Letters

To the Editor: While clinicians often communicate with patients and families, documentation of these conversations is inconsistent. Documentation is critical for continuity of patient care, medicolegal reasons and research,1 and is particularly important in the intensive care unit (ICU), where discussions regarding prognosis and withdrawal of care occur frequently. There are scant published data on documentation of conversations with patients in ICUs and their families. We conducted a retrospective audit of patients admitted to the ICU of Wesley Hospital (a 500-bed private teaching hospital in Brisbane) between 1 January and 31 August 2009 to: determine levels of documentation of communication with patients and their families by the ICU medical staff; and compare this with documentation of communication by the primary physician before and after admission to the ICU. After obtaining Wesley Hospital ethics committee approval, all patients who were cared for in the general ICU for more than 5 days were studied. During the study period, there were 862 ICU admissions, 100 of which met our inclusion criteria. The charts of only 82 patients could be successfully retrieved and these were used for final analysis. These patients had a mean age of 64 years (SD, 14 years), and a mean Acute Physiology and Chronic Health Evaluation II (APACHE II) score of 17 (SD, 7). Sixty patients (73.1%) survived to hospital discharge. In the ICU, there were 39 family conferences at any stage between an ICU clinician and the families documented for 24 patients (29% of patients; 1.6 conferences per patient). Eleven of these conferences took place in the first 48 hours, eight between 2 and 4 days, and 20 beyond 96 hours of ICU admission. The Box shows that there was a greater proportion of documented family conferences for patients who died compared with those who survived at both 2–4 days (27% v 3%; P < 0.01) and beyond 96 hours (86% v 1.5%; P < 0.001). Of note, there was no documented communication between the hospital’s admitting physician and patients or families before ICU admission or after discharge from the ICU. Intensivists use any prior discussions to make decisions about continuing or withdrawing care.2 The absence of documentation before ICU admission is clinically relevant in this context. Potential reasons for inadequate documentation include (i) discussions occurring informally at the bedside or outside of the ICU (corridor conversations) and (ii) nurses providing updates in the clinician’s absence. Better documentation in the charts of dying patients may reflect their longer ICU stays, which provide more opportunity for communication. Moreover, discussions on treatment withdrawal are more likely to be documented as they are often a shared decision-making process. Globally, levels of documentation vary (10%–90%).3,4 Potential strategies to improve documentation include: bedside reminders (such as “have you documented family conferences?”); availability of a communications officer; an internal appointments system for formal discussions; and the use of communication kits.5 Despite being a retrospective study, our findings suggest a need to improve levels of documentation, and may prompt clinicians to examine their documentation practices and develop protocols to improve record keeping. Comparison of the documented communication rates for intensive care unit (ICU) patients who died and those who survived

Riad L Silcock · Bala Venkatesh · Ranald L Pascoe · Dianne K Fisher

To the Editor: Before and since 1 July 2010, when the National Registration and Accreditation Scheme for health practitioners commenced, the claim that mandatory reporting laws will deter impaired doctors from seeking help has frequently been made. It was on the basis of this claim that Western Australia legislated to exempt health professionals from reporting impaired practitioners they are treating. At a recent conference of the Royal Australian College of General Practitioners, a representative of a medical indemnity organisation labelled the mandatory reporting laws a disgrace because the health of impaired doctors who are deterred from seeking help for this reason would be put at risk. However, she also indicated that the problem was more one of perception than reality because doctors feared triggering a mandatory report automatically if they sought help from another doctor for a perceived impairment.1 It is a problem of perception because, under the laws, only doctors whose impairment places the public at risk of “substantial” harm are required to be reported.2 If doctors have an unreasonable fear of mandatory reporting, we can infer that many are unaware of the details of the laws, in particular the reporting thresholds. Yet an argument against introducing the laws was that doctors were already under an ethical obligation to report, to the relevant authority, unprofessional conduct, impairment or performance that would put patients at risk.3 For this argument to be valid, doctors would need to be aware of the details of mandatory reporting because the Medical Board of Australia’s code of conduct for doctors states, in its list of ethical obligations, that doctors should be aware of their reporting obligations.4 So this particular argument does not appear to be valid. Because medical professionalism puts the wellbeing of the patient first; because psychological or physical health status may affect professional performance; and because the Board, like the state boards that preceded it, has a primary duty to the safety of the public — any doctor whose impairment poses a substantial risk or, in the absence of that doctor’s insight, any treating doctor who considers that a substantial risk exists should surely feel ethically compelled to report the matter to the Board. No doctor whose impairment does not pose a substantial risk should feel deterred from seeking medical help. The consequences for impaired doctors who are reported under the new legislation are no different from those of reporting a doctor when it was “merely” an ethical requirement — being placed on an impaired practitioners’ register and supported, managed and monitored, while, in most cases, continuing to practise. If impaired doctors and their treating doctors feel deterred by mandatory reporting laws, we are entitled to conclude that there was, and continues to be, significant non-compliance with the ethical obligations that arguments against mandatory reporting depend on.

Malcolm H Parker

To the Editor: A 28-year-old man presented to his general practitioner with a history of epigastric pain and constipation over 1 month. In addition, he had a history of chronic low back pain. Findings on physical examination were unremarkable. Laboratory investigations showed normo-chromic, normocytic anaemia with basophilic stippling (Box 1). His whole-blood lead level was subsequently estimated to be 4.12 μmol/L (level recommended by the National Health and Medical Research Centre for all Australians, < 0.48 μmol/L). The patient was referred for toxicological review. Further questioning revealed he had used three Ayurvedic medicines (Vatyog [Arya Aushadhi Pharmaceutical Works, Indore, India], Sahacharadi [Arya Vaidya Nilayam, Madurai, India] and Gandharvahastadi [Arya Vaidya Nilayam, Madurai, India]) for back pain, dispensed to him 3 months earlier during a trip to India. He ceased taking the medications, and a 19-day course of oral chelation with succimer was administered. His blood lead concentration fell rapidly, with a moderate rebound 6 weeks after the completion of chelation therapy. A negative blood lead result for the patient’s pregnant partner excluded environmental exposure in the patient’s home. Vatyog and Sahacharadi were analysed for heavy metals (Gandharvahastadi was not available for analysis). Sahacharadi was lead-free, but the Vatyog tablet tested contained 448 μg of lead. The patient had potentially ingested 896 μg of lead daily for 3 months. World Health Organization guidelines recommend daily lead intake should not exceed 3.5 μg/kg/day.1 Dietary intake of lead in developed nations has been reported to be about 0.1–0.7 μg/kg/day.1 Ayurvedic medicine originated in India more than 2000 years ago and relies heavily on herbal products.2 Many people take Ayurvedic medicine without any problems. In some traditional remedies, salts of heavy metals are included as active ingredients.3 A study in the United States found that one-fifth of Ayurvedic herbal products manufactured in South India and sold in Boston contained potentially harmful concentrations of lead, mercury or arsenic, or combinations of these.2 In Australia, traditional Indian and Chinese medicines authorised for supply are regulated as complementary medicines and must meet manufacturing and quality standards that ensure the absence of contaminants.3 However, there is no quality control of medications imported for personal use or purchased over the internet. Importantly, this case involved tablets prepared by a registered pharmaceutical manufacturer and presented in sophisticated packaging (Box 2). Reports of lead contamination from traditional Indian and Chinese medicines have been intermittently publicised4,5 and include recent warnings from the New South Wales Health Department concerning Ayurvedic medicines. Patients with lead toxicity commonly present with non-specific clinical features, and lead poisoning might only be suspected when a blood film shows anaemia with basophilic stippling. While enquiring about a patient’s environmental, occupational and social history is important for determining possible sources of lead exposure, clinicians should also ask patients if they have used alternative and complementary medicines and whether these were obtained in Australia or from overseas. 1 Blood film from the patient showing coarse basophilic stippling 2 Packaging from the lead-containing Ayurvedic medicine

Nilika G Wijeratne · James C G Doery · Andis Graudins

To the Editor: I describe two episodes of transient recurrence of psychosis temporally related to ingestion of an “energy drink” in an obese (125 kg) 27-year-old New Zealand Maori man who had been diagnosed with schizophrenia 8 years earlier. The patient had found that risperidone 6 mg/day effectively relieved persecutory ideas and auditory hallucinations. He had stopped using cannabis and alcohol to excess, but continued to drink up to 10 cups of instant coffee throughout the day for a “lift”, apparently without insomnia or other complications. In July 2009, he consumed a 60 mL Demon Shot energy drink and enjoyed an hour-long “buzz”. Repeating the dose did not re-create the desired effect, and instead made him uneasy, irritable and paranoid; for the first time in many months, he had recurrent thoughts, lasting several hours, of people wanting to harm him. One week later, he drank three shots over 15 minutes. He again experienced a buzz and was observed to be emotionally labile — initially laughing and talkative, and later, restless, withdrawn and argumentative. He had a rapid pulse and insomnia. These symptoms subsided over the next day, and when he was back to his usual self after a further 2 days, he described having had paranoid ideas (“gangsters after me ... scared to leave the house”) over several hours after consuming the drinks. His family also associated these symptoms with the drinks, noting a striking similarity to his illness before commencing antipsychotic treatment. He has since avoided energy drinks and continued risperidone monotherapy, remaining stable for 15 months. Demon Shot is a concentrated energy drink, widely available in Australia and New Zealand. The product’s web page describes each “insanely intense” shot as containing 200 mg caffeine, plus taurine, guarana, and B vitamins (http://www.demonenergy.com.au/products/demon-shots). As a regular smoker, the patient does have increased caffeine metabolism, and his response to consuming large amounts of coffee suggests he is not particularly sensitive to caffeine. Nevertheless, he did exceed the maximum recommended dose (two shots per day) on the second occasion, consuming at least 600 mg of caffeine (4.8 mg/kg) virtually as a single dose. The additional presence of guarana extract (48 mg per shot in Australia; unspecified in New Zealand) is probably relevant, as guarana is a further source of caffeine and may have other stimulant properties. Caffeinated drinks are popular among patients being treated for schizophrenia,1 possibly due to partial reversal of the unpleasant effects of dopamine blockade.2 Although case reports3 and a later controlled study4 suggest that high doses of caffeine may exacerbate psychosis, this has only recently been described for energy drinks.5 This case provides naturalistic challenge–dechallenge–rechallenge evidence that some patients with treated schizophrenia may be vulnerable to exacerbation of their illness by such products.

David B Menkes

To the Editor: At budget estimates hearings of the Senate Community Affairs Committee in June 2010, answers were provided to questions on notice asked by Senator Guy Barnett to the Therapeutic Goods Administration (TGA) about the use of mifepristone for medical abortion in Australia since 2006.1 The answers are of interest as they include the number of practitioners who have become authorised prescribers of mifepristone in Australia since the “Harradine Amendment” was overturned in federal parliament in February 2006, as well as details of their practice. Nationally, 81 medical practitioners now have TGA authorised prescriber approval for mifepristone: 33 in New South Wales and 18 in Victoria (several large private clinics have gained approval in these states); 15 in South Australia, six in the Australian Capital Territory, five in Western Australia, and four in Queensland. There are none in Tasmania or the Northern Territory, so women in these two regions have no access to mifepristone abortion. There is very accurate documentation of all adverse effects of mifepristone–misoprostol medical abortion, which must be reported 6-monthly to the TGA. This information is then available to the general public through Senate estimates questioning. To 31 December 2009, 2926 medical abortions using mifepristone and misoprostol were performed in Australia by authorised prescribers. These include early and late procedures. The following adverse events were reported in that period: significant haemorrhage (7; 0.24%); retained products of conception requiring dilatation and curettage (D&C) or dilatation and extraction (84; 2.9%); ongoing pregnancy requiring surgical evacuation (14; 0.48%); and nausea and vomiting (10; 0.34%). These results are all well within the parameters expected from Australian and overseas studies of mifepristone–misoprostol use,2-5especially as a high proportion of the 2926 abortions would have been performed in the second trimester, as reported in a recent WA study.2 Large overseas studies on the use of mifepristone–misoprostol for early medical abortion (to 63 days of pregnancy) show a continuing pregnancy rate of around 1%, the need for D&C in 2%–3% of cases, and heavy vaginal bleeding requiring transfusion among 1 : 500 to 1 : 1000 women.3-5 These complications are more common in second-trimester procedures than in early medical abortions.2-5 Use of mifepristone in Australia is still restricted to authorised prescribers, and it is therefore not accessible to many women who might wish to use it. However, the number of cases performed in Australia is now large enough to conclude that mifepristone is safe and effective for the Australian women able to access it.

Caroline M de Costa

To the Editor: Following the 1918 influenza pandemic, during which an estimated 50 million people died,1 the relationship between influenza and secondary bacterial infections such as Streptococcus pneumoniae (pneumococcus) became recognised as an area of considerable scientific importance. It has been shown that influenza infection results in epithelial damage, up-regulation and exposure of respiratory tract receptors, alterations in the immune response and increased adherence of pneumococcus.2 The most severe form of pneumococcal infection, invasive pneumococcal disease (IPD), has a mortality of 7%–9%.3 It occurs frequently in the Indigenous population of the Northern Territory, with an incidence of 70.5 per 100 000, compared with 7 per 100 000 in Australia overall.3 In population studies, 11%–20% of seasonal increases in IPD have been attributed to influenza.4,5 During the increase in influenza notifications in 2009 in the NT, as a result of pandemic H1N1 (2009) influenza, a concurrent increase in IPD notifications was also recorded. Our study aimed to investigate whether there was a relationship between influenza and subsequent IPD. We calculated the relative risk of IPD in the 4 weeks after laboratory-confirmed influenza compared with the background risk. Cases were defined as patients who were diagnosed with IPD within 4 weeks of laboratory-confirmed influenza being diagnosed. Data concerning cases of influenza and IPD between 2005 and 2009 were extracted from the NT Notifiable Diseases System and merged by date of birth using Stata, version 11 (StataCorp, College Station, Texas, USA), leading to 75 matches. The 75 matched records were checked individually to determine whether they met the case definition, resulting in eight potential cases. Demographic characteristics (name, sex, Indigenous status and address) were then checked to confirm the match. One case was excluded, leaving seven cases eligible for analysis. We identified 2567 cases of influenza and 346 cases of IPD in the study period. The risk of IPD within 4 weeks of influenza was calculated as 7/2567 (2.7 × 10-3). To calculate the background risk of IPD in a 4-week period, we first calculated the annual risk of IPD occurring without previous influenza (339/1 076 470 person-years) and divided this by the number of 4-week periods in a year (13). This gave a background risk of 2.42 × 10-5 and a relative risk of 112.5 (95% CI, 48.9–224.8; χ2 = 758.3; P < 0.001). This analysis is based on population surveillance data, and is therefore limited by the possibility of ascertainment bias due to variations in influenza testing. Nevertheless, our study provides additional evidence that IPD is an important complication of influenza, and reinforces the need for clinicians to promote influenza vaccination in high-risk individuals and to be aware of complications which may occur in the weeks after the initial viral infection. Additionally, achieving high coverage of pneumococcal vaccination in at-risk groups will reduce the impact of influenza.

Laura J Edwards · Peter G Markey · Heather M Cook · James M Trauer · Vicki L Krause

To the Editor: Resistant gram-negative bacteria (especially Escherichia coli) are a major and growing problem worldwide.1 Some strains are resistant to all available antibiotics.2 We report a patient with relapsing multi-resistant E. coli bacteraemia from a prostate infection. Relapses occurred despite treatment with numerous parenteral antibiotics that should have been effective. His infection was cured using azithromycin. A 69-year-old man developed dysuria and fever 2 days after a transrectal prostate biopsy; he had received ciprofloxacin for prophylaxis. He had travelled to Morocco 15 months before this episode and to Vietnam 1 year earlier. Blood and urine cultures grew E. coli that showed sensitivity to ceftriaxone, gentamicin and nitrofurantoin; intermediate sensitivity to ampicillin; and was resistant to ciprofloxacin, cotrimoxazole, tetracycline, doxycycline and cephazolin. He was treated with intravenous ceftriaxone for 2 weeks, followed by oral amoxycillin/clavulanate with a good clinical response. The patient relapsed 1 week after completion of this treatment, with dysuria and fever. E. coli was cultured from his urine, with the same sensitivities as before. He was treated for 4 days with ceftriaxone and 2 weeks with amoxycillin/clavulanate. One week after his second course of therapy, he re-presented with dysuria and fever. Urine and blood cultures grew E. coli with the same sensitivities as the previous cultures. A prostate ultrasound showed no focal abscesses. He received ceftriaxone 2 g and gentamicin 7 mg/kg, yet remained febrile. After 2 days, gentamicin was discontinued and azithromycin 500 mg orally daily was added. He became afebrile within 24 hours and his C-reactive protein level fell. Azithromycin was selected because its minimum inhibitory concentration was 4 μg/mL, and because of its efficacy against multiresistant salmonella.3 The patient was treated for 1 week with ceftriaxone and for 3 weeks with azithromycin. He remained asymptomatic and his C-reactive protein normalised. There were no further relapses. He underwent prostatectomy for confirmed prostate cancer 2 months after discontinuing azithromycin. There was no histological evidence of prostate infection. We believe that while travelling in Morocco or Vietnam, he acquired multiresistant E. coli, which was carried in his bowel and inoculated during transrectal prostate biopsy, a common infective risk of this procedure. We believe the initial treatment failure was due to poor antibiotic penetration of the acidic prostate environment, and azithromycin was responsible for the eventual cure. Case reports and clinical trials have shown successful use of azithromycin in treatment of chronic prostatitis due to Chlamydia trachomatis,4 and that azithromycin is active against E. coli.5 This case report demonstrates successful use of azithromycin for treatment of E. coli prostatitis resistant to ciprofloxacin.

Simon H T Jiang · Peter J Collignon

To the Editor: As a family history of cystic fibrosis (CF) is uncommon among children diagnosed by newborn screening, offering carrier testing in the general population is warranted Cystic fibrosis is the most common severe autosomal recessive genetic condition in children. The carrier frequency in populations of Northern European ancestry is one in 25 and the incidence of CF in Victoria, Australia is 1 in 2874.1 There is no cure for CF, but advances in management have improved life expectancy. A question about a family history of CF is often asked as part of preconception or prenatal care in populations in which CF is more prevalent. However, anecdotally, most parents of a baby with CF do not report a family history and the diagnosis is unexpected. Carrier testing would provide information that may be used by couples to make reproductive decisions, such as prenatal and preimplantation genetic diagnosis of a fetus or embryo, respectively. Australia, the United Kingdom, other European countries, and all states in the United States now offer newborn screening for CF. Apart from records of affected older siblings,1 there are no clinical data for the existence of a family history of CF among children diagnosed with CF through newborn screening. We audited the family pedigrees, collected soon after diagnosis, of all children born in Victoria in 2000–2004 who were diagnosed with CF through newborn screening. From the extended pedigrees of 82 children, we identified five families with a family history of CF. In two pedigrees, the children were first cousins; in another two pedigrees, the children were first cousins once removed; and in one pedigree, the children were second cousins. There were no families in which older siblings had been previously diagnosed with CF, but in two families the diagnosis triggered further examination of older siblings who were subsequently diagnosed. These empirical findings that most babies with CF (77/82; 94%) are born to families with no family history of CF support clinical observations. Although inquiry about a family history of CF is necessary as part of prenatal or preconception care, this is not sufficient. Even when a family history is known, most relatives do not undertake carrier testing. For example, in an audit of cascade carrier testing after a diagnosis of CF through newborn screening, only 11.8% of eligible (non-parent) relatives were tested.2 Not surprisingly, a family history-based approach to offering carrier testing will not provide the vast majority of couples with the opportunity to learn their carrier status and make informed reproductive decisions. Therefore, in addition to newborn screening to diagnose affected babies, CF carrier screening in the general population (of individuals with a risk of one in 25)3 is needed.

Belinda J McClaren · Sylvia A Metcalfe · David J Amor · MaryAnne Aitken · John Massie

To the Editor: Rahman and colleagues suggest that iodine deficiency in Victoria increased between 2001 and 2006, based on the findings of thyroid-stimulating hormone (TSH) levels in neonates at routine newborn screening.1 Indeed, their data as presented suggest a doubling of the percentage of mothers with iodine deficiency to over 9% during that period. This could be correct. Certainly, as they state, there is much evidence to suggest that there is mild iodine deficiency in Australia. However, there are caveats about the data they report which are not mentioned. Data from New South Wales do not show this trend. While they do suggest a degree of mild iodine deficiency, there was no increase in the percentage of neonates with TSH levels > 5 mIU/L of whole blood from 2002 to 2009 (Box), although the average age at sampling falls slightly (from 2.96 to 2.32 days) over this period. The World Health Organization has defined iodine sufficiency as being indicated, inter alia, when more than 3% of newborns aged 3–4 days have a TSH level > 5 mIU/L of whole blood.2 Factors that affect the TSH level in a newborn screening program include the precise age at sampling, and any changes to the method of TSH analysis used. In a Swiss study assessing the efficacy of iodine supplementation, there was a small but significant decrease in the TSH level from Day 3 to Day 4 of age.3 This is unsurprising: following the TSH surge in the first hours after birth, TSH levels decline gradually to a steady level at about Day 7.4 There could well have been a trend to earlier sampling in Victoria, within the bounds of the 2–4 days of age assay that Rahman and colleagues mention, over the period studied, but these crucial data are not given. The dried blood spot TSH assay method is not described either. A change in any aspect of the methodology; for example, if the manufacturer modified the antibody used, may produce a small, clinically insignificant but numerically significant, change in results. If the data presented by Rahman and colleagues for Victoria do not have these biases, then the situation warrants further investigation, but whatever is happening in Victoria seems not to be replicated over the border. Percentage of newborns with thyroid-stimulating hormone (TSH) level > 5 mIU/L of whole blood, detected by routine newborn screening in New South Wales, by year Year Newborns with TSH level > 5 mIU/L 2002 3.80% 2003 3.68% 2004 3.87% 2005 5.02% 2006 4.48% 2007 3.56% 2008 3.92% 2009 4.00%

Bridget M Wilcken · Veronica C Wiley

In reply: The methods used for blood sample collection and analysis remained unchanged during our data collection period. One source of thyroid-stimulating hormone (TSH) calibrators and reagents was used over the study period by a single laboratory covering all of Victoria. Material from the United States Centers for Disease Control and Prevention was used for external quality assurance, ensuring that the results were in agreement with those of other laboratories. The per cent coefficient of variation over the period ranged from 10% to 20%. The table of neonatal TSH values for New South Wales provided by Wilcken and Wiley further demonstrates the value of using TSH levels as a screening tool for population iodine status, even with a decreasing mean age of sample collection. While we dealt with the effect of sample collection time in our published article,1 here we present a table illustrating analysis of the Victorian neonatal TSH values for samples collected at 48, 72 and 96 hours after birth (Box). The percentage of elevated TSH values increased from 2001 to 2006 at each collection time and, although the percentage of elevated TSH values decreased with increasing age, these values were still indicative of iodine deficiency. Iodine status varies between regions. The National Iodine Nutrition Study (NINS) found both South Australia and Queensland iodine sufficient, while the neighbouring states of NSW and Victoria were iodine deficient.2 The results also indicated that iodine status was worse in Victoria than in NSW; therefore, we might expect similar differences in TSH values. We are now in the process of analysing Victorian TSH values for 2007 to 2010. Percentage of newborns with thyroid-stimulating hormone (TSH) level > 5 mIU/L for blood samples collected at 48, 72 and 96 hours after birth, Victoria, 2001–2006 Sample collection time (h) Percentage of neonates with TSH > 5 mIU/L according to birth year 2001 2002 2003 2004 2005 2006 48 5.73% 6.83% 8.47% 10.58% 11.86% 13.53% 72 4.20% 5.15% 6.87% 7.01% 9.13% 9.38% 96 2.49% 3.21% 4.37% 3.78% 5.98% 5.34%

Ashequr Rahman · Gayle S Savige · Nicholas J Deacon · Ivan Francis · Janice E Chesters

To the Editor: O’Leary and Major1 have opened debate on the issues raised by the New South Wales-wide introduction of the Clinical Excellence Commission Between the Flags (BTF) observation charts, which incorporate escalation thresholds for vital signs and other criteria. In the BTF charts, “yellow” zone criteria trigger a clinical review and “red” zone criteria, a rapid response.2 Evidence shows that deterioration can be recognised early, reducing serious consequences.3,4 O’Leary and Major trialled a subset of the draft BTF paediatric calling criteria in the context of triage in an emergency department, applying them retrospectively and comparing the actual triage decision with the one that would have been made using the BTF criteria alone. They also examined patient disposition for patients falling within the yellow and red zones. They concluded that “the physiological parameters ... are not suitable as a triage tool in the paediatric emergency department, do not replace an experienced triage nurse, and are a poor predictor of disposition”.1 We are not surprised by this conclusion. The Australasian Triage Scale is designed to assess a patient’s urgency based on presenting problem and general appearance, possibly combined with physiological observations.5 Trialling a subset of vital signs, on their own, against the triage process, although interesting, is unlikely to demonstrate a strong correlation in decision making for the following reasons. The BTF vital signs observations are designed for use in the context of a “track and trigger system” in a general ward, to monitor trends, not as a substitute for the triage process. Vital sign observations on their own contribute little to triage decisions. Triage decisions are poor predictors of disposition (eg, up to 67% of triage Category 2 patients are discharged home from the emergency department) (Sydney South West Area Health Service emergency department data for July 2010, for Campbelltown, Liverpool and Royal Prince Alfred hospitals). On their own, we would expect vital signs to be poorer predictors of disposition outcome than triage decisions, which have the benefit of information on presenting problem and general appearance. Context for vital sign observations has a major influence on their interpretation. We recommend that more work be done on the value of the BTF vital signs criteria as a complement, not alternative, to triage processes.

Charles H Pain · Clifford F Hughes · Marino Festa · Jodie Ekholm · Matthew O’Meara

To the Editor: Pearce and Haikerwal state that “legislation to introduce health identifiers [was] recently passed by Parliament” and that e-health “can ease the patient journey, improve quality of care and reduce costs”.1 It is critically important to comprehend the purpose of the unique health identifier. Its primary function is to be the “key” to a patient’s clinical data and information. It is not a clinical decision-support tool that will improve care. There are many systems that require multiple identifiers for a given patient. However, the success of these systems is dependent on their ability to provide information management tools.2-5 The authors correctly state that “Australia’s health care system lags behind all other sectors of our economy in the use of computerised systems”. What must be debated is their statement that, although “general practice and community pharmacy are highly computerised, the hospital sector is not”. The debate should move from the uptake numbers for computerisation to the evidence of whether these implementations have improved care — for example, through fewer adverse drug events, decreased resource use, improved quality of care and better patient outcomes. I suspect the evidence is very thin in the Australian context. The authors observe: “Uncoordinated implementation of differing, incompatible systems within and between hospitals compounds a dire lack of national coordination of effort.” This is not a new phenomenon. We actually know what works and, in some advanced systems, what does not work.6 This evidence base should provide the stimulus for pursuing the changes necessary to implement effective health information technologies. These changes are more social, professional and cultural than technical.7 In their final comments on the National E-Health Transition Authority, Pearce and Haikerwal maintain their focus on the technologies. Can we wait 5–10 years for the e-health system to be connected via the National Broadband Network? The necessary information management tools can be implemented now. Care must be a priority. It would help if the focus was on successful implementation and a willingness to listen to those who have some degree of success in this field. A review of the National E-Health Transition Authority’s website list of clinical leaders suggests that there are few clinical informaticians currently involved.

Terry J Hannan

To the Editor: E-health’s great promise is to improve health care delivery efficiency and effectiveness1 by enhancing multidisciplinary care planning and information exchange.2 General practice is now largely computerised.3 However, a chain is no stronger than its weakest link, and Australia’s otherwise slow progress towards linking secondary care into a functioning national e-health system is frustrating.1,4 We join the chorus, raising concerns about the e-health readiness of the paediatric sector. We recently surveyed the e-health readiness of paediatricians via the Australian Paediatric Research Network (APRN), a network of about 370 non-tertiary paediatricians broadly representative of all states and territories.5 Established in 2007, the APRN aims to facilitate multisite research into common chronic childhood illnesses where they are most commonly seen — in paediatric secondary care settings. Core to this goal is testing cross-sectoral care models, which requires shared e-health capabilities. The APRN’s annual web-based Multi-Topic Survey in April–May 2010 (n = 181, 48% response) contained a section on computer use with outpatients and in private rooms (Box). Almost all respondents reported having access to a computer at work. However, the reach of e-health is otherwise disturbingly low, its quality is questionable and, when used, it is mainly for clerical purposes rather than patient care. The 27% of respondents using computerised medical records reported more than 10 different systems, suggesting possible future barriers to information exchange and e-health initiatives. We applaud the rapid progress made in general practice,3 but general practitioners do not work in isolation. Our survey shows that, despite similar computer access, Australian paediatricians (and probably other physicians) are a long way behind. This has major implications for policy, shared care programs and the feasibility of e-research. E-health can only provide national benefits in our new world of chronic disease if all health sectors are involved. On a positive note, the current low rate of computerisation in paediatric patient care means that e-health implementation for secondary care could yet be coordinated and integrated from the outset. This opportunity should be seized while still possible. Resources, government funding, planning, a solution focus and the will to move forward are urgently needed. Australian Paediatric Research Network Multi-Topic Survey 2010: computer and e-health use among paediatricians Variable Proportion Computer access Access to computer at work 97% Access in each patient clinic room 85% Internet access at an acceptable speed 71% Computerised patient booking/billing 67% Clinical e-health use Computerised medical records 27% Writing new/follow-up notes 21% Ordering tests 16% Writing scripts/referrals 13%

Melissa Wake · Sarah A Davies · Harriet Hiscock · Gervase M Chaney

To the Editor: Waxman’s prompt recovery after a stroke was probably a stroke of luck,1 unless perhaps Saint Mary MacKillop was involved. Amid the conflicting evidence on the effect of thrombolysis on recovery in ischaemic stroke, one fact is beyond doubt — thrombolysis gives no greater improvement at 24 hours than placebo.2 It is the various measures of recovery at 3 months, and the risks of adverse outcomes from protocol violations in administering thrombolysis, that are less clear. The European Cooperative Acute Stroke Study III (ECASS III) showed that there is a benefit from administering thrombolysis up to 4.5 hours after stroke onset,3 and the complete Safe Implementation of Thrombolysis in Stroke International Stroke Thrombolysis Register (SITS-ISTR) data showed that this benefit can be reproduced in non-trial hospital practice.4 Simpson and colleagues say that the Australian data support this conclusion, even though outcomes show slightly higher mortality compared with rest-of-world data.5 There are several gaps in that line of reasoning. A modified Rankin score (mRS) of 0–1 (zero or minimal disability) is the standard measure of a good outcome. In ECASS III, an mRS of 0–1 was seen in 52.4% of patients who received thrombolysis compared with 45.1% of patients who received placebo.3 However, the SITS-ISTR data showed an mRS of 0–1 in just 40% of patients treated within 3 hours of stroke onset and 44% of patients treated at 3–4.5 hours.4 This is despite lesser stroke severity (National Institutes of Health Stroke Scale score: 10 in the SITS-ISTR cohort v 11.6 in the ECASS III placebo group), selective patient inclusion, and exclusion of patient data with protocol violations known to give worse outcomes. Achieving a worse result than placebo is hardly evidence for safe and effective use of thrombolysis in normal clinical practice. However, the Australian study presents the data differently, using an mRS of 0–2 and thereby concluding good outcomes in patients with greater disability than is the accepted good outcome measure.5 Rather than providing the actual data and the percentage of patients with good outcomes, a novel measure of odds ratio comparison with rest-of-world data is made. Publishing data in the same format as for the full SITS-ISTR report would allay concerns of what may be unfavourable outcomes compared with clinical trial data. Patients deserve to be told what their likely prognosis will be in hospitals that are less experienced in stroke thrombolysis than The Alfred, where Waxman received world-class stroke care.

Brendon J Smith

In reply: We were delighted to hear of Waxman’s “miraculous” recovery, and admire the world-class standard of stroke care which was delivered at The Alfred.1 The Australian data in the Safe Implementation of Thrombolysis in Stroke (SITS) database up to 2008 included tertiary referral centres and outer metropolitan hospitals, and centres with varying experience of using recombinant tissue plasminogen activator (rt-PA) in clinical practice. There was no difference in outcomes according to experience of the treating centres. We believe that the non-trial clinical data presented in our article are a realistic indication of expected outcomes for all patients receiving thrombolysis for stroke in Australia. The choice of a modified Rankin score (mRS) of 0–2 as an outcome measure in our article was largely dictated by the SITS international analysis. The aim of our study was to compare outcomes in Australia with those in the rest of the world; these were also not significantly different when an mRS of 0–1 was used as an outcome measure (mRS 0–1 at 3 months, 34.6% for Australia v 37.8% for the rest of the world; P = 0.13). The original National Institute of Neurological Disorders and Stroke (NINDS) trial data,2 now more than 15 years old, have been analysed and re-analysed and, arguably, it is time to move on from rehashing the data. However, if one wishes to follow this path, there was in fact a significant difference in improvement at 24 hours between patients in the rt-PA and placebo groups. Smith’s statement that thrombolysis gives no greater improvement at 24 hours than placebo is correct in terms of the proportion of patients with an improvement in National Institutes of Health Stroke Scale (NIHSS) score of 4 or more, but this is a low benchmark and many untreated patients also show a relatively mild improvement at 24 hours. If one considers major neurological improvement (MNI, defined as an improvement in 24-hour NIHSS score of 8 or more), this was seen in a greater proportion of patients in the rt-PA group compared with the placebo group.3 Patients with MNI have a much greater chance of achieving minimal or no disability at 3 months, which is by far the most accepted outcome measure for thrombolytic therapy. This major early recovery is frequently seen in patients who receive rt-PA, and has been termed the “Lazarus effect”. Waxman was fortunate enough to experience this effect — but it occurs far too often with rt-PA in stroke to pass the Catholic Church criteria for a miracle.

Marion A Simpson · Helen M Dewey · Mark W Parsons

To the Editor: We view with concern the listed minimum investigations for chronic suppurative lung disease and bronchiectasis in children and adults recommended in the position statement from the Thoracic Society of Australia and New Zealand and the Australian Lung Foundation, published in the Journal.1 We do not agree with the inclusion of measurement of IgG subclasses in the list. Interpretation of IgG subclasses is fraught with difficulties.2 It is unclear whether the isolated finding of low levels of one or more IgG subclasses is a risk factor for developing severe recurrent bacterial infections. Individuals with complete absence of IgG subclasses due to gene deletions may remain entirely asymptomatic, and the finding of a low or absent response to specific bacterial protein and/or polysaccharide is more closely associated with propensity to infection. In addition, the lack of reproducibility of IgG subclass assays and lack of well standardised age-specific reference ranges hinder interpretation of results. Of major concern is the lack of evidence to provide clear guidance on the efficacy of immunoglobulin replacement therapy in IgG subclass deficiency. In general, the finding of isolated IgG subclass deficiency is not sufficient grounds for commencement of immunoglobulin therapy and is not recommended as an indication for intravenous immunoglobulin therapy in the current Australian guidelines.2 We urge the authors to reconsider the inclusion of IgG subclasses in the list of minimum investigations for chronic suppurative lung disease and bronchiectasis.

David S Gillis · D Sean Riminton

In reply: We thank Gillis and Riminton for their comments highlighting the inclusion of IgG subclasses in the list of suggested investigations when reviewing a patient with chronic suppurative lung disease or bronchiectasis. We agree that a finding of one or more low IgG subclasses in isolation is not an indication for immunoglobulin therapy and that interpretation of serum IgG subclass concentrations is difficult. Indeed, this was debated at length by the position statement authors. However, after considerable discussion, we decided that measurement of IgG subclasses should be included. Although the data are weak, it is an important area of possible aetiology, and there are varying opinions regarding associations with impaired adaptive immunity.1-3 As with some other investigations, an abnormal test result in the context of a child or adult with suppurative lung disease would lead to discussion with an appropriate subspecialist for advice on interpretation and/or further evaluation. Space limitations in the Journal version of our position statement did not permit us to elaborate further.4 In the full document available online,5 we stated that: The benefits of performing IgG subclasses in isolation are particularly controversial as the role of a reduction of a subclass is likely over-diagnosed.[88-91] If IgG subclass is low, assessment of response to vaccinations is advocated.

Anne B Chang

To the Editor: In their paper, Campbell and colleagues list the potential pitfalls of introducing pay-for-performance into Australian primary care.1 They emphasise that, in Australia, the electronic medical records used in general practice would not support the introduction of a scheme such as the Quality and Outcomes Framework (QOF) that is in place in the United Kingdom. In 2009, we demonstrated that it was quite possible to apply a QOF in a large Australian general practice by adapting the in-built search facility of a commonly used medical records program.2 We concluded that the introduction of a QOF in Australia would drive up the quality of our care. The non-clinical standards of a QOF are already similar to those set by the Royal Australian College of General Practitioners for accreditation and could be readily achieved. Campbell and colleagues suggest that the QOF is expensive and without proven benefits.1 This is incorrect. As predicted, the QOF has not only led to a reduction in cardiovascular disease events, but the benefit is greatest in the lowest socioeconomic groups.3 It is now possible to measure the cost effectiveness of a QOF, which is likely to show that it pays for itself by way of reduced morbidity and health service costs.4 Rather than argue for general practitioner remuneration based on pay-for-performance to be less than 20%, one should err on the side of caution, which recognises that 20% has ensured a massive public health gain for the British population. Now that 95% of GPs use computers, at least for prescribing,5 it is plausible that an Australian QOF could be developed, but pay-for-performance may need to include some pay-for-data incentives.

Mark A J Morgan · James Dunbar

In reply: Our article, which related to clinical care, not non-clinical standards, was intended to generate debate, and the points raised by Morgan and Dunbar are important ones. We address the central issue, which relates to the intended purpose of introducing a pay-for-performance scheme, similar to the Quality and Outcomes Framework (QOF) in the United Kingdom, into Australian general practice. Morgan and Dunbar state in their letter, “Now that 95% of GPs use computers for prescribing”; this emphasises that it would not be a level playing field. Even if all indicators only related to prescribing, at least 5% of practices would be disenfranchised. That statement also appears to presume that 95% of practices use compatible and comparable clinical codes. The UK witnessed years of computerisation and clinical code usage across a wide range of clinical issues before the introduction of the QOF. Elliot-Smith and Morgan adapted “the in-built search facility of a commonly used medical records program”, and introducing a QOF-like scheme in Australia would involve this investment in all practices. The percentage of income attributed to pay-for-performance is a political decision, and most people who are involved in the QOF in the UK strongly believe that it is too high. Whatever level is set, it must result in a level playing field for practices.

Stephen M Campbell · Anthony Scott · Rhian M Parker

To the Editor: Although we welcome Boyd and colleagues’ recommendations for changes to infection control guidelines for Australian recipients of human growth hormone (hGH) and human pituitary gonadotrophin (hPG),1 we are concerned about the accuracy of the history provided and do not want it misrepresented. Boyd and colleagues refer to recipients living with anxiety since 1985, when the Australian Human Pituitary Hormone Program (AHPHP) was halted. However, Australian recipients were not aware of the risk until 1991–1992. After the first two recorded deaths from Creutzfeldt–Jakob disease (CJD) of women treated with hPG, their husbands fought for official recognition of the link between their deaths and the AHPHP. In 1992, media attention alerted recipients to the risk and initiated contact from treating doctors. Knowledge that this information had been withheld since 1985 angered recipients. Boyd et al’s article is contradictory in mentioning a possible single, discrete contamination event but also claiming that a higher-risk period is unknown. We challenge this — particularly on behalf of recipients who were informed they had received hormones from a batch identified as likely to have contained the transmissible agent that causes CJD. We dispute the stated lack of availability of detailed treatment information, as the Australian Government Department of Health and Ageing has meticulously obtained and archived patient notes of recipients from their treating doctors, which provide the most accurate source of data. Boyd and colleagues had access to records of the four women whose deaths resulted from AHPHP treatment. The records seem consistent with published data,2,3 confirming that the four women were treated between 1973 and 1978. This surely indicates a high-risk period and gives extra confidence to those treated outside this timeframe. Further, two of the three patients on whom autopsies were performed received treatment from an identical batch of hormones (44); one was treated with this batch only. The third patient received hormones from batches 43 and 45.2 Batch 44 was identified as contaminated, and it was theorised that tainted material was transferred to batch 45 via production equipment. The patient who did not undergo an autopsy received only batch 25, which was produced earlier and indicates another possible contamination event within the 1973–1978 period. Risk estimate percentages do not provide more confidence for recipients than is currently already growing. They do not bring back those who died, take away the guilt of the parents who consented to hGH therapy for their children, or change a history of suicides, broken marriages, anxiety, disgrace and despair. It is important that Boyd and colleagues do not blur the implications of their findings with the continued need for the Australian Government to maintain the Human Pituitary Hormone Trust Account, which was established to provide funding for counselling and support services for recipients, and ongoing support for this group of citizens whose lives changed forever and who continue to battle discrimination and delays when accessing health care. The recommendations inspire hope but provide no immediate relief. The passing of time brings confidence, but the record needs to respect the enormous impact on the lives affected.

Suzanne L Solvyns · David W Ralston

In reply: We thank Solvyns and Ralston for their thoughtful and heartfelt comments and their assistance in ensuring historical accuracy regarding the issues surrounding iatrogenic Creutzfeldt–Jakob disease (CJD) risk and recipients of the Australian Human Pituitary Hormone Program (AHPHP). Whether there is a need for ongoing support of AHPHP recipients for the foreseeable future was not considered in the scope of our report,1 but we unequivocally support provision of optimal health care services for them, which we believe relaxation of infection control measures will assist. The delays and manner in which many AHPHP recipients learned of their risk of CJD is certainly regrettable, but this was not the universal experience of recipients. Some were made aware of the risk by their treating practitioner in 1985,2 with their heightened personal anxiety commencing from that time. As stated in our report, we agree that the close temporal development of CJD in the four Australian human pituitary gonadotrophin recipients appears consistent with a single, significant contamination event,1 such as from tainting of a limited number of treatment batches — an observation broadly in keeping with the existence of a higher-risk treatment window. However, despite assertions to the contrary, data acquired from various sources by the Australian National Creutzfeldt–Jakob Disease Registry for analysis did not permit confident and consistent delineation of a higher-risk treatment group. This is not synonymous with saying that a higher-risk period was unknown to us.

Alison Boyd · Genevieve M J A Klug · Colin L Masters · Steven J Collins

To the Editor: Dinh and colleagues present data on head injuries among cyclists admitted to the emergency department of Sydney’s Royal Prince Alfred Hospital (RPAH) from 1991 to 2010,1 and conclude with their opinion that legislation mandating the wearing of bicycle helmets should be maintained. However, their data provide poor evidence of the efficacy of helmets and of the effectiveness of legislation requiring helmet use. Almost all studies of helmet efficacy are case–control studies, which, to be of high quality, require a well defined and matched control group. In the study by Dinh et al, the control group was made up of patients who had been wearing a helmet. Many possible explanations for injury severity were not considered. For example, were the non-helmeted cyclists drinking alcohol or riding faster? Were they more likely to be risk-takers? Were the helmet wearers also wearing high-visibility vests or using lights at night? Case–control studies of this kind are a weak form of evidence (ranking just above anecdotal evidence and case studies) and cannot answer questions of causality. The two graphs presented by Dinh et al show that head injury rates among cyclists were very low and fairly stable over most of the time period examined. RPAH treats thousands of trauma patients each year. The study reported on 287 cyclists admitted in a 2.5-year period — an average of 115 per year. Of these, only 14 (eight helmeted; six not wearing helmets) incurred major trauma — about six per year. Given that helmet legislation was in place for the entire study period, these data add no evidence for its effectiveness. In public health terms, the risk of injury needs to be weighed against the benefits of increased physical activity, reduced chronic disease and reduced air pollution due to more people cycling. Because of the low absolute rate of cycling injuries, these benefits are nine to 20 times the size of the risk.2,3 Further, it is well documented that levels of cycling in Australia dropped by 30%–40% when helmet legislation was introduced. This had the unfortunate effect of making cycling less safe for the remaining cyclists because of the “safety in numbers” phenomenon.4 The poor uptake5 of the two public bicycle-hire schemes in Australia (in Melbourne and Brisbane) indicates that helmets are a significant barrier to short-trip and spontaneous cycling, with few potential or casual users carrying a bicycle helmet in case they decide to hire a bicycle. Dinh and colleagues obviously believe in the efficacy of the bicycle helmet, but this does not mean that mandating their use is the best public health policy response. When helmet legislation was introduced, we did not have the epidemics of diabetes and obesity that now threaten to bankrupt state health budgets. Strategies to increase levels of physical activity among the whole population are desperately needed, even if there is a small risk involved.

Chris E Rissel · Paul Martin

To the Editor: About 5% of Australians of Asian, African, Middle Eastern or Mediterranean descent have glucose-6-phosphate dehydrogenase (G6PD) deficiency.1 Affected babies can develop massive haemolysis within hours of exposure to clothes stored with mothballs containing naphthalene. It has long been known that this results in severe jaundice, which may lead to kernicterus2 and profound brain damage, for which the cost is either a lifetime of dependency and very expensive care, or death. We are aware of three cases of kernicterus in babies with G6PD deficiency in Australia in the past 3 years, one of which was associated with exposure to naphthalene in mothballs. One baby died. The exact incidence of severe neonatal jaundice and kernicterus in Australia is unknown, but it is the subject of an ongoing study funded by the Cerebral Palsy Foundation and coordinated through the Australian Paediatric Surveillance Unit. In Australia, packages of naphthalene mothballs must carry a warning that the product is harmful to children. However, clinical directors of neonatal units that comprise the Australian and New Zealand Neonatal Network have unanimously agreed that warning labels give insufficient protection. They have called on the Australian Pesticides and Veterinary Medicines Authority (APVMA) to act in harmony with the European Union, which banned the sale of mothballs containing naphthalene in 2008,3 following a report by the European Chemicals Bureau.4 The adverse risk–benefit ratio for naphthalene provides strong justification for its withdrawal. A submission to this effect has been lodged with the APVMA. Some mothballs contain paradichlorobenzene, a chemical related to naphthalene and associated with haemolysis. Less toxic products that protect clothes against moths exist. Department stores in the United Kingdom have replaced moth repellents containing naphthalene with products containing natural substances, such as sandalwood and lavender. Between 2004 and 2010, the New South Wales Poisons Information Centre reported that it received about one call per week concerning children exposed to naphthalene in moth repellents (Box). The Victorian Poisons Information Centre reported 53 calls in 2008.5 While acknowledging the importance of raising awareness of the dangers of naphthalene, we believe that the safest course is prevention — that is, an Australia-wide ban of mothballs containing naphthalene. Readers who wish to report cases of naphthalene toxicity are encouraged to contact APVMA at aerp@apvma.gov.au. Number of calls to the New South Wales Poisons Information Centre reporting children exposed to napthalene in moth repellents, 2004–2010 Year Number of calls 2004 55 2005 59 2006 65 2007 67 2008 73 2009 45 2010 71 Total (average) 435 (62) Source of data: Judith Kirby, Department Head, NSW Poisons Information Centre, personal communication.

on behalf of the Advisory Committee of the Australian and New Zealand Neonatal Network

To the Editor: We agree with Mendelson and colleagues1 about the need to ensure radiation doses of computed tomography (CT) are as low as reasonably achievable while maintaining diagnostic quality of the images. Perceived benefit to the patient must be balanced against theoretical small risks from the procedure. The uncertainty surrounding risk estimates at the lower end of the radiation dose range means that it is important for radiologists to take responsibility for monitoring radiation dosimetry. This becomes especially important when imaging young people with chronic disease (such as renal calculi, chronic hepatitis, cystic fibrosis and inflammatory bowel disease), who may have many CT scans while young. As stated by Mendelson et al,1 the theoretical risks are assumed to be cumulative over a lifetime, but diminish significantly the older the person is at time of exposure. We would like to highlight work that the Royal Australian and New Zealand College of Radiologists (RANZCR) has undertaken to support the development of CT scanning standards, professional education, dose optimisation and the establishment of national diagnostic reference levels by the Australian Radiation Protection and Nuclear Safety Agency. CT dose optimisation quality improvement programs have been conducted in Queensland and Victoria. These have been funded through a grant to the RANZCR Quality Use of Diagnostic Imaging (QUDI) Program from the Australian Government Department of Health and Ageing. A dose optimisation project is planned for South Australia in 2011. The RANZCR QUDI program has developed a consumer and referrer information database of diagnostic imaging procedures, including a general item about the theoretical risks associated with ionising radiation. This information is accessible at http://www.insideradiology.com.au and supports informed choices about imaging. While Blecher2 and Mendelson et al1 are correct in indicating that the latest multidetector computed tomography (MDCT) platforms have the potential to deliver acceptably low radiation doses and high-quality images, this is not always seen in practice, based on the evidence from these optimisation programs. Some multiphase MDCT protocols have the potential to give rise to organ doses in the range where there is little doubt that a small increased risk of carcinogenesis exists.3,4 Contrary to the comment made by Blecher,2 our knowledge of the impact of radiation on humans does not arise exclusively from atomic bomb survivors. For example, we have evidence of increased breast cancer risk from patients who underwent multiple fluoroscopic examinations to monitor treatment for tuberculosis.5 Risk estimates from this and other datasets are broadly consistent with those from atomic bomb survivor data. In view of the above, we need to be vigilant in ensuring that CT examinations are clinically justified and, secondly, performed using optimised scanning protocols. Referring practitioners need to know that the potential stochastic effects resulting from CT examinations are cumulative over a lifetime. It is this cumulative effect that is not necessarily trivial.

John C P Heggie · Stacy K Goergen · Michael J Fallon

To the Editor: At a time of apparent increasing concern over radiation exposure, the article by Keijzers and Britton1 unfortunately does not provide any perspective on the clinical necessity of the imaging investigations requested by emergency department doctors. It also contains inaccurate information. Stating that “multidetector CT [computed tomography] scanners have the potential to expose patients to higher radiation doses”, and preceding it with “radiation doses per scan have increased by up to 40%”, is incorrect. CT scanner manufacturers have for some time been producing scanners that have decreased radiation doses by incorporating various techniques,2 including iterative reconstruction (a reconstruction algorithm that corrects image data by using various models, thereby improving CT image quality while reducing radiation dose). The dosage reductions reported by major manufacturers and in the literature2 vary between 25% and 80%. The efficacy of dose-reduction mechanisms and the concepts behind them have been outlined in numerous white papers and scientific articles, leading to the conclusion that “newer scanners have improved technology that facilitates lower patient doses”.3 In their seminal article, McCollough and colleagues caution that “alarmist articles” do the public a disservice. They also state that low-level radiation risk estimates, which are derived primarily from atomic bomb survivors and have considerable uncertainties at low doses (< 100 mSv), give no consideration to the medical benefit of a CT scan.2 While Keijzer and Britton’s research investigated emergency department doctors’ knowledge of radiation exposure, the sample was small, being restricted to one institution, and the study had other limitations, as acknowledged by the authors.1 The alarmist undertones have not been helped by the publication of articles in the lay press, such as “Doctors warned over X-ray risks”.4 The danger is that sensationalising the risks of CT scanning may lead to patients not undergoing scans that are clinically indicated, with consequent patient harm. The Royal Australian and New Zealand College of Radiologists (RANZCR) has issued a position statement on radiation dosage.5 As radiologists, we abide by the ALARA (as low as reasonably achievable) principle, and the RANZCR is actively pursuing ways of decreasing radiation exposure.5 In a clinical context, the risk from CT scans is small compared with health risks arising through other activities. Educating doctors about radiation exposure risks is crucial, but so is up-to-date information and common sense, devoid of sensationalism.

Alain M Lavoipierre