is visible, with a significant amount of free intraperitoneal fluid. 3: Proposed mechanism for abdominal injury while bodyboarding

Kelvin L Choo MB BS, FRACS · Deborah M Bailey MB BS, FRACS · John B Hansen MB BS, FRACS

The majority of the "Australian COX-2-Specific Inhibitor (CSI) Prescribing Group" endorse the following points: CSIs are equivalent to non-steroidal anti-inflammatory drugs (NSAIDs) as anti-inflammatory agents. CSIs and NSAIDs modify symptoms but do not alter the course of musculoskeletal disease. CSIs do not eliminate the occurrence of ulcers or their serious complications, but are associated with considerably fewer peptic ulcers, slightly fewer upper GI symptoms and, according to published reports, fewer serious upper GI complications, notably bleeding, than NSAIDs. CSIs and NSAIDs have similar effects on renal function and blood pressure. Whether any CSI poses a risk to cardiovascular safety remains subject to debate. Comorbidities and coprescribed drugs must be considered before initiating CSI (or NSAID) therapy. Patients prescribed CSIs (or NSAIDs) should be reviewed within the first few weeks of therapy to assess effectiveness, identify adverse effects and determine the need for ongoing therapy

The Australian COX-2-Specific Inhibitor (CSI) Prescribing Group*

There is persuasive anecdotal evidence that healthcare professionals have a heightened vulnerability to being stalked by their patients. This is supported by an overrepresentation of healthcare professionals in stalking-victim populations.1 Some stalking behaviours constitute little more than minor irritations, but more serious cases can ruin a clinician's career. Rarely, but tragically, stalking has cost some doctors their lives. Stalking refers to a constellation of behaviours encompassing repeated and persistent attempts to impose unwanted communications or contact upon another. While legal definitions of stalking do not specify time periods, it is increasingly apparent that harassing behaviours extending beyond two weeks constitute clinically significant stalking.2 The behaviours include telephone calls, letters, email, unsolicited gifts, following, maintaining surveillance, making complaints of professional misconduct or initiating spurious legal action against the victim, ordering or cancelling goods or services on the victim's behalf, property damage, threats, and physical or sexual violence. Stalking has emerged as a significant social problem3 and now constitutes a specific criminal offence in many jurisdictions. Although the past decade has seen a virtual explosion of interest in this phenomenon, only in recent years has stalking begun to pique the interest of the medical profession. This attention relates in part to the observation that stalking behaviours are becoming more prolific3 and may be amenable to clinical intervention.4 Although the extent of the problem has yet to be quantified, it is timely to highlight pertinent issues relating to the stalking of medical practitioners by their patients. We searched the Index Medicus/MEDLINE and PsycLIT databases for articles pertaining to stalking and related behaviours perpetrated against medical practitioners by their patients, and we review the literature in this article. We also present strategies that can be used to discourage a patient's intrusions and minimise the impact of these behaviours on clinicians' professional and personal lives. Because there is a lack of empirical data on the efficacy of specific interventions, these management strategies are based on our experience of assessing and treating stalkers and their healthcare practitioner victims during the past decade. Harassment and stalking in healthcare settingsWorkplaces can be risky environments, as burgeoning reports of violence, bullying and sexual harassment attest.5 Clinicians cannot claim immunity from these abuses.6-8 One survey of 850 forensic psychiatrists found that 42% of the 480 respondents had been harassed in some way by patients: 9 17% reported threats of physical harm, 13% reported non-violent threats (such as a lawsuit), and 3% had been physically assaulted. A study documenting the nature of threats to clinicians10 observed that intimidation could continue after the patient's discharge from inpatient care. Clinicians attending the 1994 annual meeting of the Oregon Psychiatric Society were surveyed about stalking.11 Twenty-six of the 90 respondents had been subjected to behaviours that met the authors' definition of stalking, and 37 reported other forms of distressing intrusions. In Australia, healthcare professionals were overrepresented in a sample of 100 self-referred stalking victims (nine general practitioners, three psychiatrists, two gynaecologists, one rheumatologist, a medical resident, a psychologist, a nurse and an occupational therapist).1 The vulnerability of medical practitioners to the unwanted intrusions of their patients is one legacy of a profession that comes into regular contact with lonely and disordered people. In some of these individuals, "sympathy and attention is easily reconstructed as romantic interest".12 Despite this, little systematic research has been conducted on the abnormal attachments of patients to their doctors, and it is likely many incidents of stalking remain unreported. This paucity of research relates in part to the nature of the profession. The authors of one study suggest that some staff may not have reported stalking behaviours because "denial and minimisation are common reactions to being the target of patients' aggressive behaviour".13 Denial enables clinicians to ignore threats and continue their work.10 Some doctors may fear that their victimisation will be equated with incompetence, or, in the current climate of sexual misconduct allegations, that their complaints of harassment will be met with scepticism. In popular depictions of doctor–patient liaisons, the emphasis is usually on the doctor exploiting the patient, with an implication that the imbalance in power precludes the patient from victimising the doctor.8 Resistance within the criminal justice system to prosecuting patients, especially when the patient is mentally ill,10 may also discourage reporting. Patients who stalkStalking behaviours are the product of a number of different motivations and states of mind.14 The two most common motivations for stalking clinicians are: Patients' developing romantic or child-like attachments that lead them to entertain hopes of a relationship. These hopes may arise on the basis of delusional beliefs (as in erotomania, a false conviction that one is loved by another), misplaced expectations in the lonely and desperate, or simply the unrealistic expectations of a would-be suitor. The last group is typically drawn from the socially incompetent or narcissistically self-absorbed. Patients' developing a resentment against the professional, usually connected to some supposed injury or dereliction. It has long been recognised that patients may transfer feelings of love or hate to their therapist as a consequence of the treatment situation and early experiences, a phenomenon known as transference.15 However, transferential explanations for patients' misplaced affections or anger foster in healthcare professionals an unhelpful sense of responsibility that can lead to feelings of guilt. They also foster the counterproductive assumption that the stalking and harassment can be managed within the doctor–patient interaction. The impact of stalking on healthcare professionalsIn one report, the authors noted that the harassing behaviours of their subjects had a disruptive impact on hospital staff, and consequently the functioning of the hospital.13 Staff became more vigilant and modified aspects of their work practices and lifestyle to ensure their safety. Some even changed their place of residence. Another study also highlighted the disruption created in the professional and personal lives of physicians who experienced threats.10 Examples included modifying their treatment style or refusing certain referrals, reducing their practice size, and placing restrictions on family activities. Some expressed guilt for the intense family stress that resulted. Many of the doctors who participated in the Australian survey1 said that nothing in their medical training had prepared them for this and they had emerged feeling frustrated, helpless and disenchanted with their profession. Management strategiesMany doctors have, or feel they are expected to have, a high tolerance for the criminal behaviour of patients, even when it induces fear and disrupts the lives of practitioners and their families. We believe such tolerance is misplaced, both in terms of the clinician's health and the patient's long term interests. The following strategies (summarised in the Box) can be used to discourage a patient's unwanted intrusions and to minimise their impact on all concerned. The first threat or declaration of loveStalking behaviours can emerge gradually out of apparently appropriate behaviours, such as attending more frequently at the doctor's rooms, phoning ostensibly for clarification of some aspect of treatment, or "chance" encounters outside of the work context. Alternatively, they may have a sudden onset, heralded, for example, by a declaration of love. Clinicians often respond awkwardly to these sudden declarations, giving polite disclaimers of interest on the basis of the professional nature of the relationship. All too often this results in patients' discharging themselves from the practice and returning, hopeful that now all obstacles have been removed. Worse still, the clinician may suggest an intimate relationship is precluded on the basis that they already have a partner. The patient might hear this as, say, "I would be interested but for my wife", with potentially dangerous repercussions. Stalking behaviours will be more readily recognised and their emergence may be subdued when clinicians set clear boundaries for appropriate behaviour in their patients.16 It is important to state that their relationship is, and will only ever be, a professional one. While this message should be firm and unequivocal, the clinician must endeavour to preserve the patient's dignity, lest they give the patient further reasons, particularly anger and resentment, to continue their harassment. The early stages of stalkingIt is reasonable to confront the patient and set limits by restricting the duration and frequency of appointments and any other behaviours that are aimed at maximising contact with the doctor. However, this approach has most chance of succeeding with the socially incompetent would-be suitor. Stalkers driven by morbid infatuation or resentment seldom respond to these measures alone. If a patient's intrusions continue despite the clinician's warnings, particularly when the behaviour persists beyond two weeks, there is a substantial risk of protracted harassment.2 In these circumstances, the clinician is ill-advised to continue the therapeutic relationship and should resist any temptation to engage in endless debates and negotiations. Ongoing communication will only reward the patient's efforts to maintain contact. However, transferring such patients can be difficult,11 both because of medicolegal abandonment implications and because prospective physicians may be understandably wary of accepting them. It is essential to document the termination process. In some instances it may be necessary to defuse the situation by hospitalising the patient. It is essential that the clinician carefully document each incident, noting the time and date, a synopsis of the behaviour and the names of any witnesses. These records should be kept in a secure place, as they can be invaluable in any future legal proceedings. The victim should also retain any concrete evidence of the stalking, such as answering machine messages, gifts and letters. Correspondence should not be returned, as this perpetuates contact between victim and stalker and vital evidence is lost. Affected professionals should not suffer in silence. They should inform a trusted colleague and discuss the case with a psychiatrist familiar with these issues. Other staff at the doctor's practice, particularly receptionists, must be apprised so that they do not inadvertently countenance the patient's pursuit and to enable them to adopt personal safety precautions. Spouses and other family members should be informed for similar reasons. Established stalkingStalking is a criminal offence and cases that do not respond to the above measures warrant a police report. This is particularly important when stalking involves threats or violence, or where there is a known history of sexual or physical aggression. Morbidly infatuated stalkers generally have a low incidence of threatening and violent behaviour, and, although resentful stalkers often issue threats, they seldom proceed to actual personal violence.14 Nevertheless, threats should never be disregarded, particularly in a stalking context, and the clinician must take precautions to ensure his or her personal safety and the safety of others who may be at risk. If the stalker threatens to initiate spurious legal action against the practitioner, or there is a likelihood of future action, the doctor's medical defence organisation should be alerted. Restraining orders or protective injunctions are frequently advised in stalking situations, but are not universally effective.12 Erotomanic patients are likely to be impervious to legal sanctions, although these are often useful in dissuading the socially incompetent and less entrenched resentful stalkers. Erotomanic stalking patterns will not subside without definitive treatment, and referral for psychiatric evaluation is a priority. It must be stressed that any escalation in stalking behaviours, particularly when it involves threats and overt hostility, should not be ignored or "worked through" in therapy. Threats are indications to seek advice. If a patient poses an immediate risk of harm to the therapist, he or she may require civil commitment or, for patients who are not mentally ill, prosecution under anti-stalking statutes.17 Anti-stalking strategies Don't disclose personal details to patients. Don't leave personal information where patients can access it. Don't give home contact details to professional organisations that allow public access to the information. Beware of confidentiality and privacy issues associated with electronic patient data. Consider security measures: obtain an unlisted home phone number and post office box address; install deadlocks, window locks, peepholes and exterior motion sensor lights; trim trees and shrubbery. Ensure patients understand your relationship is a professional one and will never be otherwise. If concerned about a patient's intrusions, set limits on proximity-seeking behaviours and ensure the patient is seen only when other people are nearby. Carefully document the patient's intrusions. Retain all evidence of stalking. Inform a colleague and other parties as appropriate, including medical defence organisation. If unwanted contact or communication persists, discontinue the patient's care and refer to another practitioner. If the patient poses an immediate risk to the practitioner or third parties, consider civil commitment (if mentally ill) or prosecution. Preventive strategiesIt is crucial that medical practitioners take steps to protect their privacy. They should be cautious about disclosing personal details to patients or leaving personal information where patients can access it, and avoid divulging home contact details to professional organisations or registries that appear in the public domain. With the introduction of electronic patient data, clinicians need to be familiar with the associated confidentiality and privacy issues.18 Office security is a priority in any medical rooms.9 Clinicians who practise in relative isolation without the support staff available in group practices or hospital outpatient departments are precariously placed to manage patients with a history of stalking or other threatening behaviours. It is wise to arrange a chaperone for physical examinations and, if the patient objects, to document this in the medical file. A doctor who is concerned about possible stalking behaviours by a patient should at the very least endeavour to see the patient when other staff are nearby.19 ConclusionsIn studying stalking, we are constantly seeking more effective ways of identifying and protecting the victims. At a time when healthcare resources are a widespread concern, phenomena with the potential to end medical careers cannot be ignored. Studies to date indicate that stalking of medical practitioners occurs with sufficient frequency to warrant systematic attention. Research is currently in progress to empirically measure the extent of the problem in the medical profession. Future research efforts should be directed to developing specific management strategies and formulating workplace policies. Greater emphasis must be placed on incorporating this knowledge into medical training curricula, so that doctors are less confused about their responsibilities, less inclined to ascribe these experiences to their own shortcomings, and more likely to obtain definitive help before their lives (and often those of their patients) are devastated. Because not just doctors but a range of healthcare professionals are at risk, these approaches will almost certainly have wide application.

Michele T Pathé MB BS(Hons), FRANZCP · Paul E Mullen MB BS, DSc · Rosemary Purcell BA, MPsych

Objective: To describe the genetic basis of various personality traits.Design: Prospective, blinded cohort study comparing questionnaire-reported personality traits with candidate genes for temperament, as revealed by genetic mapping in the Human Genome Project. Non-supervised questionnaires were mailed to MJA subscribers. DNA extracted from newborn screening blood samples of all New South Wales participants was used to perform mutation analysis for candidate personality genes.Setting: Tertiary medical care in New South Wales, 1 April 2000 to 1 April 2001.Participants: Healthcare professionals who admitted to reading the MJA on at least a semi-regular (monthly) basis.Main outcome measures: Correlations between occupation, personality and gene mutations were sought using a LOD score in comparison with a classic Poisson d'avril distribution.Results: Mutations were identified that suggested the existence of genes determining several personality traits. Genes coding for belligerence (bel), charisma (lub), cynicism (dub), housekeeping (vac and uum), lack of personality (dul-1), obsessive-compulsive behaviour (pic-e) and gullibility (suk-r) are described. These were found to be selectively represented in certain members of the healthcare profession.Conclusions: The seven most important healthcare personality genes have now been described for posterity.

Dominic A Fitzgerald PhD, FRACP · David Isaacs MD, FRACP

A growing range of infections can be safely and effectively treated with parenteral antimicrobial therapy at home, including cellulitis, pyelonephritis, pneumonia, endocarditis, osteomyelitis, septic arthritis and deep abscesses. Patients may be admitted to HITH directly from the emergency department or after a period of in-hospital care; they must be thoroughly assessed for suitability, including clinical stability and social circumstances, and both patient and carer consent must be obtained. Patients should be medically reviewed weekly at the hospital to monitor progress of therapy and check for possible complications, including adverse drug reactions. Antibiotic selection should be based on appropriate prescribing principles rather than purely dosing convenience. Innovative dosing regimens, including once-daily aminoglycosides, continuous-infusion β-lactams (eg, flucloxacillin), once- or twice-daily cephalosporins (eg, cephazolin) and oral fluoroquinolones (eg, ciprofloxacin) provide effective therapy for a wide range of infections that would have previously required in-hospital care. Appropriate use of HITH leads to improved patient and carer satisfaction, efficient in-hospital bed use and possibly some financial efficiencies.

Series Editors:

Community-acquired pneumonia is caused by a range of organisms, most commonly Streptococcus pneumoniae, Mycoplasma pneumoniae, Chlamydia pneumoniae and respiratory viruses. Chest x-ray is required for diagnosis. A risk score based on patient age, coexisting illness, physical signs and results of investigations can aid management decisions. Patients at low risk can usually be managed with oral antibiotics at home, while those at higher risk should be further assessed, and may need admission to hospital and intravenous therapy. For S. pneumoniae infection, amoxycillin is the recommended oral drug, while benzylpenicillin is recommended for intravenous use; all patients should also receive a tetracycline (eg, doxycycline) or macrolide (eg, roxithromycin) as part of initial therapy. Flucloxacillin or dicloxacillin should be added if staphylococcal pneumonia is suspected, and gentamicin or other specific therapy if gram-negative pneumonia is suspected; a third-generation cephalosporin plus intravenous erythromycin is recommended as initial therapy for severe cases. Infections that require special therapy should be considered (eg, tuberculosis, melioidosis, Legionella, Acinetobacter baumanii and Pneumocystis carinii infection).

Paul D R Johnson PhD, FRACP · Lou B Irving FRACP, FRACGP · John D Turnidge FRACP, FRCPA

To the Editor: Warfarin is frequently used in the same population that is at risk of age-related macular degeneration (ARMD), the commonest cause of blindness in the elderly. A recent report has suggested that warfarin may cause severe intraocular haemorrhage and loss of vision in the minority of patients who have the neovascular form of ARMD.1 We have also seen this occur in patients taking warfarin. There are methodological imperfections in the report, and the association is certainly not proven. Nevertheless, it would seem prudent to exercise some caution in the use of warfarin in patients who are or may be at risk. Our preliminary recommendations (pending a prospective study) are: If a patient has only one functioning eye (for whatever reason), an ophthalmologist's opinion should be sought as to the risk of neovascular ARMD in the seeing eye before or soon after commencing warfarin therapy. Low-risk patients are easily identified and warfarin use in these patients should have no visual sequelae. In patients considered at high risk of neovascular ARMD, the use of warfarin may carry a (currently unquantifiable) risk of visual loss. Alternatives to warfarin should be considered, and the possible risks of taking, or not taking, warfarin should be discussed with the patient. If a patient has two seeing eyes, the risk of bilateral visual loss from warfarin must be exceedingly small, and indeed there have not even been anecdotal reports of such an event. Warfarin can be used in such patients without regard to ocular status. Ophthalmologists should ask all patients they examine whether they take warfarin and should communicate to the treating doctor the presence or absence of factors that put this patient at high risk of neovascular ARMD.

Lionel M Kowal FRANZCO · C Alex Harper FRANZCO

To the Editor: I wish to report a case of quinine-induced disseminated intravascular coagulation (DIC) and haemolytic–uraemic syndrome (HUS). A 78-year-old woman presented with nausea, vomiting, diarrhoea, fever and confusion three hours after taking 150 mg of quinine for leg cramps. Five months earlier she had been admitted overnight for similar symptoms after quinine ingestion, her symptoms resolving over 12 hours without sequelae. Before this she had ingested quinine infrequently for the preceding five years without complication. Her past history included hypertension, hypercholesterolaemia and glaucoma; medications were simvastatin, lisinopril and latanoprost eye drops. On presentation her temperature was 40ºC, blood pressure was 150/90 mmHg; physical examination was otherwise unremarkable. Initial investigations showed serum creatinine concentration, 0.11 mmol/L (reference range, 0.05–0.10 mmol/L); platelet count, 124 × 109/L (reference range, 150–400 × 109/L); prothrombin time, 21.2 seconds (reference range, 11s–16s); activated partial thromboplastin time, 66.3 seconds (reference range, 25s–42s); fibrinogen concentration, 2.1 g/L (reference range, 1.5–4.0 g/L); and D-dimer level, > 4.0 mg/L (reference range, < 0.35 mg/L). No haemolysis was present on the initial blood film. The patient rapidly developed oliguric renal failure, progressive coagulopathy and thrombocytopenia. There was no focus of infection, and blood, urine and faecal cultures were negative. Urine microscopy showed 3 × 106 leukocytes per litre (reference range, < 10 × 106/L), 270 × 106 erythrocytes per litre (reference range, < 10 × 106/L) and granular casts. Her urine output improved following infusions of saline, dopamine and high dose frusemide, but renal function continued to deteriorate. The coagulopathy had resolved by 48 hours after taking the quinine, but thrombocytopenia and renal function continued to worsen, with a platelet count of 18 × 109/L and a serum creatinine concentration of 0.58 mmol/L, evidence of haemolysis with fragmentation of red blood cells, elevated concentrations of lactate dehydrogenase (2450 U/L; reference range, 110–250 U/L) and bilirubin (32 µmol/L; reference range, < 20 µmol/L), and low haptoglobin concentration (< 0.06 g/L; reference range, 0.3–2.15 g/L), consistent with haemolytic–uraemic syndrome. The patient was treated with four cycles of plasma exchange with 3 L volumes, corticosteroids and two cycles of haemodialysis over 11 days. Renal function gradually improved, although the serum creatinine concentration remained elevated at 0.17 mmol/L two months later. The most common adverse reaction to quinine is thrombocytopenia. Six cases of DIC and 10 cases of HUS following quinine ingestion have been previously reported.1-5 This is the first report of both DIC and HUS occurring together after exposure to quinine. Many of the case reports describe multiple presentations before quinine was identified as the precipitant. It is important that prescribers are aware of this rare but serious reaction which may occur following exposure to quinine, and that a history of ingestion is sought in anyone presenting with otherwise unexplained DIC or HUS.

Andrew P Morton MB BS, FRACP

To the Editor: In 1998, the Australasian Diabetes in Pregnancy Society (ADIPS) published management guidelines for gestational diabetes mellitus (GDM).1 Recently, the American College of Obstetricians and Gynecologists (ACOG) published its clinical management guidelines for GDM.2 The Table shows there are few differences from the ADIPS guidelines. At this stage, ADIPS does not consider existing evidence warrants revision of its guidelines. ADIPS will retain its existing criteria for the diagnosis of GDM based on a 75 g oral glucose tolerance test (OGTT) pending publication of the Hyperglycaemia and Adverse Pregnancy Outcome Study.3 The results of this international prospective study of 25 000 pregnant women should be available in June 2004. A second publication, the draft National evidence-based guidelines for the management of Type 2 diabetes mellitus,4 does not include GDM, but initially recommended that "women with previous GDM should be retested every three years for undiagnosed Type 2 diabetes". This periodicity was selected to retest for undiagnosed disease when the cumulative risk of developing diabetes had reached 5%. The time interval was selected on the basis of European studies. In contrast, the ADIPS guidelines recommended testing every 1–2 years, but gave no reason for this, apart from the high risk of progression to diabetes among women of certain ethnic backgrounds who had had past GDM (as high as 47% over five years in Latino women5). A further, unstated reason for the 1–2-yearly testing was the major concern that fetal exposure to undiagnosed diabetes in any subsequent pregnancies could result in malformations. The following has now been inserted into the draft Type 2 guidelines: 4 "The guideline conclusion to retest women with previous GDM every 3 years represents minimum criteria. More frequent retesting may be appropriate depending on clinical circumstances, especially during the child bearing years." ADIPS supports this amendment fully and has revised its own guidelines in relation to maternal follow-up after GDM as follows: All women with previous GDM to be offered testing for diabetes with a 75 g OGTT 6–8 weeks after delivery; Repeat testing should be performed every 1–2 years among women with normal glucose tolerance and the potential for further pregnancies; If pregnancy is not possible, follow-up testing should be performed every 3 years, with more frequent retesting depending on clinical circumstances (eg, ethnicity, past history of insulin treatment in pregnancy, recurrent episodes of GDM). Differences between management guidelines for gestational diabetes mellitus (GDM) from the Australasian Diabetes in Pregnancy Society (ADIPS, 1998) and the American College of Obstetricians and Gynecologists (ACOG, 2001) [Table corrected on 27 August 2002.] ADIPS ACOG Universal versus selective screening by blood test Universal unless low GDM incidence or resources limited No recommendation. States that "many physicians elect to screen all pregnant patients as a practical matter" Differences in definition of low risk for GDM Age < 30 years, obesity, family history of diabetes Age < 25 years, body mass index < 25 kg/m2. No known diabetes in first-degree relative Oral glucose tolerance test used 75 g, 2-hour, 2-point blood sampling 100 g, 3-hour, 4-point blood sampling Criteria for diagnosis of GDM Plasma glucose level: Fasting, ≥ 5.5 mmol/L and/or 2-hour, ≥ 8.0 mmol/L Plasma glucose level: Fasting, ≥ 5.3 mmol/L; 1-hour, ≥ 10.0 mmol/L 2-hour, ≥ 8.6 mmol/L; 3-hour, ≥ 7.8 mmol/L; (2 or more time points need to elevated) Insulin therapy commenced after medical–nutrition therapy Plasma glucose level: Fasting, ≥ 5.5 mmol/L and/or 1-hour postprandial, ≥ 8.0 mmol/L and/or 2-hour postprandial, ≥7.0 mmol/L Plasma glucose level: Fasting, ≥ 5.3 mmol/L and/or 1-hour postprandial, ≥ 7.2–7.8 mmol/L and/or 2-hour postprandial, ≥ 6.7 mmol/L

on behalf of the Australasian Diabetes in Pregnancy Society

To the Editor: While we enjoyed reading Sewell's article summarising advances in paediatrics, there was no mention of developments in general paediatric surgery.1 About a third of all patients admitted to a paediatric hospital have surgical conditions. Viable advances would thus have significant implications for many children and their families. Prevention: Although vaccination against infectious diseases remains vital, the greatest killer of children in Australia is trauma.2 Detailed analysis of patterns of injury morbidity and mortality enable recommendations for prevention to be made. Legislation to reduce the risk of scalding was enacted in 1999 in NSW to regulate the temperature of hot water in new homes. Similar measures, in addition to educating families and their local doctors, have been proposed to prevent driveway and horse-related trauma in children.3,4 Overseas data confirm that such interventions may be effective in helping to reduce Australia's present mortality rate from injury of 9.5 per 100 000 children in 1991–1995 to Sweden's rate of 5.2 per 100 000.5 Diagnosis: Laser Doppler imaging of paediatric burns will enable the surgeon to determine the requirement for operative intervention within 48 hours of the burn, expediting treatment and reducing costs.6 In conjunction with the use of cultured keratinocytes, the risk of subsequent scarring should be minimised.7 Antenatal diagnosis of hydronephrosis and hydroureter has assisted in our understanding of the natural history of urological disease in childhood, helping refine the indications for surgical intervention.8 Intervention: The safety of early surgical intervention in childhood is now well established. Many common conditions such as hypospadias are now optimally treated before the child's first birthday, requiring earlier referral.9 Minimally invasive surgery has now evolved into a useful additional technique in children, in conjunction with the development of appropriate indications, suitable instruments and specialist surgical skills.10 While brevity may be an editorial necessity, paediatric surgery encompasses many areas. Although our selection represents a personal choice, advances require active involvement and consultation with colleagues across all specialties.

Andrew J A Holland · Daniel T Cass · John Pitkin

In reply: The points made by three senior surgeons at The Children's Hospital at Westmead Hospital, emphasising important areas of progress in general paediatric surgery, are valid and point to the wide range of advancing activities in paediatric care. Given the difficulty of covering all areas in a brief article, I am pleased that the Letters to the Editor section of the Journal provides another opportunity to broaden the discussion.

Jillian R Sewell

To the Editor: Both the American Diabetes Association (ADA)1 and the World Health Organization (WHO)2 have lowered the fasting plasma glucose (FPG) level for the diagnosis of diabetes from 7.8 mmol/L to 7.0 mmol/L. The Australian Diabetes Society (ADS) has also adopted the lower level.3 However, these organisations differ in the procedure for diagnosis they recommend. The recent article by Hilton and colleagues4 compared these procedures, with particular attention to including an oral glucose tolerance test (OGTT). We, on the other hand, have investigated the impact of lowering the diagnostic FPG level to 7.0 mmol/L. Data were obtained by the Geelong Osteoporosis Study from an age-stratified sample of women randomly selected from electoral rolls for the Barwon Statistical Division5 and adjusted to match the national age profile. Venous FPG level was determined after an overnight fast, together with blood pressure (BP, seated) and anthropometric measurements, in 944 women aged 20–91 years (mean age, 47.5 years; SD, 17.8 years). History of diabetes was ascertained by questionnaire. The prevalence of self-reported diabetes and diabetes defined by an FPG level of 7.0 mmol/L or higher was 4.3% (95% CI, 3.0%–5.6%; 41 women), whereas using an FPG level of 7.8 mmol/L or higher gave a prevalence of 3.8% (95% CI, 2.6%–5.0%; 36 women). With the lower cut-off level, 29% of women (12) were unaware of their diabetes, compared with 19% (7) using the higher FPG level cut-off point. Characteristics of those identified using the lower cut-off FPG level are shown in the Table. After age-matching all patients with diabetes with control participants, diabetes was significantly associated with obesity (body mass index, > 30; odds ratio [OR], 4.2; 95% CI, 1.5–11.6) and central body fat distribution (waist/hip ratio, > 0.8; OR, 8.0; 95% CI, 2.3–25.9); and non-significantly associated with higher blood pressure (systolic, > 140 mmHg; diastolic, > 85 mmHg; OR, 2.0; 95% CI, 0.8–4.9). The new criterion for diagnosing diabetes identifies a subgroup of the population with a high proportion of obesity and android habitus, with a tendency to higher blood pressure. The recommendation of lowering the diagnostic FPG level increases the prevalence of diabetes by an apparently small proportion, but would diagnose diabetes in an additional 34 000 women in Australia. Characteristics (mean ± SD) of diabetic women (FPG ≥ 7.0 mmol/L) and controls (FPG < 7.0 mmol/L). Characteristic Diabetics (n = 41) Controls (n = 903) P* Age (years) 65.1 ± 11.1 46.7 ± 17.7 < 0.0001 Weight (kg) 74.5 ± 16.2 68.6 ± 14.4 0.03 Height (cm) 158.6 ± 5.6 161.9 ± 6.5 0.0007 BMI (kg/m2) 29.6 ± 6.1 26.2 ± 5.3 0.001 Waist/hip ratio 0.88 ± 0.06 0.80 ± 0.07 < 0.0001 Systolic BP (mmHg) 139 ± 21 121 ± 21 < 0.0001 Diastolic BP (mmHg) 83 ± 16 76 ± 12 0.007 * t test. FPG = fasting plasma glucose; BMI = body mass index; BP = blood pressure.

Julie A Pasco PhD · Mark A Kotowicz MB BS, FRACP · Margaret J Henry PhD · Geoffrey C Nicholson PhD, FRACP, FRCP

To the Editor: The recent letter by McElduff and Hitchman1 has some very practical implications. They were able to show that pregnant women having a glucose challenge test (GCT) in the afternoon were nearly twice as likely to have a positive result as women tested in the morning, so that more women tested in the afternoon were diagnosed with gestational diabetes mellitus (GDM). If the function of the GCT is to aid in the diagnosis of GDM, then either all women should be tested in the afternoon or the glucose "cut-point" for the morning test should be reduced. But does the GCT now have any relevance? In the United States, where testing for GDM often still involves a three-hour glucose tolerance test (GTT) using a 100 g glucose load and four blood samples, the GCT was introduced to reduce the number of women who had to have this long and, because of the higher dose of glucose, relatively unpleasant procedure. In Australia, where a two-hour, 75 g GTT is used (requiring two blood samples), it is not as important to offer a simpler initial test. With the use of an initial GCT, about a quarter of women will need to have a GTT for confirmation, and the definitive diagnosis of GDM will be delayed. Further, the GCT is not specific and some women who may have GDM will not have a GTT. In addition, there will inevitably be some women who are GCT-positive, some of whom will have GDM, who do not return for the definitive GTT. Thus, while a GCT may be convenient for a busy hospital clinic with space limitations, it may not necessarily be in the best interests of the patient. Whether a GCT is ultimately helpful or possibly a hindrance requires further evaluation.

Robert G Moses MD

To the Editor: We conducted a joinpoint analysis of death certificate data on prostate cancer from the Australian Bureau of Statistics. Between 1979 and 1994, mortality rates increased by 2.1% per year (95% CI, 1.6% to 2.5%). However, between 1994 and 1999, mortality decreased by 4.2% per year (95% CI, – 5.8% to – 2.4%). The total decrease in mortality rates for the five years to 1999 (the most recent year for which data were available) was 22.6% (95% CI, – 32.9% to – 12.7%). Joinpoint analysis is a statistical method that measures changing trends over time. It chooses the best-fitting points (called joinpoints) at which the rate of increase or decrease changes significantly.1 We did not look at the data and then choose 1994 as the start of the decreasing trend in prostate cancer mortality. The significant decrease since 1994 (and the consistent increase between 1979 and 1994) were identified by the joinpoint analysis. Whether early diagnosis and treatment of prostate cancer subsequent to screening with prostate-specific antigen (PSA) tests can save lives is still an open question that is best answered by randomised-controlled, long term trials. Nevertheless, it is important that we try to understand the recent decrease in population-based mortality. The mortality decline started in 1995, about five years after PSA testing became widely available in Australia. The use of PSA testing increased dramatically, reaching a peak in most States in 1994 and 1995.2 In 1996, the Australian Health Technology Advisory Committee reviewed the evidence and recommended against screening.3 Since then, the number of PSA tests has decreased.2 Recent mortality declines have also been observed in the United States and the United Kingdom following increases in the use of PSA testing. The mortality decline in the US has been greater than that in the UK, coinciding with more PSA testing in the US than the UK.4 At least three questions arise from these observations: Is it plausible that the decrease could be due to some factor other than PSA testing, such as better treatment? How do we explain these results to men who want to make an informed choice about whether to be tested? Is this type of evidence strong enough to warrant a change in the current recommendations on PSA testing?

Michael Coory MB BS, PhD · Peter Baade BSc, PhD

Comment: Using joinpoint analysis, Coory and Bade have identified a significant trend towards lower mortality rates from prostate cancer since 1994, and relate this chronologically to the increased use of prostate-specific antigen (PSA) testing in Australia. The authors raise three reasonable questions regarding this finding: Is it plausible that the decrease could be due to some factor other than PSA testing, such as better treatment? Ecological data such as those presented by Coory and Baade are subject to the pitfalls of ecological fallacy and confounding.1 Ecological fallacy exists when there is an apparent association (eg, more men in the population screened, fewer men in the population dying), but there is no association at an individual level. Confounding would occur when there are other factors that account for the outcome, or distort the relationship between PSA testing and mortality. Treatment is one such factor, but changes in diet and other factors could also play a role (eg, lycopene from tomato-based foods2). How do we explain these results to men who want to make an informed choice about whether to be tested? If these data were presented to men, then it would have to be explained that the data do not provide evidence of the benefits of screening and that many other factors could explain the relationship. Further, men need to be informed of the risks of false negative and false positive results associated with screening, and the risks of complications associated with treatment. It is important to assist men to make a balanced decision. Interestingly, studies in which men have been involved in such informed decisions show that information can reduce the probability that men will choose to be screened.3,4 Is this type of evidence strong enough to warrant a change in the current recommendations on PSA testing? Ecological associations generate hypotheses that are worthy of further investigation. Randomised controlled trial evidence at best, or case–control studies at least, would be required to provide evidence of the benefit of screening using PSA. Some such studies are currently in progress, including the prostate, lung, colorectal and ovarian cancer screening trial of the National Cancer Institute,5 and the European Screening Study for prostate cancer.6 Those in favour of PSA testing argue that it is the only means of diagnosing prostate cancer at an early and potentially curable stage.7 Against this is the argument that many prostate cancers are not clinically significant, as they are slow growing and will not spread, and there is a danger of diagnosis and treatment adding substantially to men's psychological and physical morbidity without benefits in terms of survival or quality of life.8 In the absence of evidence from randomised controlled trials or case–control studies, screening for prostate cancer using PSA testing remains controversial.9

Julie B Byles

To the Editor: A previously well 21-month-old boy presented with burns on his right thigh. His mother had been making popcorn in a domestic popcorn machine while holding the baby about a metre from the machine to watch it churning out the popcorn through a chute into a bowl. Soon after the corn started to pop, the baby screamed, pointing to his right thigh. The mother immediately removed his nappy and found an unpopped corn kernel lodged between the nappy and the thigh. General examination revealed two 4 mm burns on the child's right thigh (Figure). Kenacomb cream (Bristol-Myers Squibb) was applied four times daily, and the lesions healed after five days, leaving two depigmented scars. Burns to the right thigh caused by a popcorn maker. Although popcorn machines are gaining increasing popularity there are few reports of injury. One report describes corneal burns in three adults caused by steam from microwave popcorn.1 We believe that this is the first reported case of paediatric injury, highlighting the potential hazards of hot-air popcorn machines. Corn's ability to pop lies in the fact that the kernels contain a small amount of water (14% of weight) stored in a circle of soft starch inside the hard outer casing.2 When heated to about 232ºC,3 the water expands, creating pressure within, until eventually the casing gives way. The kernels explode and pop, allowing the water to escape as steam, turning the kernels inside out. It is likely that the temperature of kernels that do not pop (known to popcorn connoisseurs as "old maids")2 can reach as high as 200ºC. Because of their increased density compared with popped kernels, old maids can, as occurred in our case, be thrown further than popped kernels. It is also worrisome that some popcorn machines are designed so that they can be used by children.4 The information booklet that accompanies the popcorn machine in this case does state "close supervision is necessary when this appliance is being used by or near children", and has a warning that as "some hot unpopped corn may be thrown from the machine during the popping process be sure to place the [machine] facing away from you or do not stand directly in front of the machine whilst it is in operation". The mother of the child in our case did not note the warning in the instruction booklet. She was not standing directly in front of the machine. Further, she was of the impression that the fluffy popped corns are not hot. With the increased use of popcorn machines, consumers must be aware of the potential dangers of injuries, especially to young children. Our case emphasises the importance of keeping children at a safe distance while preparing hot food, including apparently harmless, fluffy popcorn.

T H H Guan Koh MA, MB BChir, FRCPCH, FRACP

False negatives Although not yet ready for clinical application, a new screening test, using stool specimens, could detect early colorectal cancer. The earliest detectable molecular abnormality in both familial and sporadic colorectal cancer is a mutation of the adenomatous polyposis coli (APC) gene. Recently, researchers from the United States and Sweden overcame major technical hurdles to detect APC mutations in the faecal DNA of patients with colorectal tumours. APC mutations were detected in stools of 26 of the 46 patients with neoplasia, and none of the 28 control patients. Six (of 12) patients whose neoplasia was not detected with the stool screen were found to have APC mutations in their tumours. Thus, 78% of patients (95% CI, 65%–89%) had mutations potentially detectable by this new assay. So far the assay appears to be highly specific but is lacking sensitivity. N Engl J Med 2002; 346: 311-319 N Engl J Med 2002; 346: 302-304 False assumption A recent report from a large epidemiological study challenges the belief that 8 hours’ sleep per night is ideal. The association between sleep duration and mortality risk was examined in the Cancer Prevention Study II. More than a million Americans completed a health questionnaire in 1982, including questions about sleep duration. In 1988 it was possible to determine survival, or date of death, for 98% of participants: 636 095 women aged 57 (SD, 11) years and 480 841 men aged 58 (SD, 10) years. Adjustments were made for 32 variables found to be predictive of mortality risk. Men and women who reported sleeping 7 hours per night were least likely to have died. Those who reported sleeping > 8 hours experienced significantly increased mortality hazard, as did those who slept < 6 hours. The risk was increased by more than 15% for those sleeping > 8.5 hours or < 4.5 hours. Arch Gen Psychiatry 2002; 59: 131-136 False promises? A Cochrane review has concluded that there is little evidence for or against the effectiveness of over-the-counter cough preparations. To be included, studies involved persons aged 16+ years with cough of less than three weeks’ duration and of presumed viral origin, and no signs on auscultation or x-ray of the chest. Outcomes included frequency and severity of cough. Only 15 randomised, placebo-controlled trials, involving 2166 participants, could be included in the review. Study quality was not high, with details of randomisation or blinding often absent, or studies too small to detect clinically important differences. Five trials of antitussives, two of expectorants, two of antihistamine–decongestant combinations and three of other combinations had conflicting results. A positive trial of a mucolytic was of poor quality and, based on two trials, antihistamine was no more effective than placebo. BMJ 2002; 324: 1-6 True lies People with high levels of post-traumatic stress are likely to be inconsistent in the accounts they give of a traumatic memory. More discrepancies occur with details that they consider peripheral, and with the passage of time. Thirty-five Bosnian and Kosovar refugees, who already had leave to remain in the United Kingdom, agreed to participate in research about memory. Using an interpreter, each was asked to recall both a traumatic and a non-traumatic event, then answer 15 standard questions about each event. Three to 32 weeks later, each person was asked to recall the events again and to answer the same questions. Overall, the mean discrepancy rate was 0.32 (95% CI, 0.28–0.36). Scores from a rating scale for post-traumatic stress were used to divide the group for comparisons. This research was prompted by concern that inconsistencies in recall currently undermine a refugee’s credibility as a witness. This may be central to determining his or her refugee status. BMJ 2002; 324: 324-327 True images A new non-invasive technique for imaging deep venous thrombosis (DVT) looks promising, even for detecting clots in the pelvis and calf. While other methods show the clot as a filling defect, Magnetic Resonance Direct Thrombus Imaging (MRDTI) detects the thrombus itself as a high signal against a suppressed background. A recent prospective study of MRDTI in the United Kingdom recruited 101 patients with suspected DVT immediately after routine venogram. An experienced radiologist, blinded to the venogram result (the “gold standard” for comparison), diagnosed DVT using MRDTI with high sensitivity (96% [95% CI, 89%–99%]; 51 of 53 tests) and specificity (90% [95% CI, 79%–96%]; 43 of 48 tests). MRDTI should have particular advantages in pregnancy, for patients in plaster and in diagnosing recurrent DVT. Ann Intern Med 2002; 136: 89-98