Letters

Michael J Sinnott Emergency Physician, Princess Alexandra Hospital, Ipswich Road, Woolloongabba, QLD 4102 michael_sinnottAThealth.qld.gov.au To the Editor: In physiology, the Frank–Starling curve demonstrates that cardiac muscle initially responds to an increased workload with an increased force of contraction.1 However, after the point of maximum efficiency is reached, further workload produces a decrease in both the force of contraction and the ejection fraction, leading to cardiac failure. It now appears that the same curve could describe the current situation in many Australian emergency departments. We used to operate at point A on the curve (Box). If there was a mini-disaster or a moderately large number of victims of a road accident, the department was able to increase output to cope with the situation. The “adrenalin stimulation” experienced by all members of the team meant that the department coped, and that staff were left with a sense of satisfaction. Now our department finds itself at point B on the curve. Extra workload can result in a decrease in performance and output. The patients obviously suffer, but so do the staff. The once-challenging and enjoyable parts of the job now generate frustration and exacerbate the background dysfunction. On a recent weekend, the emergency department experienced an influx of sick elderly patients as a result of a local heatwave, with temperatures reaching 42°C. The problem was identified as a mini-disaster only in retrospect. At the time it was thought to only exemplify another bad day. This is an example of the syndrome of “learned helplessness”2 that staff are experiencing. Politicians and health administrators need to understand that our public hospital emergency departments are struggling with their daily workloads and are no longer equipped to deal with medium- to large-scale emergencies. Access block as a medical model At point A, an increase in workload leads to increased performance to cope (moving to point A1). At point B, an increase in workload leads to a decrease in performance (to point B1).

Michael J Sinnott

Stacey L Emmett,* Joseph E Ibrahim,† Amanda Charles,‡ David L Ranson§ * Research Officer, † Physician, ‡ Clinical Research Nurse, § Deputy Director, Clinical Liaison Service, Victorian Institute of Forensic Medicine and the State Coroner’s Office, 57–83 Kavanagh Street, Southbank, VIC 3006 staceyeATvifm.org To the Editor: The Royal College of Pathologists of Australasia Autopsy Working Party highlighted the decline in the numbers of hospital autopsies.1 Forensic and hospital autopsies are a valuable safety and quality tool for improving healthcare systems. Autopsies provide an accurate cause of death and are a valuable audit tool to evaluate medical diagnostic processes and therapeutic interventions. Declines in both types of autopsies are cause for concern. A forensic autopsy is an integral part of the coronial process. The Coroner’s role is to establish the identity of the deceased, where he or she died, the cause of death and, perhaps most importantly, how the person died.2 Without an autopsy, it can be difficult to determine the cause of death. At the Victorian Institute of Forensic Medicine (VIFM), about 80% of all deaths that are reported to the State Coroner (in Melbourne and Geelong) undergo a full forensic autopsy. Senior next-of-kin can object to an autopsy being performed. Section 29 of the Coroner’s Act 1985 (Vic) details the objection process.2 The decision to grant an objection is dependent on the opinion of the Coroner and his or her view on the circumstances of death. We reviewed the number and rate of forensic autopsies performed by the VIFM between 1992 and 2002, as well as the number and rate of objections under Section 29. The rate of forensic autopsies remained relatively stable over the decade, at 80% of deaths reported to the State Coroner. There were 94 successful Section 29 applications in 1992. This accounted for 3.25% of all deaths reported to the Coroner. Objections to autopsies have been steadily increasing since 1992. In 2002, there were 212 successful applications, accounting for 7.1% of all deaths reported to the Coroner and representing about a 4% rise over the decade. When a coronial autopsy has been requested, but not carried out because of a Section 29 objection, substantial additional work is required. This includes medical record reviews; external forensic examinations; reviews of statements from treating doctors, independent experts and family; and inquests. The other major ramification associated with objections to autopsies is the discrepancies between causes of death that are determined clinically and at autopsy (with 28% of presumed causes wrong in one study).3 Without an autopsy, important pathology may remain unrecognised, and this can substantially affect the accuracy of the stated cause of death. The inherent right of next-of-kin to object to coronial autopsies will remain. However, healthcare professionals and coroners need to be aware of the public health implications associated with objections to autopsies.

Stacey L Emmett · Joseph E Ibrahim · Amanda Charles · David L Ranson

Gaston R B Arnolda Honorary Associate, Department of Public Health, University of Sydney, QEII Institute for Mothers and Babies, Building DO2, Sydney, NSW 2006. garnoldaATperinatal.usyd.edu.au To the Editor: O’Grady and Noland, concerned about the consequences of privacy legislation,1 draw attention to “. . . the findings of an Australian survey in which 61% of adults believe that even their de-identified health information should not be used for research purposes without their consent”. Unfortunately, they do not point out that the survey,2 commissioned by the Office of the Federal Privacy Commissioner, had a 20% response rate, making it effectively useless in determining what Australian adults really think about the use of their de-identified health information. Alarmed at the possibility that “evidence” of this quality could be used to aid decision-making that had important implications for rigorous research, I turned to Google (<www.google.com>) for assistance. A Google search using the words “61% de-identified health information roy morgan” generated 21 hits, 15 of them unique, and only five related to the subject. One of the five was the letter by O’Grady and Noland, one was the report of the survey, and three specifically cited this survey result: Privacy Victoria,3 Privacy NSW,4 and the Office of the Federal Privacy Commissioner5 all used the result in formal submissions to reviews of privacy-related issues — without revealing the survey’s appalling response rate. We have privacy commissioners who are powerful advocates of the principles of respect for privacy and autonomy. Perhaps the time has come for Australia to have “public interest commissioners” who can powerfully advocate for the public interest in high-quality health research.

Gaston R B Arnolda

Gary C Morgan Executive Chairman, Roy Morgan Research Pty Ltd, PO Box 2282U, Melbourne, VIC 3001. Gary. MorganATroymorgan.com In reply: In a world in which people are increasingly busy and mobile, increasingly concerned about invasions of privacy and increasingly approached to participate in surveys, survey response rates that could be readily achieved 25 years ago are now very much more difficult to attain. High response rates (ie, 60% and over) are still very desirable and can still be achieved. We have conducted surveys on sensitive issues, such as drug-taking, and achieved response rates that would probably satisfy even Arnolda. But this requires very intensive field activity that is not always justified by the nature of the project. The survey in question was not a health survey. It set out to provide general background information, exploring comparative levels of concern about, and the relationships between, a very wide range of privacy issues on a scale adequate to allow relatively small groups within the population to be examined. It was not intended to yield precise and critical measurements. Measures of “concern” or “reluctance” are highly context-dependent, “soft” measures, subject to interpretation, both by the respondents and by end-users. External validity is therefore not the issue: a response rate of 80% would not have made the figures demonstrably more “accurate”. Given the imprecise nature of the measures obtained, overengineering the sample relative to other components of the survey design would have been a waste of (public) money, better devoted to further research. The survey was part of a wider-ranging project and was planned in close consultation with the Office of the Privacy Commissioner in the light of their needs and priorities. It is appropriate that this issue arose out of a debate on privacy. As privacy constraints bite harder, whether imposed by statute or codes of practice, or arising through increasing resistance from subjects, medical researchers are going to have to come to terms with problems of non-response in the same way that social-survey researchers have. How are researchers going to react when only 20% of potential subjects consent to have their information used? Will they use words like “appalling” and “effectively useless” to dismiss any studies based on such a subset? Or will they perhaps learn to use them, with due caution, for the valuable information they nevertheless contain? One does not have to look very far in the history of medicine or public health to find major advances in knowledge using less than perfect statistics.

Gary C Morgan

Sharon J Gardiner,* Evan J Begg,† Carl M J Kirkpatrick,‡ Robert B Buckham¶ * Drug Information Pharmacist, † Professor of Medicine, Christchurch School of Medicine and Health Sciences, Private Bag 4345, Christchurch, NZ; ‡ Lecturer, School of Pharmacy, University of Queensland, Brisbane; ¶ Drug Information Pharmacist, Christchurch Hospital, NZ sharon.gardinerATcdhb.govt.nz To the Editor: We wish to commend the Australasian Diabetes in Pregnancy Society (ADIPS) ad hoc working party for providing an update on the safety of metformin in pregnancy.1 However, we would like to comment on the information they provided on the safety of this drug in breastfeeding. Metformin can be regarded as a well studied drug with respect to its distribution into human breastmilk;2,3 most drugs are not as well served in this regard. As Simmons et al indicated,1 the infant “dose” in breastmilk is small at less than 0.4% of the maternal dose, corrected for body weight. This is substantially lower than the arbitrary cut-off of 10% used to guide drug use during lactation and thus implies safety.4 Further evidence for the safety of this drug in breastfeeding arises from failure to detect metformin in blood sampled from four of six infants exposed via breastmilk (limit of detection, 5–10 μg/L) and lack of adverse effects noted in nine exposed infants.2,3 Simmons et al stated that infant exposure to metformin could be reduced by breastfeeding immediately before maternal dose ingestion and then avoiding feeding for at least 2–3 hours after the dose. For drugs with a short elimination half-life, this recommendation — avoiding feeding at peak drug concentrations in the milk — may reduce infant exposure. However, this is not the case for metformin. Two studies investigating metformin in breastfeeding have shown that the metformin peak plasma concentration occurs about 2–4 hours after the dose, while milk concentrations are “flat” across the entire dosing interval. This is distinctly different from most drugs, in which the drug concentrations in milk mimic the rise and fall of plasma concentrations, consistent with passive diffusion.2,3 The flat profile observed with metformin raises the possibility that the distribution of metformin into or out of breastmilk may involve an active process such as organic cation transporter(s), in addition to passive diffusion Given this unusual concentration profile in breastmilk, infant exposure (albeit small) will not be reduced by the practice of avoiding breastfeeding for a few hours after maternal dose ingestion, as suggested by Simmons et al. In other words, mothers may feed their infants at any time during the dosing interval and this will not affect infant exposure to metformin. We believe that there is sufficient evidence for metformin to be considered a safe therapeutic option in the treatment of diabetes or polycystic ovary syndrome in breastfeeding mothers, with the usual caveat of weighing up the risk–benefit ratio in each case.

Sharon J Gardiner · Evan J Begg · Carl M J Kirkpatrick · Robert B Buckham

David Simmons,* Barry N J Walters,† Janet A Rowan,‡ H David McIntyre§ * Professor of Medicine, Waikato Clinical School, Waikato Hospital, Hamilton, NZ; † Clinical Associate Professor, Department of Women’s and Children’s Health, King Edward Memorial Hospital, Subiaco, WA; ‡ Physician, Department of Obstetrics, National Women's Hospital, Auckland, NZ; § Director of Endocrinology, Mater Hospital, Brisbane, QLD. simmonsdATwaikatodhb.govt.nz In reply: We thank Gardiner et al for their commendation and support for our update on the safety of metformin in pregnancy.1 We agree with their analysis regarding the timing of the use of metformin during lactation. Nevertheless, we would like to highlight that the safety of metformin can not be assumed from the studies they quote, as these included very few subjects. Such studies, helpful as they may be, provide no imprimatur for the long-term safety for the growing infant and subsequent adult. While no babies had side effects reported during these studies, this may not be the case for other babies. Should a woman decide against the use of insulin to control hyperglycaemia postnatally, then the risk of potential known and unanticipated side effects of metformin should be discussed while obtaining informed consent for metformin use. However, during this discussion, it would also be prudent to weigh-up metformin use against breastfeeding with continued hyperglycaemia, an activity associated with greater obesity and impaired glucose tolerance in the offspring.2

David Simmons · Barry N J Walters · Janet A Rowan · H David McIntyre

Clive D Hadfield Gastroenterologist, 30 Megan Street, Cairns, QLD 4870. chadfieldATaustarnet.com.au To the Editor: In their recent study, Scott and colleagues demonstrated benefit from a program to standardise clinical management of cardiac conditions in Queensland hospitals.1 They found differences in the effect on “low-intensity intervention” hospitals compared with “high-intensity intervention” hospitals. The former were, by and large, district-type hospitals and the latter tertiary hospitals. The study found that about 50% more patients in the larger hospitals had assessments of left ventricular function. Three times as many patients in the larger hospitals accessed rehabilitation. Nearly three times as many patients in the smaller hospitals were readmitted with a diagnosis of acute coronary syndrome within 30 days, perhaps a surrogate for angiography rates, which were not reported differentially. It may be that the most urgent intervention required is “high-intensity” funding of district hospitals, so that they can achieve rates of echocardiography, rehabilitation and coronary angiography approaching those of tertiary hospitals. This intervention would need no further justification than that the population served by the district hospitals has paid its share for these treatments. Let us hope that the remaining comparative outcome data are published.

Clive D Hadfield

Ian A Scott,* Irene C Darwin,† Kathy H Harvey,‡ Andy B Duke,§ Nicholas D Buckmaster,¶ John Atherton,** Hazel E Harden,†† Michael Ward,‡‡ for the CHI Cardiac Collaborative * Director of Internal Medicine, Princess Alexandra Hospital, Ipswich Road, Woolloongabba, QLD 4102; † Program Manager, ‡ Project Manager, § Senior Analyst, Collaborative for Healthcare Improvement, Queensland Health; ¶ Director of Medicine, Caboolture Hospital; ** Director of Cardiology, Royal Brisbane Hospital; †† Program Manager, Integrating Strategy and Performance, Queensland Health; ‡‡ Program Director, Queensland Health Skills Development Centre, Royal Brisbane Hospital. ian_scottAThealth.qld.gov.au In reply: We agree with Hadfield that optimising cardiac care may require extra resources targeted at increasing access of patients in regional Queensland to specific interventions, such as coronary angiography, cardiac rehabilitation and echocardiography, in addition to the quality-improvement strategies used within our collaborative. We contend that both approaches are necessary, and that the magnitude of improvement achieved by either will depend on the intensity with which they are applied. Indeed, the “high-intensity” quality-improvement hospitals in our study were defined on the basis of more funding being made available to undertake quality-improvement activities at those sites. We concede that some of the differences in quality indicators between “high-intensity” and “low-intensity” quality-improvement hospitals may be attributable to inequities in capital expenditure on service delivery that we did not measure. However, some of the differences may have also arisen from variation in systems for identifying and referring those patients who have most to gain from receiving the care targeted by our collaborative.

Ian A Scott · Irene C Darwin · Kathy H Harvey · Andy B Duke · Nicholas D Buckmaster · John Atherton · Hazel E Harden · Michael Ward

Barry E Christophers Retired General Practitioner, 1/12 Tollington Avenue, East Malvern, VIC 3145 To the Editor: I write concerning the recent article about my letters to the MJA in the 1950s and 1960s drawing attention to Indigenous health issues.1 Mention is made in the article of the campaign waged by the Federal Council for the Advancement of Aborigines and Torres Strait Islanders concerning the exclusion of Queensland Aboriginal patients with tuberculosis from the generous allowance paid to other TB patients. This campaign was successful. The Tuberculosis Act was amended so that Aboriginal people were not excluded from receiving this allowance. The Australian Medical Association supported this campaign. Without its support it would have failed.

Barry E Christophers

Subhash C Arya,* Nirmala Agarwal† * Clinical Microbiologist; † Chief of Gynaecology and Obstetrics, Sant Parmanand Hospital, 18 Alipore Road, Delhi-110054, India. subhashjiAThotmail.com To the Editor: The recent editorial by Burgess and McIntyre on the recommended vaccination schedules in Australia1 points to the fiscal constraints on offering the new, costlier vaccines. Overcoming these constraints would not be insurmountable if vaccinations were to be linked with annual festivals and celebrations in the life of individuals and the community. Birthday celebrations are important for infants and preschool children. Rather than giving conventional birthday gifts, varicella vaccine, costing $40, would be most appropriate. For those in the sixth or higher decades of life, gifts of influenza vaccine, 23-valent pneumococcal polysaccharide vaccine or the adult formulation of the diphtheria–tetanus vaccine would be memorable on Mother’s Day or Father’s Day and silver, golden or platinum wedding anniversaries. Similarly, slight adjustments to the allocation of funds for celebrating festivals such as Christmas and New Year, Dewali or Eid could make costly vaccines available to all. Vaccine producers, like department stores, could gear up for a Christmas vaccine sale. The public should be motivated to consider vaccines the most appropriate gifts. This is bound to address any poor coverage of costlier vaccines, such as varicella or the pneumococcal polysaccharide vaccine.

Subhash C Arya · Nirmala Agarwal

Ronan J Murray,* Kishore Sieunarine,† Peter B Ward,‡ John W Pearman§ * Senior Microbiology Registrar, § Clinical Microbiologist, † Vascular Surgeon, Royal Perth Hospital, Perth, WA; ‡ Senior Scientist, Department of Microbiology, Austin Repatriation Medical Centre, Heidelberg, VIC. ronan.murrayAThealth.wa.gov.au To the Editor: Previous articles in the Journal have described the emergence of Staphylococcus aureus with reduced susceptibility to vancomycin (also known as heteroresistant vancomycin-intermediate Staphylococcus aureus, or hVISA) in populations where methicillin-resistant S. aureus (MRSA) is endemic in healthcare settings.1,2 We describe a case of infection caused by hVISA from a region where healthcare-associated MRSA infection is relatively uncommon.3 A 79-year-old woman with an extensive medical history, including type 2 diabetes mellitus and multiple bypass procedures for lower-limb ischaemia, presented with critical ischaemia of the right lower leg. After above-knee amputation, she developed a discharge from the stump wound from which multiresistant MRSA was cultured. Despite receiving several courses of intravenous vancomycin (a total of 25 days of therapy over 5 months), the infection did not resolve. Extensive debridement surgery, with removal of multiple grossly infected vascular grafts, was performed, and MRSA was cultured from the graft material. Subsequently, the patient developed a discharging sinus from which MRSA with reduced susceptibility to glycopeptide antibiotics was cultured (vancomycin minimal inhibitory concentration [MIC], 8 mg/L; teicoplanin MIC, 24 mg/L). This isolate was shown to be hVISA by population analysis profiling (PAP). When the original MRSA isolate was subsequently tested by PAP, heterogenous subpopulations of bacteria with reduced susceptibility to vancomycin were present which had not been detected by routine susceptibility testing (ie, the isolate was already hVISA). Review of the patient’s medical records from other Perth healthcare institutions revealed no evidence of vancomycin administration before the initial isolation of MRSA, or contact with known MRSA-colonised patients or healthcare workers. Multilocus sequence typing and staphylococcal cassette chromosome mec allotyping identified the MRSA strain as ST239-MRSA-III, a multiresistant “international” MRSA clone frequently isolated in Australia, mainly on the east coast.4 Despite further surgery and institution of alternative antimicrobial therapy (initially rifampicin and fusidic acid and subsequently linezolid), the patient died of ongoing ischaemia and uncontrolled infection. Prolonged or repeated use of vancomycin in patients with implanted prostheses that are infected with MRSA should be discouraged, not only because it is commonly futile, but also because it may promote the emergence of subpopulations of S. aureus with reduced susceptibility to vancomycin, as occurred in this case. The fact that these resistant subpopulations were detected in an isolate before the commencement of vancomycin therapy (and then only with specialised testing) reinforces our recommendation.

Ronan J Murray · Kishore Sieunarine · Peter B Ward · John W Pearman

James D Kidd Retired General Practitioner, The Medical Centre, 125 River Road, Emu Plains, NSW 2750. To the Editor: Although I am retired, I wish to make some comment on the controversy about the use of ciprofloxacin in the treatment of chronic suppurative otitis media.1 I still feel uneasy when there is mention of the topical use of an antibiotic that may be used either orally or parenterally. Many years ago, when I had a large practice, including paediatric patients, chronic suppurative otitis media was common, although most cases responded to the classical ear drops. However, some persisted and, not infrequently, a new patient would present with this problem. As mentioned, most were the result of a pseudomonas infection. Pseudomonas was then a common problem in chronic leg ulcer of the elderly, and I had found that treatment with Burrow’s solution (aluminium sulfate [2.25 g], acetic acid [33%], tartaric acid [0.45 g], calcium carbonate [1 g], purified water [7.5 mL]) was very successful and continued to be successful right up to my retirement, even in new cases which had been treated unsuccessfully with ciprofloxacin. Before the advent of ciprofloxacin, I used Burrow’s solution ear drops in many adult cases of chronic suppurative otitis media with great success. As the number of children with this problem grew, I attempted to get advice on the use and any toxicity of Burrow’s solution ear drops in children. I could not find anyone at the Children’s Hospital with any experience, but the consensus was that it was unlikely to be toxic. Although at first I had difficulty in getting the chemist to make up ear drops for adults, by this time there was little problem getting them made for children, and the results were dramatic. There were recurrences, but these responded as well as they did the first time they were treated. Advice on correct aural hygiene after swimming and bathing was important. Burrow’s solution kills pseudomonas. Am I too old fashioned?

James D Kidd

W Ian Cameron Chief Executive Officer, New South Wales Rural Doctors Network, Level 3, 133 King Street, Newcastle, NSW 2300. icameronAT.nswrdn.com.au To the Editor: Bourke is a small rural town in far northwest New South Wales with a population of about 3500. The Bourke Shire Council has recently published a fascinating small book entitled 100 lives of Bourke, which uses Bourke cemetery as a “window to the past”.1 Included in the 100 vignettes based on headstones in the cemetery is one for Dr George Faithfull, Bourke’s third doctor. Dr Faithfull was born in Calcutta, India. He started medical studies in Edinburgh, Scotland, but did not complete them. He travelled to Australia, arriving in Victoria, and worked his way north (Dr Don Faithfull, grandson of George, personal communication). By 1885, he was working as a chemist in Bourke.2 In view of his experience as a “medical man” and his previous medical studies, he later became registered as a medical practitioner under Clause 3 of Act 70, The Medical Practitioners Further Amendment Act of 1900, where it was stated that: It shall be lawful for the Medical Board or its Doctors to place upon a separate register the name of any person who has passed through a course of study as Medical Practitioner in NSW during five years before the passing of this Act. Dr Faithfull was a doctor in Bourke and Goodooga from 1900 to 1908. He must have been one of the very first doctors to have gained what is now termed in NSW “conditional registration in an area of need”. To have become conditionally registered while practising as a chemist in Bourke, he must also be an early example of rural-based distance learning — a cornerstone of our new regionalised general practice training. Dr George Faithfull and his wife, Mary Faithfull (née Whitfield). (Photo, Dr Don Faithfull.) It is sobering to think that the changes made over the last few years by medical boards, and the streamlining of processes outlined in MedicarePlus, had precursors over a hundred years ago. Doctors and their families were not exempt from the morbidity and mortality of the time. Three of Dr Faithfull’s 11 children died young and are buried in Bourke cemetery. Another Bourke doctor, Dr Sides, had three children die in infancy, and Dr Dey lost a six-month-old son. In 1901, a locum doctor died of heat stroke. More recently, in 1992, Bourke cemetery became the final resting place for another overseas-trained doctor, Professor Fred Hollows, who achieved international recognition for his work in ophthalmology in disadvantaged populations, particularly Indigenous Australians.

W Ian Cameron

Robert W Pegram,* Lisa Valenti† *Senior Lecturer, Department of General Practice, University of Adelaide, Adelaide, SA; †Senior Analyst, General Practice Statistics and Classification Unit, University of Sydney, Wentworthville, NSW. robert.pegramATadelaide.edu.au To the Editor: Young and Dobson’s article1 examining the bulk-billing status of services provided to women generated much debate.2,3 To add to that debate we undertook an analysis of 5546 Medicare-claimable general-practice encounters. Data were provided by 200 general practitioners between May and July 2002, using the BEACH (Bettering the Evaluation and Care of Health) methodology.4 We examined which encounter, GP and patient characteristics determine billing status (patient-billed or bulk-billed). From time and day of service we determined which consultations were “after hours”. Over two-thirds of services (69.8%; 95% CI, 65.4%–74.3%) were bulk-billed. One in fourteen services (7.1%; 95% CI, 2.2%–12.1%) were delivered “after hours” as defined by the Australian Government Department of Health and Ageing5 (ie, between 6 pm and 8 am on weekdays or between 1 pm Saturday and 8 am Monday on weekends). The results after simple and multiple logistic regression analysis are shown in the Box. After-hours consultations were significantly more likely to have been bulk-billed than those held during standard office hours (odds ratio [OR], 1.9). Patients aged < 15, 15–24 years and > 75 years were significantly more likely to be bulk-billed than working-age adults (P < 0.0001). Also significantly more likely to be bulk-billed were patients from non-English-speaking backgrounds (OR, 7.3), living in an urban area (OR, 2.6), holding a healthcare card (OR, 3.5) and/or coming from a low socioeconomic status background (OR, 2.3). There was no significant association between the likelihood of being bulk-billed and the age or sex of the GP, the practice size or the number of problems managed at the encounter. Interestingly, the variable with the largest impact on bulk-billing rates was whether patients were from a non-English-speaking background. These patients were over seven times more likely to be bulk-billed than patients from an English-speaking background. This study adds further support to the findings of Young and Dobson1 that patients in urban areas were significantly more likely to be bulk-billed for general practice consultations than their rural counterparts. We can go one step further and say that consultations given after hours were also significantly more likely to be bulk-billed. The conclusion is that bulk-billing decisions by GPs are not uniformly influenced by timing, location and patient characteristics. This has implications for assessing the likely impact of bulk-billing strategies such as MedicarePlus.6 Factors influencing the likelihood of bulk-billing in Australian general practice* Simple logistic regression analysis Multiple logistic regression analysis OR (95% CI) (n = 5546) Adjusted OR (95% CI) (n = 4793) Time of consultation “After hours” status (standard hours : after hours) 1.5 (0.9–2.5) 1.9 (1.1–3.3) Other variables Non-English-speaking background (no : yes) 8.8 (4.8–16.3) 7.3 (3.8–14.0) Aboriginal or Torres Strait Islander descent (no : yes) 2.0 (0.7–6.0) ns Patient new to practice (new : not new) 1.5 (1.0–2.2) ns Rural/urban place of residence (rural : urban) 2.2 (1.3–3.7) 2.6 (1.5–4.7) Having health care card (no : yes) 3.4 (2.4–5.0) 3.5 (2.3–5.2) Socioeconomic status† (higher SES : low SES) 3.2 (1.8–5.7) 2.3 (1.2–4.5) Practice size (5+ GPs : solo GP) 2.4 (1.0–5.8) ns (5+ GPs : 2–4 GPs) 1.4 (0.9–2.3) ns Patient age (years) (25–64 : < 15) 1.5 (1.1–2.0) 1.4 (1.0–1.9) (25–64: 15–24) 1.4 (1.2–1.8) 1.4 (1.1–1.9) (25–64 : 65–74) 1.8 (1.4–2.5) 1.2 (0.8–1.6) (25–64 : ≥ 75) 2.5 (1.7–3.8) 1.7 (1.1–3.8) Patient sex (female : male) 1.1 (0.9–1.3) ns GP age (years) (25–54 : ≥ 55) 1.6 (1.0–2.8) ns GP sex (female : male) 1.1 (0.7–1.8) ns Number of problems managed per encounter‡ 1.0 (0.9–1.2) ns GP = general practitioner. ns = not significant at 5% level. OR = odds ratio. * For each variable, the first-mentioned category is the reference. † Assessed by SEIFA (Socioeconomic Indexes for Areas) categories of the Australian Bureau of Statistics. ‡ The reference point for this variable is the number of problems managed (1, 2, 3 or 4), measured against whether the patient is bulk-billed. The OR here indicates that for each unit increase in problems managed the odds of the encounter being bulk-billed do not change.

Robert W Pegram · Lisa Valenti

Susan D Huckson Project Manager, National Institute of Clinical Studies, 499 St Kilda Rd, Melbourne, VIC 3004. shucksonATnicsl.com.au To the Editor: Stacey, in his recent editorial on preventing pressure ulcers, stated that it is a major challenge to successfully implement guidelines in clinical practice and that a combination of strategies is required.1 In this context, your readers may be interested in the new pressure ulcer resource guide, developed by the National Institute of Clinical Studies (NICS).2 The resource guide provides health professionals and consumers with easy access to: the latest guidelines; literature reviews; health professional and consumer information resources; current Australian research activities, journal articles and papers; and links to a range of groups undertaking significant work in this area. The guide was developed in consultation with leading experts and professional groups with a strong interest in the management of pressure ulcers in Australia. This guide is an example of the efforts of NICS to help healthcare organisations and practitioners improve patient care by providing them with access to the best available evidence, proven strategies, tools and resources. The guide can be freely accessed from the NICS website www.nicsl.com.au (under “Quick Links”).

Susan D Huckson

Adrian B Wenban Associate Governor, Australian Spinal Research Foundation, PO Box 1047, Springwood, QLD 4127. adrianwenbanATwanadoo.es To the Editor: Bogduk, in his recent clinical update on management of chronic low back pain, stated that “zygapophysial joint pain can be relieved by radiofrequency neurotomy (E2, E3), and techniques are emerging for treating sacroiliac joint pain and internal disc disruption (E2, E3, E4)”.1 (E2, E3 and E4 refer to the levels of evidence: E2 evidence is obtained from at least one properly designed randomised controlled trial; E3 from pseudorandomised controlled trials or comparative studies; and E4 from case series, either post-test or pre-test and post-test.) I have two concerns about this statement. Both involve the fact that at least three systematic reviews (E1 level of evidence) have now been published on treatment of lumbar zygapophysial joint-mediated low back pain with radiofrequency procedures.2-4 Together, these three reviews cast considerable doubt on Bogduk’s claim about the role of radiofrequency neurotomy. My first concern is that the earliest of those systematic reviews,2 which came to a very different conclusion from that of Bogduk, was published in 2001, well before he submitted his clinical update to the Journal on 16 July 2003. Yet, he does not mention that review, even though he was obviously aware of it, as evidenced by his letter to the journal Regional Anesthesia and Pain Medicine expressing concern with the methods underpinning the review.5 Even if Bogduk disagrees with the conclusions of that review, justification for excluding it from the evidence base of his recent clinical update seems warranted. Secondly, two more recent systematic reviews3,4 have come to conclusions very different from those of Bogduk on the role of radiofrequency neurotomy in treatment of lumbar zygapophysial joint pain. The first of those two reviews, from the Cochrane Collaboration Back Review Group, concludes: “There is . . . conflicting evidence for its [radiofrequency denervation] effectiveness for lumbar zygapophysial joint pain. There is limited evidence suggesting that intradiscal radiofrequency may not be effective in relieving discogenic low back pain. Further high-quality randomized controlled trials are needed, with larger patient samples and data on long-term effects, for which current evidence is inconclusive.”3 The second review concludes: “Current studies fail to give more than sparse evidence to support the use of interventional techniques [zygapophysial injections and radiofrequency denervation] in the treatment of lumbar zygapophysial joint-mediated low back pain. This review emphasizes the need for larger, prospective, randomized controlled trials with uniform inclusion and exclusion criteria, standardized treatment, uniform outcome measures and an adequate duration of follow-up period so that definitive recommendations for the treatment of lumbar zygapophysial joint-mediated pain can be made.”4 In summary, authors who knowingly exclude important relevant evidence from clinical updates should at a minimum justify the exclusion of that evidence. Given the collective weight of recent relevant systematic reviews, considerable doubt now exists about the role of radiofrequency procedures in the treatment of lumbar zygapophysial joint pain.

Adrian B Wenban

Nikolai Bogduk Director, Department of Clinical Research, Royal Newcastle Hospital, David Maddison Building, Newcastle, NSW 2300. mgillamATmail.newcastle.edu.au In reply: Although systematic reviews have identified three studies of lumbar radiofrequency neurotomy, they do not take into account technical errors in the procedure. Neither the study of Gallagher et al1 nor that of Leclaire et al2 used techniques that could coagulate the target nerves accurately, consistently, and thoroughly.3 Therefore, these studies are not a measure of the efficacy of the procedure when performed correctly and have no place in a systematic review. Nor are systematic reviews permitted to include complementary, observational studies like that of Dreyfuss et al.4 Yet, this study sets the benchmark for what outcomes can be achieved, if and when correct techniques are used. The “conflicting” evidence reported by systematic reviews arises because inaccurate surgical techniques are used. When studies using inaccurate techniques are eliminated, there is no conflict. The literature reduces to one controlled study5 and one complementary study.4 These were the studies that I cited in the clinical update on chronic low back pain.6 In the face of that literature, the conclusion that I proffered, and which Wenban has quoted, is valid. Zygapophysial joint pain can be relieved by radiofrequency neurotomy. Perhaps the extra words that are required are provided that correct techniques are used.

Nikolai Bogduk

Winston Chong Chair, Interventional Radiology Reference Group, Level 9, 51 Druitt Street, Sydney, NSW 2000. ranzcrATranzcr.edu.au To the Editor: I refer to the recent editorial on “Metformin and serious adverse effects”.1 I would like to highlight the position of the Royal Australian and New Zealand College of Radiologists (RANZCR) on the use of metformin hydrochloride when administering intravascular contrast media. The RANZCR has adopted an evidence-based approach in formulating its guidelines. The College guidelines on metformin hydrochloride and intravascular contrast media are available at <www.ranzcr.edu.au/open/policies/diagnostic_imaging/pol1_2.htm>. The current guideline is that there is no need for patients to stop taking metformin hydrochloride for 24–48 hours before administration of an intravascular contrast medium. Stopping or continuing to take metformin depends on the patient’s renal status, and the likelihood of inducing renal dysfunction when intravascular contrast is administered. If discontinuation is required, then the drug only needs to be stopped for 48 hours, commencing on the day of administration of intravascular contrast.

Winston Chong

Janelle C Nisbet,* Joanna M Sturtevant,† Johannes B Prins‡ * Endocrinology Registrar, † Renal Specialist Pharmacist, ‡ Director of Diabetes and Endocrinology; and Professor of Endocrinology, University of Queensland, Princess Alexandra Hospital, Ipswich Road, Woolloongabba, QLD 4102. jprinsATsoms.uq.edu.au In reply: Chong’s letter highlights some important issues, and his points are well made. As we imply in our editorial,1 the evidence base on which to base guidelines and decisions is poor, which is one of the reasons that the guidelines differ widely between countries and organisations. Specialty-specific guidelines must also take into account practicalities. We elected to follow the more conservative end of the guideline spectrum in our suggestions, accepting that, in many circumstances, these would be difficult or impossible to follow. From a radiological perspective, it would require a significant change in practice to implement guidelines such as those we suggested, and the evidence base supporting such a change does not exist. Metformin is a short-acting drug and stopping it at the time of a potentially hazardous procedure will almost always be effective in preventing drug-related complications. Overall, the aim of our article was to raise awareness of the potential hazards of metformin use, and to encourage practitioners to follow available and relevant guidelines.

Janelle C Nisbet · Joanna M Sturtevant · Johannes B Prins

To the Editor: We report a case in which routine use of transesophageal echocardiography (TOE) during cardiac surgery almost certainly prevented a patient’s death. This is important, as the Australian Government Department of Health and Ageing has recently decided that TOE during routine coronary artery cardiac surgery should not attract a Medicare benefit as there is no Level 1 or 2 evidence of its efficacy.1 A 65-year-old man with critical aortic valve stenosis and severe left ventricular dysfunction, requiring an intra-aortic balloon pump, was scheduled for mechanical aortic valve replacement. He had been in atrial fibrillation intermittently, but was in sinus rhythm for the 24 hours before surgery and his heparin had been at therapeutic levels since balloon pump insertion 36 hours before. Like most cardiac anaesthetists in Australia, we routinely perform TOE during cardiac surgery. Following induction of anaesthesia, TOE examination confirmed severe aortic stenosis and poor left ventricular function, but the left atrial appendage was not specifically examined and a routine preoperative transthoracic echocardiogram had shown no other abnormalities. With the patient on cardiopulmonary bypass, the diseased valve was replaced. Before weaning the patient from bypass, a TOE examination showed a large free 2.5 × 2 cm mass in the left atrium resembling thrombus (Box [a]). The surgeon then reported having invaginated the left atrial appendage while de-airing. While the patient was still on bypass, a large organised clot was removed from the left atrium (Box [b]) — this had presumably developed sometime during the preoperative period. Subsequent separation of the patient from bypass was uneventful, and he made a good recovery. The thrombus would not have been suspected or found without the TOE, and almost certainly would have migrated from the left atrium into the left ventricle following the return of cardiac output, and would likely have precipitated a sudden cardiac arrest with acute valvular obstruction. Despite the lack of Level 1 or 2 evidence, most cardiac anaesthetists and surgeons consider TOE an invaluable diagnostic and monitoring tool, particularly for assessment of left ventricular function and filling. While TOE has not rendered the Swan–Ganz catheter obsolete, it provides similar information with fewer complications. This case is a dramatic illustration of the usefulness of TOE during cardiac surgery, but less dramatic examples occur much more frequently. The Department of Health and Ageing decision not to support TOE (except in valve repair or replacement) seems shortsighted. While we support the Department’s evidence-based approach, it would be interesting to know how many procedures listed in the Medicare Benefits Schedule would withstand the same scrutiny. Randomised trials are difficult to perform, but, at the very least, TOE has excellent peer consensus, strong anecdotal evidence and large series reviews2,3 supporting its routine use in cardiac surgery. Atrial thrombus in left atrium detected by transoesophageal echocardiography during cardiac surgery (a) Echocardiogram showing the mass. (b) The clot that was removed.

Chris J Cokis · John Faris

Sebastiaan J M van Hal,* Jeffrey J Post† * Advanced Trainee, † Infectious Diseases Physician, Prince of Wales Hospital, Barker Street, Randwick, NSW 2031. j.postATunsw.edu.au To the Editor: Community-acquired methicillin-resistant Staphylococcus aureus (CAMRSA) infection is an increasingly recognised cause of skin and soft tissue infection, as well as more serious manifestations, including necrotising pneumonia, endocarditis, osteomyelitis and severe sepsis.1,2 We report another serious manifestation of CAMRSA infection. A 24-year-old Australian-born white, HIV-negative, male prison inmate was transferred to hospital. He had previously injected drugs, but not for more than 2 years. He reported a 10-day history of cervical neck pain, left-sided pleuritic chest pain and a 2-cm occipital carbuncle, which developed after a laceration on the scalp from a close haircut with barber’s shears. Examination revealed a temperature of 37.9°C, sinus tachycardia, profound neck stiffness and signs of left basal consolidation with a left pleural effusion. There were no focal neurological signs or other signs of meningism. Investigations revealed a neutrophil leukocytosis of 22.7 × 109/L (reference range, 1.7–7.0 × 109/L) with mild elevation of hepatic transaminases and hypoalbuminaemia. CAMRSA was isolated from blood cultures and a swab of the occipital carbuncle. The organism was susceptible to erythromycin, vancomycin, rifampicin and fusidic acid. A chest radiograph showed extensive left-sided consolidation consistent with pneumonia. A magnetic resonance imaging scan of the entire spine revealed asymmetric thickening and enhancement of the epidural tissue from C2 to C5, with no discrete abscess, osteomyelitis or discitis. The patient made a full recovery after 4 weeks of intravenous therapy with vancomycin (1 g every 12 hours) and clindamycin (600 mg every 6 hours). This is the first reported case of epidural tissue infection (epiduritis) caused by CAMRSA, and adds to the spectrum of potential clinical manifestations. In addition, this is the first reported case of CAMRSA infection occurring in an Australian prison inmate. There have been several reported outbreaks of CAMRSA in US correctional facilities.3,4 Although no similar outbreaks have been described in Australia, clinicians should be aware that correctional facilities may contribute to the spread of CAMRSA. The most common manifestations of CAMRSA infection in the prison setting in the United States are skin and soft tissue infections, with invasive infection occurring in 1.7% of patients.4 Clinicians providing care to inmates should be aware that strategies have been identified to reduce the incidence of CAMRSA infection.5 Clinicians need to be aware of the increasing incidence of CAMRSA infection, the diverse and potentially severe manifestations, and the treatment and preventive strategies available.

Sebastiaan J M van Hal · Jeffrey J Post

Sandy Reid Head, School of Rural Health, University of New South Wales, PO Box 5695, Wagga Wagga, NSW 2560. s.reidATunsw.edu.au To the Editor: The editorial by Wearne and Wakerman on the subject of rural training and its relationship to rural practice is timely.1 There is widespread belief that the rural problem lies in small country centres, and it is generally not realised that most of the regional centres, in New South Wales at least, are seriously short of both general practitioners and specialists. The statistics are highly unreliable, as many people, such as me, are listed as GPs even though we do not practise. It seems likely that improved selection processes and the undergraduate rural training initiatives will increase student interest in rural practice. Many students express intent to practice in a rural or regional centre, but are later deflected from this by the necessity for training after graduation. Several large centres are not primary allocation centres, so a student who trained there may never get back. (The Australian Medical Students’ Association study puts the figure for rural internships as 9% of the total.)2 All too often, regional centres are understaffed, and junior staff lack adequate supervision and educational possibilities. Their negative experience is highly visible to students. The allocation of registrars and senior resident medical officers, who supervise or make time for consultants to supervise, rests between the major hospitals and the various colleges. Regional hospitals are often low on their priority list when shortages occur. Yet there are regional hospitals that rate very highly in the quality of the training experience they provide. Regional hospitals that do not enjoy this reputation need to take active steps to improve their training and supervision or their plight will worsen. Training packages also need to be developed for those who express an interest in rural practice so they can plan their future. This must be a collaborative process between the different seconding agencies: without such a process, much of the impetus of the undergraduate initiatives will be lost.

Sandy Reid

Barney J McCusker Orthopaedic Surgeon, Wavell House, PO Box 1998, Mount Gambier, SA 5290. To the Editor: I recently had the opportunity to analyse the home postcodes of medical students at the University of Adelaide over five recent consecutive years. Although this analysis is rather simplistic, its findings are quite startling (see Box). Between 50 and 60 students per year were from the Adelaide metropolitan area. My colleagues in the Department of Human Services tell me that we struggle every year to fill the junior posts at our metropolitan public hospitals. A considerable number of students (25–41) were from outside South Australia. It is only natural that these students would want to go home after completing their degree, and I believe that considerable effort is made every year to encourage these students to stay and staff junior positions in our hospitals. There has been much discussion about students from outside Australia and I will not add to the debate on this complex issue. The most alarming feature is the small number of students from rural South Australia. This ranged from zero to three or four. Moreover, I have included in this category students whose home address was in the Adelaide Hills, which is between 5 and 20 km from the Adelaide metropolitan area. We spend a lot of time encouraging students who grew up and were educated in cities to come to rural settings, using various schemes of bonding and financial inducement, with limited success. After 18 years’ practice in a rural area, I am firmly of the belief that we should be starting much earlier. If a person is born and reared in the country, and has their secondary schooling in the country, it is much easier to transplant that person back to the country after tertiary eduction. If these figures of students’ home residence are representative, we will always be fighting an uphill battle to attract doctors to rural areas. This is not the fault of our universities, as they are constrained by federal regulations. I believe the solution lies in Canberra. In our parliaments, we have proportional representation. As far as medical students are concerned, I think South Australia has disproportionate representation. Medical students attending the University of Adelaide, by area of residence (determined by home postcode) (years not specified) Total number of students 145 139 131 110 107 Students from metropolitan Adelaide 56 50 60 50 53 Students from outside South Australia 32 41 36 28 25 Students from outside Australia 53 45 35 28 26 Students from rural South Australia 4 3 0 4 3

Barney J McCusker

S Bruce Dowton,* Danielle Brown† * Chair, † Executive Officer, Committee of Deans of Australian Medical Schools, Faculty of Medicine, University of New South Wales, Sydney, NSW 2052. b.dowtonATunsw.edu.au Comment: Data collected by the Committee of Deans of Australian Medical Schools (CDAMS) would appear to broadly support the pattern of medical student origins reported by McCusker from the University of Adelaide. CDAMS has been collecting medical student statistics annually for the past few years. Box 1 sets out the basic enrolment figures for Australian and international students across all years in 2003. In 2003, CDAMS also commenced collecting data on interstate mobility (ie, students attending medical school in another state) among first-year medical students at the 12 medical schools. As a general principle, some level of interstate mobility brings positive benefits by fostering diversity in the medical student body and broadening educational and social opportunities for individual students. However, as McCusker rightly points out, this can create problems when interstate medical graduates choose to return to their home states (or move to other states) when they enter their intern training years. In this regard, South Australia faces particular problems, as it has the highest proportion of students entering medicine from interstate. Box 2 shows that just under half of all commencing medical students at Adelaide and Flinders medical schools in 2003 were from other states. This may well have an impact when these students come to choose their intern training locations. Box 3 sets out state averages for first-year students with a rural background. The figure for South Australia as a state suggests that it is reasonably comparable with other states. However, there is a sizeable gap between the rural student cohorts at Flinders and Adelaide medical schools: in 2003, 26% compared with 12%, respectively, were from a rural background. All medical schools in Australia have committed their energies to a range of collaborative programs, together with the Australian Department of Health and Ageing, aimed at ensuring a sustainable medical workforce that is equitably distributed across the nation. 1: Australian and international medical students at Australian medical schools, 2003 No. of students Percentage of total (n = 8684) Australian students, all years 7111 82% International students, all years 1573 18% 2: Interstate mobility (number of students attending medical school in another state — first-year Australian students, 2003) State No. of students Total (%) SA 85 174 (49%) Tas 25 64 (39%) Vic 82 344 (24%) Qld 62 346 (18%) NSW 63 446 (14%) WA 14 140 (10%) National total 331 1514 (22%) 3: Rural background (first-year Australian students, 2003) State No. of students Total (%) SA 31 174 (18%) Tas 14 64 (22%) Vic 80 344 (23%) Qld 93 346 (27%) NSW 71 446 (16%) WA 20 140 (14%) National total 309 1514 (20.5%)

S Bruce Dowton · Danielle Brown

To the Editor: We read with interest the article by Mitchell et al and the editorial by Cole on comparatively adverse outcomes in adolescents and young adults with cancer.1,2 The improvements in cancer survival, most remarkably demonstrated in children, have failed to similarly benefit older adolescents and young adults, a population with a higher and increasing incidence of cancer.3 Most authors advocate increasing research, networking resources and information, enhancing support for clinical trials and facilitating participation in them. Recently announced funding for a comprehensive cancer-care program for adolescents and young adults at the Peter MacCallum Cancer Institute in Melbourne is a welcome step towards these goals. The specific emotional and psychosocial needs of this age group are also poorly addressed within traditional models of care.4 Patients in this group express strong preference for peer support, opportunities to be cohorted with each other and access to specific support services.4 The short and long term sequelae of cancer, as well as of its treatment, in adolescents and young adults create particular challenges for both healthcare professionals and the broader community. In December 2003, an entire issue of the European Journal of Cancer was dedicated to adolescent oncology — this science, clinical care and the needs of the patient population.5 An important development in the United Kingdom has been the establishment of eight “teenage cancer units” with a comprehensive approach to all aspects of care.6 Preliminary research in Australia and New Zealand by one of our group (J E) highlights gaps in service and support perceived by consumers, concerns yet to be addressed by appropriate policy and funding.7 The paediatric model of care has provided not only excellent survival but also an exemplary family-centred and comprehensive support system. The recent emphasis on the transition process for adolescents with a variety of chronic illnesses has highlighted systemic differences between the paediatric and adult models. Some paediatric oncologists, especially in the United States, have proposed raising the upper age limit for eligibility to childhood cancer units as a solution. A more realistic approach, particularly in Australia, requires establishing specialist facilities operated in partnership between (preferably collocated) adult and paediatric units, with subspecialisation of the multidisciplinary workforce, an age-appropriate environment and peer support from groups like the Australian Organisation for Young People Living with Cancer (CanTeen). This approach is capable of addressing all of the above issues, from better science through to psychosocial expertise appropriately targeted for age. The ideal number and distribution of such centres/partnerships should be guided by the paediatric track record for balancing centralisation and quality with access and family focus.

Les White · Jane Ewing · Anne M Senner · Madeleine King · Belinda Goodenough