Research

Objectives: To investigate the perceptions of Australian health and medical researchers 4 years after the Wills Report recommended and led to a substantial increase in health and medical research funding in Australia.Design, setting and participants: A telephone poll of 501 active health and medical researchers, conducted between 28 April and 5 May, 2003.Main outcome measures: Researchers’ views on the adequacy of funding, infrastructure and support, salary, community recognition, the excitement of discovery and research outcomes such as publication and patenting in research.Results: Research funding was the most important concern: 91% of researchers (455/498) viewed funding as “very” or “extremely” important to their role, but only 10% (52/500) were “very” or “extremely” satisfied with the level of funding. Research infrastructure and support were seen as “very” or “extremely” important by 90% of researchers (449/501), while only 21% (104/501) were “very” or “extremely” satisfied. Researchers in medical research institutes were significantly more likely to be satisfied (27% [56/205] “very” or “extremely” satisfied) with the level of infrastructure and support than those working in universities (15% [41/268] “very” or “extremely” satisfied; P = 0.001). Among the factors that motivate researchers, the excitement of discovery stood out in terms of both high importance and satisfaction. Publications were viewed as more important research outcomes than patenting or commercial ventures.Conclusions: Funding and infrastructure support remain overwhelmingly researchers’ greatest concerns. University-based researchers were less satisfied with infrastructure and support than those in independent medical research institutes.

Louise G Shewan BA(Hons), PhD · Andrew J S Coats DM, DSc, FRACP · Jane A Glatz BSc(Hons), PhD, MBA · Christine C Bennett FRACP, MPaed

The Helicobacter story illustrates some of the human hallmarks of revolutionary research Not so long ago, peptic ulcer disease played havoc with people’s lives. Its sufferers endured chronic and debilitating pain and ran the risk of a life-threatening gastrointestinal haemorrhage or ulcer perforation. Throughout most of the 20th century, the conventional wisdom was that peptic ulcer disease was caused by gastric juice corroding vulnerable mucosa; the dictum “no acid — no ulcer” ruled the day, and neutralisation of gastric acid was the mainstay of management (Box 1). Fixed in this belief, gastroenterologists and surgeons vigorously argued the relative merits of different medications and surgical procedures in reducing gastric secretion, while research addressed the influence of “associated” factors such as social status, smoking and stress on gastric secretion and mucosal resistance. Marshall and Warren, 1984. Reproduced from Helicobacter pioneers.3 By the turn of the century, all this was relegated to medical history by the groundbreaking research of two Australians, Barry J Marshall and J Robin Warren. They scuttled the prevailing acid-mucosal model by showing that peptic ulcer disease is an infectious disease caused by Helicobacter pylori. For this research, Marshall and Warren who, “with tenacity and a prepared mind, challenged prevailing dogmas”,1 were awarded the 2005 Nobel Prize for Physiology or Medicine. The story of their research journey is well known. But its essentials are worth repeating because they illustrate some of the human hallmarks of revolutionary research. These include: being at the right place at the right time, and seeing what other people had seen but thinking what nobody else thought;2 the role of serendipity; a passion for research that abandons personal safety with self-experimentation; and the inevitable resistance of the medical establishment as research undermines current dogma. Being, seeing and thinkingAs part of the physician training program in the 1980s at Royal Perth Hospital in Western Australia, registrars were encouraged to pursue a research project. In 1981, Marshall found himself in such a position as he commenced a rotation in the hospital’s gastroentero-logy service. His boss, Dr Tom Waters, suggested that he talk to Dr Robin Warren, a pathologist at the hospital, about the mucosal spiral bacteria Warren had observed microscopically in some gastric biopsies. Marshall was intrigued and, on reviewing the endoscopic and clinical details of 25 patients exhibiting these mysterious bacteria, found that they were associated with endoscopic diagnoses of duodenal ulcer (n = 2), gastric ulcer (n = 7), gastritis (n = 12), and erosions and scars (n = 4). But there was no consistent clinical pattern. A literature review revealed that these mucosal organisms had been noted intermittently for at least a hundred years, but there were no clear patholo-gical or clinical patterns.3,4 Marshall and Warren decided to explore the significance of the spiral bacteria by gathering more clinical material and by attempting to culture the organism from gastric mucosa obtained at endoscopy. Marshall and Warren were at the right place: members of the Royal Perth Hospital gastroenterology service were prepared to provide such specimens and members of the hospital’s microbiology department were willing to freely provide people and resources for the culture quest. It also was the right time. Apart from the happy meeting of an enthusiastic registrar unencumbered by dogma and a more senior person who had an observation waiting to be explored, flexible endoscopy had become a widely used clinical procedure and provided the means to obtain fresh specimens for research. The presence of mucosal gram-negative bacteria (initially thought to be a Campylobacter-like organism3) was soon confirmed, but all attempts at their culture were unsuccessful. On taking stock, in late 1981, Marshall discussed his progress with two senior members of the Royal Perth Hospital gastroenterology service, Tom Waters and Chris Sanderson. The latter, in true Australian (laconic) style, advised: “Barry, you should stop buggerising around and do a proper study!”3 And so, a prospective study of 100 patients, with the investigators blinded, was born, and the next characteristic of revolutionary research satisfied: thinking and addressing what other people presumably had not thought about. Marshall believes his relative inexperience helped him think in an original way about the bugs in the biopsies. “If the dogma is incorrect, it’s better to know nothing about it.” (Marshall, personal communication, October 2005.) The 100-patient study aimed to pursue the following questions: is the organism present in the normal stomach? can it be cultured? can its presence be correlated with the type and severity of abnormal gastric histology? how is it related to patients’ symptoms and disease status? Marshall and Warren may not have known it, but their journey to Stockholm had begun. Serendipity plays its partThe attempt to culture the elusive organism proved fruitless until religion entered the saga. The gastric mucosal specimens for culture were treated as routine faecal specimens or throat swabs: if at 48 hours no unusual organisms were seen, the culture plates were discarded. In 1982, the Easter holidays occurred from 9 April (Good Friday) to the end of 12 April (Easter Monday). The busy weekend technicians were preoccupied with an outbreak of methicillin-resistant Staphylococcus aureus in the hospital and presumably did not get around to examining the culture plates on Saturday. On the next working day, Tuesday 13 April, small transparent colonies of H. pylori were present.3 Marshall and Warren now had tangible evidence to advance their research. The first solid dataThe 100-patient study was completed by May 1982 and its outcomes were seminal.3,4 Of the 100 patients who had undergone endoscopy, 65 had gastritis, and there was a strong association between gastritis and the presence of the spiral organism. The latter were found in all patients with duodenal ulcer and 80% of patients with gastric ulcer. In contrast, their presence was rare in patients with non-steroidal drug-related ulcers. The bacteria could be cultured, and were a new genus with features of both Campylobacter and Vibrio species. (They were subsequently shown to be H. pylori, but that is another story.) Resistance from the medical establishmentIn January 1983, to stake their individual legitimacy in the evolving H. pylori story, Warren and Marshall submitted two separate research letters outlining their preliminary data to The Lancet. Publication was delayed, as Robin Fox, an editor at The Lancet, wanted to know why there were two separate letters. These were eventually published in June of that year.5 Locally, an abstract outlining preliminary findings of the 100-patient study was not accepted by the Gastroenterological Society of Australia for presentation at its annual conference (Box 2), and difficulties attended the submission of the definitive paper of the 100-patient study to The Lancet in January 1984. Once again, the journal’s Editor-in-Chief, Ian Munro, was challenged — this time because he was not able to find reviewers who would agree on the importance of the paper. Munro sent Marshall a “temporising letter”, after the first round of review advising Marshall the he believed that The Lancet should publish the paper and that he was trying to find reviewers who would agree with him (Marshall, personal communication, October 2005). The paper was published in June 1984.6 An accompanying editorial noted in classical reserved style: “If the authors’ hypothesis of cause and effect should prove valid this work is very important indeed.”7 Self-experimentationAt the end of 1982, Marshall had left Royal Perth Hospital and taken up a senior registrar post in general medicine and gastroenterology at the Fremantle Hospital in Western Australia. Again, he was at the right place with the right people. The staff at the hospital were aware of Marshall’s research at Royal Perth Hospital and encouraged him to continue at Fremantle. These included Ian Hislop, head of gastroenterology, David McGechie, a microbiologist with excellent laboratory facilities, and the pathologist, Ross Glancy. At Fremantle Hospital, Marshall and his colleagues showed that bismuth salts (which had been used to treat gastritis and peptic ulcer disease for many years) killed H. pylori in vitro; and, in clinical studies, that bismuth cleared H. pylori but the infection would recur unless metronidazole was added to the regimen. Yet during this time, Marshall was frustrated with The Lancet’s seeming procrastinations and his own failure to develop an animal model for the disease. Possibly because of these frustrations, he decided to infect himself with H. pylori. He asked Hislop to perform a gastric biopsy on him and then ingested a pure culture of H. pylori (109 organisms). All was well for 5 days, but then he developed halitosis, morning nausea, and recurrent vomiting of acid-free gastric juice. A gastric biopsy on Day 10 showed severe acute gastritis and many H. pylori (Box 3). The symptoms spontaneously resolved after 14 days, but Marshall’s wife, Adrienne, had had enough and demanded that he immediately commence antibiotics or “be evicted from the household to sleep under a bridge”.3 The Medical Journal of Australia connectionAfter the 1984 Lancet paper came out, Marshall received a call from the then MJA editor, Alistair Brass. He congratulated Marshall on the Lancet paper, but suggested that, next time he published, it might be closer to home. In a recent interview with the Journal, Marshall noted that he had always planned to submit his initial work to The Lancet, but has no objections to publishing in Australia: “If it has a local flavour I send it to the MJA.” (Marshall, personal communication, October 2005.) And so it came to pass that the account of his self-experimentation8 and observations on the in-vitro sensitivities of H. pylori and further clinical correlates9 were submitted to this Journal and promptly published. In a citation analysis performed by the MJA to celebrate its 90 years of publication, these two articles ranked second and third among the Journal’s 10 most cited articles.10 Marshall and Warren join four other Australian born Nobel Laureates in Physiology or Medicine: Howard Florey for his discovery of penicillin, MacFarlane Burnet and Peter Doherty for immunology, and John Eccles for neurobiology. Marshall and Warren’s achievement is made more extraordinary by the fact that they performed all their ground-breaking work not in well endowed and cloistered medical research institutes, but in the orderly chaos of hospitals. And their attempts to convince medical orthodoxy were not smooth: as noted by their Nobel Prize citation, they challenged prevailing dogmas with “tenacity and a prepared mind”.1 The Helicobacter story, and the journey to this particular Nobel prize, doubtless benefited from great timing, the right people and places, original thinking, serendipity, tenacity and passion. But the outcome is unarguable: Marshall and Warren have irrevocably changed clinical practice and have alleviated much human suffering. 1 Advertisement from the MJA, 1957 2 Letter from the Gastroenterological Society of Australia* * Reproduced from Helicobacter pioneers.3 3 Silver stain of Marshall’s gastric biopsy on Day 10 after ingesting Helicobacter pylori Epithelial cells have rounded up in shape without intracellular mucin, and have many closely adherent black H. pylori organisms. Reproduced from Helicobacter pioneers.3

Martin B Van Der Weyden MD, FRACP, FRCPA · Ruth M Armstrong BMed · Ann T Gregory MB BS, GradCertPopHealth

Objective: To describe the peer review process of the Western Australian Audit of Surgical Mortality (WAASM), selected outcomes and recommendations.Study design: Prospective audit using peer review of all cases of patients who died while under the care of a Western Australian surgeon between 1 January 2002 and 30 June 2004. Audit reviews were current to 30 September 2004.Participants and setting: 194 of 202 surgeons participated after a patient died under their care.Main outcome measures: Surgeon participation, deficiencies of care, deep vein thrombosis (DVT) prophylaxis, futile surgery, postmortem reviews, proportion of operations performed by consultant surgeons.Results: The audit process was complete for 896 of 1647 reported deaths (54%), while a further 503 (31%) were still under review at 30 September 2004. Twenty deaths associated with terminal care were excluded from analysis. Median patient age was 80 years, and 799 of the 876 patients who died (91%) had significant comorbidities that increased the risk of death. Deficiencies of care were reported in 179/876 (20%). In 45/876 deaths (5%) the deficiency of care was assessed to have caused the death, and 15 deaths were considered preventable. The risk of a deficiency of care was 1.9 times higher in elective admissions than emergency admissions. Autopsy was undertaken in 83/768 (11%) deaths with complete data. Changes in practice were noted in some areas targeted by WAASM, such as improved DVT prophylaxis. A problem with fluid management was recorded.Conclusion: Most patients who died were elderly, had complex comorbidities and were treated appropriately. The WAASM has helped to change surgical practice and emphasises the importance of ongoing systematic audit. The participation of surgeons demonstrates their commitment to accountability and supports the intention of the Royal Australasian College of Surgeons to extend the process throughout Australia and New Zealand.

James B Semmens MSc, PhD · Frank M Sanfilippo BPharm, PGradDipPharm, PhD · S Aqif Mukhtar MSc, GradDipPH, MEMIM · Natasha S Haynes BA(Hons) · Jenny A Mountain BA, MBA · R James Aitken FCS(SA), FRACS

Objective: To assess the effect of a multifaceted hand hygiene culture-change program on health care worker behaviour, and to reduce the burden of nosocomial methicillin-resistant Staphylococcus aureus (MRSA) infections.Design and setting: Timetabled introduction of interventions (alcohol/chlorhexidine hand hygiene solution [ACHRS], improved cleaning of shared ward equipment, targeted patient decolonisation, comprehensive “culture change” package) to five clinical areas of a large university teaching hospital that had high levels of MRSA.Main outcome measures: Health care worker hand hygiene compliance; volume of ACHRS used; prevalence of patient and health care worker MRSA colonisation; environmental MRSA contamination; rates of clinical MRSA infection; and rates of laboratory detection of ESBL-producing Escherichia coli and Klebsiella spp.Results: In study wards, health care worker hand hygiene compliance improved from a pre-intervention mean of 21% (95% CI, 20.3%–22.9%) to 42% (95% CI, 40.2%–43.8%) 12 months post-intervention (P < 0.001). ACHRS use increased from 5.7 to 28.6 L/1000 bed-days. No change was observed in patient MRSA colonisation or environmental colonisation/contamination, and, except in the intensive care unit, colonisation of health care workers was unchanged. Thirty-six months post-intervention, there had been significant reductions in hospital-wide rates of total clinical MRSA isolates (40% reduction; P < 0.001), patient-episodes of MRSA bacteraemia (57% reduction; P = 0.01), and clinical isolates of ESBL-producing E. coli and Klebsiella spp (90% reduction; P < 0.001).Conclusions: Introduction of ACHRS and a detailed culture-change program was effective in improving hand hygiene compliance and reducing nosocomial MRSA infections, despite high-level MRSA endemicity.

Paul D R Johnson MB BS, PhD, FRACP · Rhea Martin RN, MPH, CSIC · Laurelle J Burrell RN, BAppSc · Elizabeth A Grabsch BSc, MPH, GradDipClinEpi, GradDipHAdmin · Susan W Kirsa BPharm, GradDipHospPharm · Jason O’Keeffe RN, BN, CSIC · Barrie C Mayall FRACP, FRCPA, BSc, DipEpid · Deidre Edmonds RN, CSIC · Wendy Barr RN · Christopher Bolger RN · Humsha Naidoo MB BChB, FRACMA · M Lindsay Grayson MD, MSc, FRACP, FAFPHM

Objective: To estimate the risks of maternal and perinatal morbidity and mortality in a second pregnancy, attributable to caesarean section in a first pregnancy.Design and setting: Cross-sectional analytic study of hospital births in New South Wales, based on linked population databases.Participants: 136 101 women with one previous birth who gave birth to a singleton infant in NSW in 1998–2002.Main outcome measures: Crude and adjusted odds ratios (aOR) and 95% confidence intervals (95% CI) for maternal and perinatal morbidity and mortality.Results: 19% of mothers had a caesarean section in their first pregnancy. Compared with mothers who had had primary vaginal births, mothers who had had primary caesarean section and undewent labour in the second birth were at increased risk of uterine rupture (aOR, 12.3; 95% CI, 5.0–30.1; P < 0.0001), hysterectomy (3.5; 1.5–8.4; P < 0.01), postpartum haemorrhage (PPH) following vaginal delivery (1.6; 1.4–1.7; P < 0.0001), manual removal of placenta (1.3; 1.1–1.6; P < 0.01), infection (6.2; 4.7–8.2; P < 0.0001) and intensive care unit (ICU) admission (3.1; 2.1–4.7; P < 0.0001); among mothers who did not undergo labour (ie, had an elective caesarean section), there was a lower risk of PPH (0.6; 0.5–0.7; P < 0.0001) and ICU admission (0.4; 0.3–0.5; P < 0.0001). For infants there was increased risk of preterm delivery (1.2; 1.1–1.3; P < 0.0001) and neonatal intensive care unit admission following labour (1.6; 1.4–1.9; P < 0.0001) in the birth after primary caesarean section. The occurrence of stillbirth was not modified by labour.Conclusions: Caesarean section in a first pregnancy confers additional risks on the second pregnancy, primarily associated with labour. These should be considered at the time caesarean section in the first pregnancy is being considered, particularly for elective caesarean section for non-medical reasons.

Lee K Taylor DRANZCOG, FAFPHM · Judy M Simpson PhD, CStat · Christine L Roberts DrPH, FAFPHM · Emily C Olive FRANZCOG, DDU · David J Henderson-Smart FRACP, PhD

Objective: To describe empiric community-acquired pneumonia (CAP) management in Australian hospital emergency departments (EDs) and evaluate this against national guidelines, including use of the pneumonia severity index and antibiotic selection.Design: A multicentre, cross-sectional, retrospective audit, April 2003 to February 2005.Setting: 37 Australian hospitals: 22 principal referral hospitals, six large major city hospitals, four large regional hospitals, four medium hospitals and one private hospital.Participants: Adult patients with a diagnosis of CAP made in the ED. Data on 20 consecutive CAP ED presentations were collected in participating hospitals.Outcome measures: Documented use of the pneumonia severity index, initial antibiotic therapy prescribed in the ED, average length of stay, inpatient mortality, and concordance with national guidelines.Results: 691 CAP presentations were included. Pneumonia severity index use was documented in 5% of cases. Antibiotic therapy covering common bacterial causes of CAP was prescribed in 67% of presentations, although overall concordance with national guidelines was 18%. Antibiotic prescribing was discordant due to inadequate empiric antimicrobial cover, allergy status (including contraindication to penicillin), inappropriate route of administration and/or inappropriate antibiotic choice according to recommendations. There was no significant difference between concordant and discordant antibiotic prescribing episodes in average length of stay (5.0 v 5.7 days; P = 0.22) or inpatient mortality (1.6% v 4.1%; χ2 = 1.82; P = 0.18).Conclusions: Antibiotic therapy for CAP prescribed in Australian EDs varied. Concordance with national CAP guidelines was generally low. Targeted interventions are required to improve concordance.

for the CAPTION Study Group*

Objective: To estimate the incremental effects on cost and quality of life of cardiac rehabilitation after an acute coronary syndrome.Design: Open randomised controlled trial with 1 year’s follow-up. Analysis was on an intention-to-treat basis.Setting: Two tertiary hospitals in Sydney.Intervention: 18 sessions of comprehensive exercise-based outpatient cardiac rehabilitation or conventional care as provided by the treating doctor.Participants: 113 patients aged 41–75 years who were self-caring and literate in English. Patients with uncompensated heart failure, uncontrolled arrhythmias, severe and symptomatic aortic stenosis or physical impairment were excluded.Main outcome measures: Costs (hospitalisations, medication use, outpatient visits, investigations, and personal expenses); and measures of quality of life. Incremental cost per quality-adjusted life year (QALY) saved at 1 year (this estimate combines within-study utility effects with reported 1-year risk of survival and treatment effects of rehabilitation on mortality). Sensitivity analyses around a base case estimate included alternative assumptions of no treatment effect on survival, 3 years of treatment effect on survival and variations in utility.Results: The estimated incremental cost per QALY saved for rehabilitation relative to standard care was $42 535 when modelling included the reported treatment effect on survival. This increased to $70 580 per QALY saved if treatment effect on survival was not included. The results were sensitive to variations in utility and ranged from $19 685 per QALY saved to rehabilitation not being cost-effective.Conclusions: The effects on quality of life tend to reinforce treatment advantages on survival for patients having postdischarge rehabilitation after an acute coronary syndrome. The estimated base case incremental cost per QALY saved is consistent with those historically accepted by decision making authorities such as the Pharmaceutical Benefits Advisory Committee.

Tom G Briffa PhD, MSc · Simon D Eckermann PhD(Ec), MSc, GradDipHEc · Alison D Griffiths BA Hons · Anthony C Keech MB BS, MScEpid, FRACP · Phillip J Harris MB BS, DPhil, FRACP · M Rose Heath RN · Saul B Freedman PhD, FRACP, FACC, FESC · Lana T Donaldson RN, MPH · N Kathryn Briffa BAppSc(Physio), PhD

Objectives: To ascertain what general practitioners’ priorities are for achieving optimal outcomes in people with asthma, and the barriers they face in delivering this care.Design: A qualitative study using the Nominal Group Technique (a highly structured meeting to gain information from experts about a particular issue) was conducted between August 2002 and September 2003. GPs in six discussion groups were asked “What do you think is needed to achieve best outcomes for asthma care?” To augment analysis of the discussion, sessions were taped and transcribed.Participants: Forty-nine GPs were recruited: 34 from metropolitan and 15 from rural areas.Results: All groups nominated asthma education for patients and continuing professional education for GPs as major priorities, but they also described educational and structural barriers to achieving these priorities. Other priorities were: medication adherence, facilitating regular patient review, negotiated treatment/management plans, making the correct diagnosis, increased remuneration and consultation time, and safer asthma medications and access to these. Health promotion initiatives and increased public awareness were also priorities. Spirometry was a significant cause of uncertainty. Overall, written asthma action plans were not considered a high priority.Conclusions: Remarkable consistency was found between GPs’ priorities for delivering best asthma care. Our study identified barriers to asthma guideline adherence, including accessible, relevant education for GPs, and structural, time and cost barriers GPs must overcome in providing asthma treatment and patient education.

Dianne P Goeman MA, DipSoc · Jo A Douglass MD, FRACP · Chris D Hogan MB BS, FRACGP · Rosalie A Aroni PhD · Michael J Abramson PhD, FRACP · Susan M Sawyer MD, FRACP · Kay Stewart PhD · Lena A Sanci MB BS, FRACGP, PhD

Objective: To assess the value of cusum analysis in hospital bed management.Design: Comparative analysis of medical patient flows, bed occupancy, and emergency department admission rates and access block over 2 years.Setting: Internal Medicine Services and Emergency Department in a teaching hospital.Interventions: Improvements in bed use and changes in the level of available beds.Main outcome measures: Average length of stay; percentage occupancy of available beds; number of patients waiting more than 8 hours for admission (access block); number of medical patients occupying beds in non-medical wards; and number of elective surgical admissions.Results: Cusum analysis provided a simple means of revealing important trends in patient flows that were not obvious in conventional time-series data. This prompted improvements in bed use that resulted in a decrease of 9500 occupied bed-days over a year. Unfortunately and unexpectedly, after some initial improvement, the levels of access block, medical ward congestion and elective surgical admissions all then deteriorated significantly. This was probably caused by excessive bed closures in response to the initial improvement in bed use.Conclusion: Cusum analysis is a useful technique for the early detection of significant changes in patient flows and bed use, and in determining the appropriate number of beds required for a given rate of patient flow.

Claire M Burns RN, BNursing · Cameron J Bennett MB BS, FRACP · Colin T Myers MB ChB, FACEM · Michael Ward MB BS, FRACP

Objective: To identify the relative importance of extrinsic determinants of doctors’ choice of specialty.Design: A self-administered postal questionnaire.Setting: Australian vocational training programs.Participants: 4259 Australian medical graduates registered in September 2002 with one of 16 Australian clinical colleges providing vocational training programs.Main outcome measures: Choice of specialist vocational training program; extrinsic factors influencing choice of program, and variation by sex, age, marital status and country of birth.Results: In total, 79% of respondents rated “appraisal of own skills and aptitudes” as influential in their choice of specialty followed by “intellectual content of the specialty” (75%). Extrinsic factors rated as most influential were “work culture” (72%), “flexibility of working arrangements” (56%) and “hours of work” (54%). We observed variation across training programs in the importance ascribed to factors influencing choice of specialty, and by sex, age and marital status. Factors of particular importance to women, compared with men, were “appraisal of domestic circumstances” (odds ratio [OR], 1.9), “hours of work” (OR, 1.8) and “opportunity to work flexible hours” (OR, 2.6). Partnered doctors, compared with single doctors, rated “hours of work” and “opportunity to work flexible hours” as more important (OR, 1.3), while “domestic circumstances” was more important to doctors with children than those without children (OR, 1.7). In total, 80% of doctors had chosen their specialty by the end of the third year after graduation.Conclusions: Experience with discipline-based work cultures and working conditions occurs throughout medical school and the early postgraduate years, and most doctors choose their specialty during these years. It follows that interventions to influence doctors’ choice of specialty need to target these critical years.

Mary G Harris MPH, PhD, FCHSE · Paul H Gavel BEc, MEc · Jeannette R Young MB BS, MBA, FRACMA

Objective: To determine immunity to tetanus and diphtheria in the Australian population.Design and setting: Analysis, using double antigen enzyme immunoassays, of a representative sample of sera (1950 samples tested for diphtheria and 2884 for tetanus) collected opportunistically from Australian laboratories between July 1996 and May 1999.Main outcome measure: Immunity to diphtheria and tetanus, defined as negative (susceptible) when the antitoxin level was < 0.01 IU/mL, positive (immune) when it was ≥ 0.1 IU/mL, and low positive (partially immune) when it was in the range 0.01–< 0.1 IU/mL.Results: About 99% of children aged 5–9 years had diphtheria and tetanus antitoxin levels ≥ 0.01 IU/mL (immune or partially immune). Antitoxin levels declined with age and generally more markedly for diphtheria than tetanus. For subjects aged 50 years and over, less than 60% were immune or partially immune to diphtheria and less than 75% to tetanus. Men and women had similar diphtheria antitoxin levels, while women had lower levels of tetanus antitoxin compared with men of the same age, with the difference being most marked in the age group ≥ 70 years (37% v 60%; P < 0.001).Conclusions: Immunity in children appears to be good, but adults, especially older people, may not be adequately protected. Recent changes to the Australian Standard Vaccination Schedule should improve immunity in cohorts now aged < 50 years. However, additional efforts are required to protect those over 50 years (especially travellers), who are most susceptible.

Heather F Gidding BAppSci, GradDipEpidBiostats, MAppEpid · Josephine L Backhouse BRurSc · Gwendolyn L Gilbert MD, FRACP, FRCPA · Margaret A Burgess MD, FRACP

Objective: To determine whether community management of mild to moderate community-acquired pneumonia (CAP) is as effective and acceptable as standard hospital management of CAP.Design: Randomised controlled trial.Setting: Christchurch, New Zealand, primary and secondary care.Participants: 55 patients presenting or referred to the emergency department at Christchurch Hospital with mild to moderately severe pneumonia, assessed using a validated pneumonia severity assessment score, from July 2002 to October 2003.Interventions: Hospital treatment as usual or comprehensive care in the home delivered by primary care teams.Main outcome measures: Primary: days to discharge, days on intravenous (IV) antibiotics, patient-rated symptom scores. Secondary: health status measured using level of functioning at 2 and 6 weeks, patient satisfaction.Results: The median number of days to discharge was higher in the home care group (4 days; range, 1–14) than in the hospital groups (2 days; range, 0–10; P = 0.004). There was no difference in the number of days on IV antibiotics or on subsequent oral antibiotics. Patient-rated symptom scores at 2 and 6 weeks, median change in symptom severity from baseline to 6 weeks, and general functioning at 2 and 6 weeks did not differ between the groups. Patients in both groups were satisfied with their treatment, with a clear preference for community treatment (P < 0.001).Conclusions: Mild to moderately severe CAP can be managed effectively in the community by primary care teams. This model of comprehensive care at home can be implemented by primary care teams with suitable funding structures.

Dee A Richards MB ChB, DPH(Dist) · Les J Toop MD, MRCGP, FRNZCGP · Michael J Epton MRCP, FRACP, PhD · G Ian Town MB ChB, FRACP, DM · Robin D Dawson MB ChB, MRCGP, DCH · Michael C Hlavac MB ChB, FRACP · Graham R B McGeoch MB ChB, FRNZCGP · Simon M H Wynn-Thomas BM BS, MRCGP, FRNZCGP · Paul D Abernethy BA(Psych), RN · Anja M Werno MD, FRCPA

Aim: To describe the characteristics, outcomes and treatment complications of patients with pulmonary embolism (PE) who were treated at home and as outpatients in an ambulatory care program.Methods: Retrospective descriptive study of patients with PE who were treated in the ambulatory care unit during 2003. Ambulatory care unit data and medical record information were reviewed. Data collected included demographic and clinical data, standard clinical indicators of unplanned admission during treatment program, incidence of major bleeding, recurrent venous thromboembolism (VTE), and death within 3 months of admission into the ambulatory care program.Results: 130 patients with PE were treated: 46% were treated totally as outpatients and 54% as early discharge patients. Mean age was 66.4 years; 61% were women. The program was successfully completed for 89% of patients; one patient was lost to follow-up. There were three episodes of major bleeding (2%; 95% CI, 0.5%–7%), all in patients aged > 70 years. Four patients died (3%; 95% CI, 0.8%–8%) within 3 months of admission into the program, but none in the first week, no death being directly attributable to PE. There were seven episodes of recurrent VTE (5%; 95% CI, 2%–11%).Conclusion: Appropriately selected patients with sub-massive PE can be treated as outpatients and in the home. Although the outcome is good in most patients, a significant proportion will require admission, emphasising the need for a well defined protocol and close medical supervision. Further study will more closely define at-risk patients and refine the care pathways.

Bin S Ong MB BS, FRACP · Margaret A Karr MPH, MSc(Med) · Daniel K Y Chan MD, FRACP, MHA · Anthony Frankel MB BCh(Wits), FCP(SA), FRACP · Qing Shen BMed, MSc

Objectives: To determine the incidence of childhood type 1 diabetes mellitus (T1DM) in New South Wales from 1997 to 2002; to compare with previously published rates (1990–1996); and to analyse trends in incidence from 1990 to 2002.Design, setting and participants: Prospective population-based incidence study. Primary ascertainment of incident cases aged < 15 years was from the Australasian Paediatric Endocrine Group NSW children’s diabetes register. Secondary ascertainment was from the National Diabetes Supply Scheme until 1999 and from the Australian Institute of Health and Welfare thereafter. Childhood population data were obtained from the Australian Bureau of Statistics.Main outcome measures: Age-standardised incidence; trends in incidence by calendar year, and sex and age at diagnosis.Results: There were 3260 incident cases (1629 boys, 1631 girls) in the 13 years. Case ascertainment was 99.7% complete using the capture–recapture method. Mean age-standardised incidence per 100 000 person-years was 20.9 (95% CI, 19.9 to 21.9) from 1997 to 2002 compared with 17.8 (95% CI, 17.0 to 18.7) from 1990 to 1996; there was a plateau in incidence between 1997 and 2002. Overall, the incidence increased on average by 2.8% per year (95% CI, 1.9% to 3.8%, P < 0.001) and increased with age, being 12.2 (95% CI, 11.3 to 13.1) in 0–4 year olds; 18.9 (95% CI, 17.8 to 20.0) in 5–9 year olds and 26.7 (95% CI, 25.4 to 28.1) in 10–14 year olds. The increase per year in 0–4 year olds (3.9%) was not significantly higher than in older children. The mean incidence of T1DM was 19.8 (95% CI, 18.8 to 20.7) in girls and 18.8 (95% CI, 17.9 to 19.7) in boys (P = 0.02).Conclusions: The incidence of childhood-onset T1DM has increased significantly in all age groups in NSW since 1990. Resource planning in the management of childhood diabetes in NSW should take these findings into account.

Craig E Taplin MB BS · Maria E Craig PhD, FRACP, MMed(ClinEpid) · Margaret Lloyd RN · Martin Silink AM, MD, FRACP · Neville J Howard FRACP, FRCP · Claire Taylor MB ChB · Patricia Crock FRACP

Objectives: To describe trends in the notification of infectious syphilis in New South Wales, the characteristics of homosexually active men recently notified with early syphilis, and the seroprevalence and incidence of syphilis, as well as associated risk factors, in a Sydney cohort of HIV-negative homosexually active men.Design, setting and participants: Secondary analysis of New South Wales infectious syphilis surveillance data from 1998 to 2003; a case series of 57 homosexually active men diagnosed with early syphilis in inner Sydney from December 2002 to January 2004; and a prospective cohort study of syphilis among 1333 HIV-negative homosexually active men in Sydney recruited from June 2001 to December 2003.Main outcome measures: Rates of notification of infectious syphilis in New South Wales and in areas of inner Sydney; behavioural and clinical features of men with syphilis in the case series; and incidence of syphilis and hazard ratios (HRs) associated with sexual behaviours in the cohort study.Results: Infectious syphilis notifications in inner Sydney rose more than 10-fold (from 6 in 1999 to 162 in 2003), and the increase was confined to men. Of 57 men with early syphilis in the case series, 54% were HIV-positive and 32% reported no symptoms of syphilis. These 57 men were highly sexually active and likely to report recreational drug use. In the cohort study, 1292 men (97% of participants) consented to syphilis testing; the incidence of syphilis was 0.78 per 100 person-years, and risk factors included reporting unprotected anal intercourse with HIV-positive partners (HR, 5.31; 95% CI, 2.00–184.93) and insertive oral sex (HR, 4.55; 95% CI, 1.14–18.18).Conclusion: Syphilis has been re-established among homosexually active men in Sydney, and HIV-positive men are over-represented. Frequent screening is needed in this population to curb the transmission of both syphilis and HIV.

Fengyi Jin MPH · Garrett P Prestage PhD · John M Kaldor PhD · Andrew E Grulich PhD, FAFPHM · Susan C Kippax PhD · Catherine M Pell MB BS · Basil J Donovan MD, FAChSHM

Objective: To examine the effects of asbestos exposure and tobacco smoking on the level and rate of change of the diffusing capacity of the lung for carbon monoxide (Dlco).Design and participants: A cohort study of 934 people (including both mine workers and town residents) exposed to crocidolite (blue asbestos) at the asbestos mines and in the town of Wittenoom, Western Australia, between 1943 and 1966. Dlco measurements were taken during a follow-up period from 1992 to 2002.Main outcome measures: Baseline levels of Dlco and change in levels over time.Results: 2980 Dlco measurements were done on 934 people (of whom 818 were men and 724 were workers) who underwent a median of 2 (range, 1–17) measurements during the follow-up period. Radiographic asbestosis at baseline and asbestos exposure at a younger age were associated with lower Dlco values. The average rate of decline in Dlco was 0.33 (95% CI, 0.31–0.35) units per year, plus an additional decrement of 0.22 (95% CI, 0.12–0.32) units per year if the participant had radiographic asbestosis at the beginning of the follow-up period. Compared with never-smokers, current smokers and ex-smokers had lower Dlco at baseline, but smoking status did not affect the change in Dlco during the follow-up period.Conclusions: Our results confirm a continuous deleterious effect of crocidolite on Dlco, especially on people with asbestosis. Smoking was associated with lower Dlco levels, but was not a significant predictor of rate of change in Dlco. Smoking status did not affect the relationships between crocidolite exposure and the level or rate of change of Dlco in this population.

Helman S Alfonso PhD · Lin Fritschi PhD · Nicholas H de Klerk PhD · Nola Olsen MSc · Jan Sleith MSc · Arthur (Bill) W Musk PhD, FRACP

Objective: To compare trends and rates of HIV and sexually transmitted infections in Indigenous and non-Indigenous people of Western Australia.Design and setting: Analysis of WA notification data for chlamydia, gonorrhoea, and primary and secondary syphilis in 2002, and for HIV infections from 1983 to 2002.Main outcome measures: Rates of HIV and sexually transmitted infection by Indigenous status.Results: In 2002, there were 3046 notifications for chlamydia, 1380 for gonorrhoea and 64 for syphilis. When information on Indigenous status was available, Indigenous people accounted for 41% of chlamydia and 76% of gonorrhoea notifications, with Indigenous : non-Indigenous age-standardised rate ratios of 16 (95% CI, 14–17) and 77 (95% CI, 67–88), respectively. Indigenous people accounted for 90.6% of syphilis notifications (age-standardised Indigenous : non-Indigenous rate ratio, 242 [95% CI, 104–561]). From 1985 to 2002, HIV notification rates for non-Indigenous people in WA declined and rates for Indigenous people increased. From 1994 to 2002, there were 421 notifications of HIV infection in WA residents, 52 (12.4%) in Indigenous people and 369 (87.6%) in non-Indigenous people. Indigenous people accounted for 39% and 6.2% of all notifications in WA females and males, respectively. The Indigenous : non-Indigenous rate ratios were 18 (95% CI, 12–29) for females and 2 (95% CI, 1–3) for males.Conclusions: Indigenous Western Australians are at greater risk of HIV transmission than non-Indigenous people. Strategies to prevent further HIV infection in Indigenous Australians should include control of sexually transmitted infections.

Michael R Wright BSW, MAE(IH) · Carolien M Giele RN, BSc(Hons), MPH · Phyll R Dance BA, PhD · Sandra C Thompson FAFPHM, PhD

Objective: To reduce pethidine prescribing in hospital emergency departments (EDs).Design: Multi-centre drug use evaluation (DUE) process.Setting and participants: Emergency departments in 23 public hospitals (22 in New South Wales, 1 in Victoria) from 1 September 2002 to 31 August 2003. Participating hospitals included seven principal referral hospitals, six major non-teaching hospitals and 10 district or community hospitals. Data for comparison were collected from 12 non-participating hospitals.Interventions: Hospital coordinators at each participating hospital were provided with support to implement a range of prescribing interventions in their ED in each of three DUE cycles. Interventions included educational materials (guidelines, posters, prescribing reminders), audit and feedback, and small-group discussions. Three audits of pethidine prescribing were undertaken. Prescribing was compared with evidence-based guidelines and non-concordance identified.Main outcome measures: Number of dosage units of parenteral analgesics issued to the ED from each hospital’s pharmacy department was recorded monthly and aggregated in 3-month periods.Results: In the 12 months between the preintervention period and the equivalent post-intervention period, pethidine use decreased by 62% in project hospitals (4669 to 1793 units) and 56% in control hospitals (1476 to 648 units). Six months after project completion there was a significantly greater reduction from baseline in participating hospitals (71%; 4669 to 1348 units) compared with non-participating hospitals (64%; 1476 to 532 units; P < 0.001). There was a concurrent increase in use of both morphine and tramadol.Conclusion: There was a sustained reduction in pethidine use during the study period, which may indicate successful promotion of safer analgesic prescribing. It is not clear whether changes were a result of collaborative DUE methods or other factors.

Karen I Kaye BPharm, DHP · Susan A Welch BPharm · Sharon R Davis BPharm, DipNut · Linda V Graudins BPharm, DHP, FSHPA · Andis Graudins FACEM, PhD · Tai Rotem BSOCSCI · Richard O Day MD, FRACP

Objective: To compare the rates of adverse reactions and parental approval ratings for three different techniques for anterolateral thigh vaccination in children aged 2, 4, 6 and 18 months.Design: Randomised, observer-blind trial.Participants: 375 children who received pertussis-containing vaccines in a regional New South Wales town between 29 May 2001 and 30 June 2002.Interventions: Children were randomised to receive intramuscular injection with acellular pertussis-containing and Haemophilus influenzae type b vaccines with one of three recognised injection techniques (Australian, World Health Organization or United States).Main outcome measures: Local adverse reactions (bruising and redness/swelling), systemic adverse reactions (irritability, perceived fever, persistent crying/screaming, drowsiness, vomiting/poor feeding) and parental acceptance were assessed 24 hours after injection.Results: 361 children (96%) were evaluated 24 hours after vaccination. The WHO technique resulted in significantly fewer children, than with the other two techniques, with the systemic adverse reaction variable “irritability” (P = 0.0039). There was a significant difference between the technique groups overall for the local adverse reaction “bruising” with acellular pertussis-containing vaccines (P = 0.0418), due to a lower reaction rate in the WHO group compared with the US group (P = 0.0356).Conclusion: The WHO technique appears to be the optimal technique for anterolateral thigh injection in children — it ensures that the injection is intramuscular, results in fewer adverse reactions, and is the easiest technique to perform as it does not require angling of the needle to the long axis of the femur.

Ian F Cook MFamMed, PhD, FACRRM · John Murtagh MD, FRACGP

Aim: The Asthma 3+ Visit Plan is an initiative to promote organised asthma care in general practice. This study aimed to identify factors associated with uptake of the plan by general practitioners, and their views on barriers and facilitators to implementation of the plan.Design: Postal survey sent to a random sample of GPs.Participants and setting: 315 GPs in five Divisions of General Practice in metropolitan Sydney, surveyed sequentially between 1 October 2002 and 31 May 2003.Outcome measures: Awareness and use of the Asthma 3+ Visit Plan; GP and practice factors associated with use of the plan; and GP views on barriers and facilitators to implementing the plan.Results: The response rate was 55.7%, and 72.1% of participants were male; participants’ mean age was 50.5 years. Most GPs (91.2%) were aware of the plan and and 44.9% had used it. GP and practice factors associated with use of the plan were use of the six-step Australian Asthma Management Plan, confidence in aspects of asthma care, practice accreditation, sign-up for asthma incentives, and computerisation. Major barriers to implementing the plan were workload/paperwork and administrative complexities. Patient factors that influenced completion of the plan were their concept of the severity of their asthma, compliance with follow-up, and patient attitudes towards asthma care.Conclusion: The perceived workload and administrative complexity of the asthma incentives are barriers to uptake. Factors relating to the illness rather than social factors are seen as the most important influences on completion of the plan by patients.

Nicholas A Zwar PhD, FRACGP · Iqbal Hasan MB BS, MPH · Elizabeth J Comino PhD, MPH · Mark F Harris PhD, FRACGP

Objective: To measure the independent effect on length of general-practice consultations of a range of characteristics of the general practitioner (GP), practice, patient and consultation, as a basis for considering future GP workforce needs.Design: Secondary analysis of data from the BEACH (Bettering the Evaluation and Care of Health) study.Setting and participants: Data were obtained from 1904 GPs Australia-wide on 70 758 consultations between 1 January 2001 and 31 December 2002; all consultations that were claimable from the Australian Government’s Medicare system as General Practice Attendances and had recorded start and finish times were included.Main outcome variables: Characteristics of the GP, practice, patient and consultation that were significantly related to consultation length, determined by multiple regression analysis.Results: The following variables had an independent positive effect on consultation length: GP female, older, graduated in Australia, FRACGP-qualified, and in rural practice; patient female, older, new to practice, with higher socioeconomic status, no health concession card, more reasons for encounter, and more problems managed; and management of specific problem types (social, psychological and female genital problems), management of chronic disease, and provision of clinical treatments.Conclusion: The independent relationship of some GP, practice, patient and consultation characteristics with length of consultation may affect future GP supply. These factors should be considered in modelling future general practice workforce needs.

Helena C Britt BA, PhD · Lisa Valenti BEc · Graeme C Miller PhD, FRACGP

Objective: To assess pharmaceutical advertisements in prescribing software, their adherence to code standards, and the opinions of general practitioners regarding the advertisements.Design, setting and participants: Content analysis of advertisements displayed by Medical Director version 2.81 (Health Communication Network, Sydney, NSW) in early 2005; thematic analysis of a debate on this topic held on the General Practice Computer Group email forum (GPCG_talk) during December 2004.Outcome measures: Placement, frequency and type of advertisements; their compliance with the Medicines Australia Code of Conduct, and the views of GPs.Results: 24 clinical functions in Medical Director contained advertisements. These included 79 different advertisements for 41 prescription products marketed by 17 companies, including one generic manufacturer. 57 of 60 (95%) advertisements making a promotional claim appeared noncompliant with one or more requirements of the Code. 29 contributors, primarily GPs, posted 174 emails to GPCG_talk; there was little support for these advertisements, but some concern that the price of software would increase if they were removed.Conclusions: We suggest that pharmaceutical promotion in prescribing software should be banned, and inclusion of independent therapeutic information be mandated.

Ken J Harvey MB BS, FRCPA · Agnes I Vitry PharmD, PhD · Elizabeth Roughead BPharm, MAppSc, PhD · Rosalie Aroni PhD · Nicola Ballenden MA, MPH · Ralph Faggotter MB BS

Objectives: To examine the effect of demographic change on employment patterns for general practitioners, medical specialists and nurses since 1986, and to compare their patterns of retirement.Design and setting: Secondary analysis of previously unpublished Australian Bureau of Statistics Census data for the years 1986, 1991, 1996 and 2001.Main outcome measures: Age distribution of GPs, specialists and nursing workforce; attrition rates as GPs, specialists and nurses left the workforce; and hours worked according to age group.Results: The age profile of the GP, specialist and nursing workforce has aged since 1986 (P < 0.001), with the “baby boomer” generation making up more than half the workforce in 2001. A large proportion of GPs continued to work beyond the traditional retirement age of 65 years, with nurses retiring at a younger age than doctors (P < 0.001). All GP cohorts worked fewer hours in 2001 than they did in 1986 (P < 0.001), with “generation X” GPs working fewer hours than the baby boomers did at the same age (P < 0.001).Conclusions: Attrition of baby boomer clinicians will place unprecedented pressure on the medical workforce, and policy makers face a critical challenge to ensure workforce needs are met over the next 20 years. Policies and incentives to encourage ongoing employment among older clinicians, albeit at reduced hours, are crucial if the Australian health workforce is to be adequate to meet the growing community demand of the 21st century.

Deborah J Schofield PhD · John R Beard PhD, FAFPHM

Objective: To evaluate a hospital-based psychiatric consultation service for patients referred by general practitioners (GPs), and the effect on its use of a focused marketing strategy aimed at GPs.Design and setting: Postal survey of GPs in the catchment area (inner north Brisbane, Queensland), September to November 2003; and assessment of referrals, March to August 2003.Main outcome measures: Patient referrals, satisfaction among GPs who had referred, and awareness and opinions of the service among GPs who had not referred, compared with results of a similar survey conducted before marketing.Results: In the 6 months after marketing, 43 patients were referred by 23 GPs, an average of 7.2 patients per month, compared with 2.5 per month in the first 12 months of the service. Survey responses were received from 13 of 36 GPs who had referred patients and 97 of 282 GPs who had not (response rate, 35%). Satisfaction among GPs who had referred remained high, and 12/13 felt the service should continue. Among GPs who had not referred, 76% were aware of the service, up from 26% in the previous survey, and 99% liked the concept of the service.Conclusion: Given the ongoing low utilisation of this service, we question whether this model is accepted by most GPs in our district. Possibly, they prefer more traditional models, where treatment is taken over by psychiatrists in the public or private system. We believe there is a need to increase the capacity and scope of publicly funded services to treat mental health problems.

Alex E Simpson MRCGP, DTMH, FRANZCP · W Brett Emmerson MHA, FRANZCP, FRACMA · Aaron D J Frost BBehSc, BPsych(Hons) · Jacinta L Powell MHA, FRANZCP

Aim: To evaluate the efficacy and safety of an annual intramuscular injection of cholecalciferol for vitamin D deficiency.Design: Prospective open-label study.Participants: Five men and 45 women (mean age 66.3 years) with vitamin D deficiency who were given a single therapeutic intramuscular injection of 600 000 IU (15 mg) cholecalciferol (vitamin D3).Outcome measures: Serum levels of calcium, creatinine, 25-hydroxyvitamin D3 (25OHD3) and parathyroid hormone, as well as early morning 2-hour urine calcium/creatinine excretion index. Specimens were collected at baseline and after 4 and 12 months of therapy. Data are reported as mean ± 1 SD.Results: Vitamin D deficiency was severe (< 12.5 nmol/L) in one participant, moderate (12.5–24 nmol/L) in 14, and mild (25–49 nmol/L) in 35. Twenty-four participants (48%) had secondary hyperparathyroidism. Following intramuscular cholecalciferol injection, serum 25OHD3 levels normalised in all participants and remained above 50 nmol/L throughout the study. Serum 25OHD3 levels were significantly higher at 4 months (114 ± 35 nmol/L), and 12 months (73 ± 13 nmol/L) compared with baseline (32 ± 8 nmol/L) (P < 0.001), increasing by an average of 128% over the 12 months. There was a corresponding decrease in serum parathyroid hormone levels at 4 months (6 ± 3 pmol/L) and at 12 months (5.2 ± 3 pmol/L), with a 30% decrease at 12 months from baseline (7.4 ± 4 pmol/L) (P < 0.01). Primary hyperparathyroidism was unmasked in one participant at 4 months and mild hypercalcaemia (serum calcium, < 2.70 mmol/L) was noted in two participants (4%) at 12 months. Serum creatinine levels remained normal in all participants throughout the study, while increases in 2-hour urine calcium/creatinine excretion index were seen in 10 participants (20%) at 12 months, three of whom had had elevated values at baseline.Conclusions: Once-yearly intramuscular cholecalciferol injection (600 000 IU) is effective therapy for vitamin D deficiency. While this therapy appears to be safe, the potential for developing hypercalciuria needs to be examined in a large randomised controlled trial.

Terrence H Diamond MB ChB, MRCP, FRACP · Kenneth W Ho MB BS · Peter G Rohl MB BS, FRACP · Matthew Meerkin FRCPA, FAACB, FACB