Health services administration

To the Editor: I endorse the viewpoint expressed by Jenkins and colleagues.1 They refer to the dearth of hospital medical generalists, at a time of increasing numbers of elderly patients with multiple comorbidities. Nowhere is this lack of appropriate clinical skill to match demand so apparent as in our regional centres. More often than not, junior doctors are responsible for older patients, and are required to manage disparate inputs from several medical and/or surgical subspecialists. Particularly for surgical patients with medical comorbidities, the junior resident medical officer (generally supervised by a visiting surgeon) is an inappropriate medical “case manager”. There is a desperate need in our regional centres to train and employ hospital-based generalists. At the same time, there is a need to move away from the visiting medical officer (VMO) fee-for-service model, designed around the needs of private practitioners, and move towards a hospital-based specialist model, in which specialists are available, accessible and part of the fabric of our hospitals. While the VMO model has been useful, it is no longer appropriate as a basis for default clinical care arrangements. Only by moving towards a hospital-based generalist model, as suggested by Jenkins et al, will we see an improvement in hospital culture, junior doctor supervision and training, and, most importantly, medical governance for best patient care in our regional hospitals.

Joanna R Sutherland

To the Editor: I commend Jenkins and colleagues for an engaging article on the future of general medicine in Australia.1 It mirrors a debate that is occurring in many acute-care hospitals, particularly in the context of a nationwide shortage of acute-care beds and increasing numbers of older patients presenting with chronic and multisystem disease.2 Currently, emergency physicians see the majority of acutely unwell patients who present to Australasian hospitals, particularly large urban centres, which is of great benefit to these patients.3-5 Thus, creating a new specialty of acute-care physicians to look after these undifferentiated patients seems like unnecessary duplication of service. Indeed, this duplication is one of the main issues in the journey of patients through the acute-care hospital system, particularly in tertiary institutions. After presentation to an acute-care hospital, patients often experience multiple consultations in multiple venues before admission to a home medical ward. They are often seen by a junior emergency doctor, then a senior emergency doctor, then a registrar from a subspecialty, and finally by the general medical registrar — and only then are they “admitted” and the often lengthy wait for a hospital bed commences. This journey could be dramatically improved by a single emergency department consultation with a senior emergency doctor (registrar or above), with stabilisation and immediately necessary investigations being done at that stage. After this, the patient could be either discharged to the community or admitted directly to a home medical ward where they can be seen by the home inpatient team. This would abolish much of the duplication that currently occurs and ultimately make the hospital visit safer for the patient and more cost-effective for the hospital.

Alan E O’Connor

To the Editor: Since publishing its first national report on mortality data from 2009,1 the Australian and New Zealand Audit of Surgical Mortality (ANZASM) has provided coverage of surgical mortality in participating hospitals across Australia. In their recent article promoting the role of clinical-quality registries in improving the quality of health care in Australia, Evans and colleagues2 describe the importance of national registries, particularly in high-cost areas of medicine. Evans et al present the proposed national quality indicators from a 2009 Australian Institute of Health and Welfare report3 and categorise them as current national indicators, indicators requiring data development and those for which a suitable data source has not been identified or substantial development is required to operationalise the indicator. Indicator 36, “Independent peer review of surgical deaths”, is categorised as the third type. The rationale for this indicator stated that the template of the Scottish Audit of Surgical Mortality had been adapted for use in Australia by some states and territories. The report recommended that data from these sources be reported nationally, ensuring that methods of collecting the data would become standardised across the participating states and territories. We wish to highlight that this is now the case. The ANZASM is an independent, peer review audit process overseen by the Royal Australasian College of Surgeons (RACS) and funded by state and territory health departments. The audit is designed to identify and monitor improvements in the quality of surgical care through the collection and analysis of patient mortality data. It aims to improve the the delivery of safe, efficient and effective surgical care by identifying, improving and preventing system and process errors. The database is standardised across all sites. In January 2010, to ensure complete participation by RACS Fellows in this activity in all states and territories across Australia, participation was deemed a mandatory continuing professional development activity (Category 1, surgical audit and peer review). Feedback is provided to individual surgeons on their cases, and overall results are summarised in a de-identified manner as case-note reviews and annual reports that discuss system issues arising on a state and national basis. These issues are analysed further, and recommendations for quality improvement in surgery are disseminated by the ANZASM to the broader surgical community and the respective regional departments of health through its annual reports, case-note-review booklets and, more recently, through workshops and seminars.

Guy J Maddern · Julian A Smith · Wendy Babidge · Gordon S Guy

To the Editor: The research by Rosenwax and colleagues1 and Lowthian and colleagues2 published in the Journal highlights the need for increased capacity in end-of-life care within primary care to reduce the inappropriate use of acute health care services at the end of life. Providing high-quality care for people diagnosed with advanced chronic conditions is among the most complex challenges for general practitioners.3 GPs and other primary care providers are able to provide appropriate palliative and end-of-life care when they are well supported by relevant specialists.3 For patients to be well cared for in the community, it is also necessary for informal carers to have the strength, the will and the skill to provide such care, as well as timely access to support and medical care. The recent National Health and Hospitals Reform Commission’s report4 and the Australian Government’s National Primary Health Care Strategy5 both recognise the need to build “the capacity and competence of primary health care services”4 to support their dying patients. These documents make recommendations that begin to address the current difficulties of caring for these patients in the community. Of significance are recommendations for increased support for carers; improved shared-care arrangements; and better access to specialist palliative care, support and funding for advance care planning and improved access to primary health care professionals.4 This includes a commitment to address workforce shortages and improving out-of-hours access to medical care.5 Such recommendations are positive and will be helpful when they are fully realised. However, issues within primary care — both at the community and individual general practice levels — also need to be addressed. People for whom a palliative approach is appropriate need to be systematically and proactively identified in a timely way. Needs assessment and care planning should be undertaken to ensure that problems and preferences for care are identified and mechanisms are put in place to support such care. To promote optimal end-of-life care, a coordinated, multidisciplinary approach is as important in the community as it is in the hospital setting. Good communication and collaboration between primary care providers, the patient’s specialists and specialist palliative care providers are imperative. Also essential is an ongoing dialogue with the patient and family to enable a clear understanding of the goals of treatment and to proactively plan for likely adverse events. Routinely planning for likely scenarios will potentially reduce the use of acute services and encourage the provision of care in more appropriate environments.

Claire E Johnson · Geoffrey K Mitchell

Objective: To evaluate the prescribing practices of Australian dispensing doctors (DDs) and to explore their interpretations of the findings.Design, participants and setting: Sequential explanatory mixed methods. The quantitative phase comprised analysis of Pharmaceutical Benefits Scheme (PBS) claims data of DDs and non-DDs, 1 July 2005 – 30 June 2007. The qualitative phase involved semi-structured interviews with DDs in rural and remote general practice across Australian states, August 2009 – February 2010.Main outcome measures: The number of PBS prescriptions per 1000 patients and use of Regulation 24 of the National Health (Pharmaceutical Benefits) Regulations 1960 (r. 24); DDs’ interpretation of the findings.Results: 72 DDs’ and 1080 non-DDs’ PBS claims data were analysed quantitatively. DDs issued fewer prescriptions per 1000 patients (9452 v 15 057; P = 0.003), even with a similar proportion of concessional patients and patients aged > 65 years in their populations. DDs issued significantly more r. 24 prescriptions per 1000 prescriptions than non-DDs (314 v 67; P = 0.008). Interviews with 22 DDs explained that the fewer prescriptions were due to perceived expectation from their peers regarding prescribing norms and the need to generate less administrative paperwork in small practices.Conclusions: Contrary to overseas findings, we found no evidence that Australian DDs overprescribed because of their additional dispensing role.

David Lim DrPH · Jon D Emery MB BCh, FRACGP, DPhil · Janice Lewis MBus, DBA, FACHSE · V Bruce Sunderland BPharm, DCC, PhD

Objective: To describe how Medical Benefits Schedule (MBS) chronic disease (CD) item claims vary by sociodemographic and health characteristics in people with heart disease, asthma or diabetes.Design, setting and participants: A cross-sectional analysis of linked unit-level MBS and survey data from the first 102 934 participants enrolled in the 45 and Up Study, a large-scale cohort study in New South Wales, who completed the baseline survey between January 2006 and July 2008.Main outcome measure: Claim for any general practitioner CD item within 18 months before enrolment, ascertained from MBS records.Results: The proportion of individuals making claims for MBS CD items was 18.5% for asthma, 22.3% for heart disease, and 44.9% for diabetes. Associations between participant characteristics and a claim for a CD item showed similar patterns across the three diseases. For heart disease and asthma, people most likely to claim a CD item were women, older, of low income and education levels, with multiple chronic conditions, fair or poor self-rated health, obesity and low physical activity levels. The pattern of claims was slightly different for participants with diabetes in that there was no significant association with number of chronic conditions, smoking or physical activity.Conclusions: Many individuals with self-reported CD do not claim CD items. People with diabetes and individuals with greatest need based on health, socioeconomic and lifestyle risk factors are the most likely to claim CD items.

Kirsty A Douglas DipRACOG, MD, FRACGP · Laurann E Yen BSc, MPsych · Rosemary J Korda BAppSci, MAppSci, PhD · Marjan Kljakovic MB ChB, FRNZCGP, PhD · Nicholas J Glasgow BHB, MB ChB, MD

To the Editor: Equitable access to health care in Australia is facilitated by bulk-billing so that patients incur no out-of-pocket costs for medical services. From 1995 to 2001, there was a steady decline in bulk-billing of general practice consultations and rates of bulk-billing were lower for women living in rural areas than for those from urban areas.1 In 2004, Medicare incentives for bulk-billing were introduced — additional rebates for bulk-billed services provided to concession card holders or children under 16 years, and a higher rebate for services provided to eligible patients in rural and remote areas, selected metropolitan areas with a shortage of general practitioners or low bulk-billing rates, or anywhere in Tasmania.2,3 We assessed the bulk-billing rates for participants in the Australian Longitudinal Study on Women’s Health4,5 following the introduction of these items. We analysed 2002–2008 data on out-of-pocket costs for general practice consultations (services with item numbers 1–98, 601, 602, 697 or 698 in the Medicare Benefits Schedule). Cohorts of older, mid-aged and younger woman (born 1921–1926, 1945–1951 and 1973–1978, respectively) who had consented to the release of Medicare data were included in the analysis. They were classified according to area of residence recorded at Survey 5 (conducted in 2007–2009), using the Accessibility/Remoteness Index of Australia Plus (ARIA+).6 Claims for services, including items charged to the Department of Veterans’ Affairs, were identified from linked Medicare data.7 The study was approved by the Human Research Ethics Committees of the University of Newcastle and University of Queensland. Medicare data were available for 3631 older women, 6697 mid-aged women and 3546 younger women (Box). In 2002, 61% of older women in major cities had no out-of-pocket costs, and this proportion was lower for older women in regional and remote areas. From 2005, there was a marked increase in the proportion of older women with no out-of-pocket costs across all areas, especially in remote and very remote areas (where 87% had no out-of-pocket costs in 2008). Older women from inner regional areas were most disadvantaged in terms of bulk-billing, even after the introduction of bulk-billing incentives. Mid-aged and younger women were less likely to have no out-of-pocket costs than older women but showed similar, albeit less dramatic, increases in bulk-billing. Our data show an overall improvement in access to bulk-billing, although some inequity remains for women in inner regional areas. This contrasts with earlier findings of declining rates of bulk-billing and increasing out-of-pocket costs, particularly in rural areas and for older women.1 The large increases in bulk-billing that we observed for older women are likely to be due to increased use of general practice services overall8 and a higher likelihood of having a concession card. The impact of the concession card holder incentive may have been greater than the geographical targeting. A strength of this study is that the results are based on a large national random sample. A limitation is that women who consented to the release of Medicare data had higher levels of education than non-consenters,9 which may have resulted in underestimation of the proportions of women who had all their consultations bulk-billed. Also, while bulk-billing incentives are aimed at areas defined by Rural, Remote and Metropolitan Areas classification, our data were analysed according to the ARIA+ classification (which is now the standard classification for accessibility and remoteness and is stable over time). The Medicare incentives scheme for bulk-billing should be evaluated further to assess the potential for reducing inequity for people in inner regional areas and for disadvantaged groups who may have a greater need for services but less access. Women who had at least one claim but incurred no out-of-pocket costs for general practice consultations, 2002–2008* 2002 2003 2004 2005 2006 2007 2008 1921–1926 birth cohort Major city (n = 1598) 971 (61%) 875 (55%) 935 (59%) 1131 (71%) 1148 (73%) 1181 (75%) 1183 (75%) Inner regional (n = 1351) 635 (48%) 565 (42%) 634 (47%) 855 (64%) 879 (66%) 922 (69%) 909 (68%) Outer regional (n = 551) 269 (49%) 266 (49%) 309 (57%) 391 (72%) 414 (76%) 403 (74%) 410 (75%) Remote or very remote (n = 62) 35 (57%) 28 (48%) 35 (59%) 46 (78%) 51 (84%) 52 (85%) 54 (87%) 1946–1951 birth cohort Major city (n = 2495) 720 (31%) 624 (27%) 595 (26%) 670 (29%) 711 (30%) 794 (34%) 803 (34%) Inner regional (n = 2711) 429 (17%) 367 (15%) 434 (17%) 579 (23%) 645 (26%) 680 (27%) 743 (29%) Outer regional (n = 1264) 237 (21%) 225 (19%) 254 (22%) 334 (29%) 364 (32%) 409 (35%) 432 (37%) Remote or very remote (n = 213) 59 (31%) 58 (30%) 54 (28%) 63 (33%) 77 (40%) 84 (45%) 86 (44%) 1973–1978 birth cohort Major city (n = 1944) 599 (33%) 457 (26%) 445 (25%) 455 (26%) 475 (28%) 514 (29%) 523 (30%) Inner regional (n = 903) 161 (20%) 135 (16%) 140 (17%) 173 (21%) 176 (21%) 208 (25%) 193 (23%) Outer regional (n = 494) 92 (21%) 83 (19%) 91(21%) 93 (21%) 103 (24%) 107 (25%) 116 (26%) Remote or very remote (n = 115) 30 (30%) 30 (30%) 30 (29%) 31 (31%) 34 (35%) 41 (43%) 36 (38%) * Sixty-nine women from the 1921–1926 cohort, 14 women from the 1946–1951 cohort, and 90 women from the 1973–1978 cohort had no claims for these items during 2002-2008, and n values vary slightly for each year depending on the number of women with a claim for that year.

Xenia Dolja-Gore · Julie E Byles · Deborah J Loxton · Richard L Hockey · Annette J Dobson

United Kingdom primary care trusts resembled the primary health care organisations (PHCOs) that have been proposed for Australia — for example, Medicare Locals. They resulted in a loss of innovation, creativity, motivation and morale among general practitioners and other front-line staff. English primary care trusts are being abolished and £80 billion will be handed over to GP commissioners. Management theory and practical experience shows repeatedly the dangers of reorganising into larger units. Lessons for Australia are to defer deciding on the size of PHCOs until their purposes are clear, to enshrine the principle of subsidiarity, and to opt for networking of the current Divisions of General Practice over mergers. So far, debate on the functions and structures of PHCOs has been muted. It is now time for vigorous debate.

James A Dunbar MD, FRCPEdin, FRACGP

In April this year, the Federal Minister for Health and Ageing, Nicola Roxon, stated that by July 2012 all Australians will be able to “sign up for a personally controlled e-health record . . . [that] will enable better access to important health information currently held in dispersed records around the country”.1 Laudable aims, but can patients and clinicians trust the reliability and confidentiality of this personally controlled e-health record (PCEHR)? The National eHealth Transition Authority’s draft concept of operations document proposes that individuals will be able to access a data repository (“My PCEHR”) and tools (“My Access Controls”) to make this dispersed information available to their chosen health care providers.1 The authors failed to give essential details on how the PCEHR will work, but showed insight in their assessment that the scope and extent of information to be shared in the PCEHR is dependent on the readiness of the health care sector to participate. The concept of democratisation of personal health information is central to the PCEHR, reflecting the populist philosophy of the present Web 2.0 and social networking environment where information is freely published and shared.2 This increases the potential for leakage of information, albeit often unintended, from clinician-held electronic health records (EHRs) via the disparate members of the PCEHR network (Box). Information that, if leaked, might have potential adverse impacts includes family history of disease, and information that may reflect negatively on other health professionals, friends or family members.2 Information leakage, along with complex access and provenance arrangements and individuals “hiding” rather than “denying access” to PCEHR information, will discourage clinicians from participating in a system where they are uncertain about the completeness of the information. Personal health information should stay within a confidential patient–clinician therapeutic relationship, a concept promoted by the patient-centered medical home movement3 and endorsed by the National Health and Hospitals Reform Commission.4 Information exchange within a multidisciplinary care network, no matter who controls it, should be facilitated within the patient–clinician relationship. Ensuring consistency of shared terminology and a method to address misrepresentations will improve the quality of personal health information. Such alignment of patients’ and clinicians’ perceptions and understandings of health concepts and processes will improve health literacy, and this will encourage consumer empowerment, not consumer populism.5 The creation of shared information resources to support collaborative team care must take into account the business, architectural, workflow, provenance and governance requirements from both patients’ and clinicians’ perspectives. However, the success and sustainability of this system also requires careful alignment of the patient’s and the clinician’s EHRs (Box) within a health care home that provides a safe, effective continuum of care within a trusting patient–clinician relationship. This will minimise risks and engender the required confidence to make the PCEHR program work. Pathways for the collecting, sharing, and possible leakage of information in an individual personally controlled electronic health record Reproduced with permission from: Hannan T. What the hell is a PCEHR and what does it have to do with me anyway? The InformaticsInsider [internet] 2011; Jun (1). http://www.austemrs.com.au/page/informatics_insider.html

Siaw-Teng Liaw · Terry Hannan

A unique migration: South African doctors fleeing to Australia. Peter C Arnold. USA: CreateSpace, 2010 (252 pp, $35.00). ISBN 9781452830780. As signalled by its title, Peter Arnold’s treatise is an analysis of South African doctors who migrated to Australia during the latter half of the 20th century. Arnold himself was part of this exodus, soon after graduating in medicine from the University of Witwatersrand in Johannesburg in 1961. He subsequently crafted an illustrious career in Australian medicine, moving between his roles as general practitioner in Sydney’s eastern suburbs, President of the General Practitioners’ Society in Australia, Chairman of the Federal Council of the Australian Medical Association and Deputy President of the New South Wales Medical Board. In fact, Arnold is but one of more than 2000 South African medical graduates who have adopted Australia as their home, and South Africa now joins the United Kingdom, India and New Zealand as a major contributor to Australia’s medical workforce. The book follows a logical order, with chapters on “Why did they leave?”, “Why did they choose Australia?”, “The Australian experience” and “Bringing the kids, but leaving Granny behind”. It also includes a review of the various theories on what made this migration distinctive. Above all, A unique migration is an accessible thesis, in which the wealth of data never overwhelms the narrative. At the same time, it is underpinned by admirable scholarship and an almost encyclopaedic attention to detail, neatly reduced to a series of informative diagrams. It reveals Australia’s debt to this unique migration, which now accounts for one in 30 of the country’s doctors, including luminaries such as Priscilla Kincaid-Smith, Sidney Sax and Michael Denborough. Perhaps the most controversial aspect of the book is whether this great trek was the result of “push factors” such as the social upheaval following the end of apartheid, or “pull incentives” such as Australia’s comparable geographic features. A unique migration is an intriguing journey into a previously undocumented aspect of Australian immigration, and a rewarding book for interested readers. It is compulsory reading for those dealing with the medical workforce. Conflict of interest statement: I know Peter Arnold from when I was editor of the MJA and he was a director of AMPCo, the publisher of the Journal.

Martin B Van Der Weyden

CorrectionTypographical error in base number for calculation of estimates: In “Counting the cost: estimating the number of deaths among recently released prisoners in Australia” in the 18 July 2011 issue of the Journal (Med J Aust 2011; 195: 64-68), the incorrect number 50 504 was used as a basis for calculating some estimates instead of the correct number, 50 405. This has resulted in small errors in some numbers in two tables and one paragraph of the Results in the article. These errors are not substantive and do not alter the conclusions of the study. The numbers have been corrected in the online version of this report (http://www.mja.com.au/public/issues/195_02_180711/kin10879_fm.html).

Stuart A Kinner · David B Preen · Azar Kariminia · Tony Butler · Jessica Y Andrews · Mark Stoové · Matthew Law

How to ensure a better match between grant applicant and reviewer expertise All researchers have experienced dismay at the failings of peer review.1,2 These include cursory or ill informed reviews from those apparently unschooled in applicants’ disciplines and with superficial understanding of the quality and significance of proposals, publications and achievements being described by aspiring applicants. Only the greenest researchers assume that those who assign papers or grants to reviewers possess Solomonic wisdom in matching the expertise of researchers and reviewers. The reality is very different. Reviewer assigners rely on reviewer databases, particularly for areas beyond their specific expertise. It is here that many problems start. Some names are in these databases by virtue of having previously submitted a grant or paper, regardless of quality or success. Self-chosen research keywords may bear little correspondence to actual expertise. Many relatively inexperienced names appear and searches suggest many who have published little. Grant review panels seek reputable reviewers but reviewer refusal means suboptimal reviewers are often used. No published information is available on grant review refusal rates, but data obtained from the BMJ Publishing Group for 300 866 review requests sent out by 20 journals between 2002 and 2009 show that 32.4% of those asked declined to review.3 Reviewer assigners can then resort to selecting names on the basis of reviewer supplied keywords, without detailed knowledge of the reviewer. Although this can sometimes produce superb reviews, it can also produce those that are confidently written but ill informed. The core problem with current approaches to reviewer selection is that a small number of assessors are involved, typically two or three. Although it would be impractical to involve more people in grant assessments, the rating of track record — typically attracting at least 25% of overall score — could easily break free from the reputational lottery affected by the opinions of a few, sometimes ill informed, people that may affect the fate of a grant application. There is an easily managed, low-cost way of democratising the rating of researcher track record that would end the practice of relying on a small number of sometimes inexpert peers rating applicants’ track records. As part of a project examining the nature of researcher influence, we invited all Australia-based researchers in six research public health fields who had published 10 or more Institute for Scientific Information-indexed publications in the past 10 years to nominate up to five Australian researchers in their respective fields whom they considered to be the most “influential” researchers. Full descriptions of the selection of participants are provided elsewhere.4,5 Participation was high: 83% of invited eligible Australia-based researchers (175/211) completed the rating task, which took about 10 minutes. One hundred and eighty-two individuals across these fields received at least one nomination, producing a distribution of “votes” on researcher influence across these researchers. In five of the six fields, there was impressive consensus on which researchers were most influential, with a long tail of researchers receiving a smaller number of votes, and many none. In all six fields, a large number of researchers (108/182; 59%) received only one vote from all those voting. Of these, 48 (26% of all nominations) were self-nominations. Peer-influence rating generally correlated strongly with four commonly used bibliometric indices of research impact6 (the h, m and q2-indices and the m-quotient).7-10 The traditional approach to track-record ranking involves a small number of researchers being selected as reviewers by an often idiosyncratic process that reflects selector knowledge and preferences and reviewer self-nomination and availability. The existing process lacks transparency and probably provides biased results from an unrepresentative sample of reviewers. Our study demonstrated that the process can be opened up by seeking votes from all research-active peers in each relevant field. Most researchers agreed to participate and found the task unproblematic. We received no critical feedback from those involved. We asked participants to select five researchers they considered to be “most influential” in their fields. The concept of “researcher influence” is arguably wider than that of research track record, but we could have just as easily asked participants to rate the narrower concept of “research track record”. Equally, we could have asked participants to nominate and rank a larger, but non-onerous, number of influential researchers (say, 20), which would have produced a more finely grained distribution of voting, facilitating the overall ranking into quartiles or quintiles that ranged from those considered by many to be influential through to those considered influential by none of their peers. Our proposalIt could be made mandatory for all researchers submitting grant applications to rank the track records of a reasonable number of their peers. Each applicant could be sent a random, computer-generated list of research peers (say, 10) from their respective research fields who where also seeking funding and to rank them as a condition of grant application submission. The generation of these lists could be stratified to ensure that all researchers, regardless of experience, were guaranteed to be ranked by 10 randomly selected peers. Every applicant would thus receive a score of between 100 (if all 10 peers ranked them first) and 10 (if all ranked them 10th). Track record could then be taken as this aggregated raw score. To assist ranking, links could be provided to publications, citations and grant successes. In Australia, the National Health and Medical Research Council now requires all applicants to complete an online track record where all publications and competitive grants are entered into a database which could be linked for this purpose. There are potential weaknesses in this process, although none that are weaker than the approach now operating. Early career researchers would be disadvantaged as they would be unlikely to be competitive with more experienced researchers and would receive lower rankings. This could be addressed by finally dividing all researchers into either “open” or “early career” divisions. Some researchers who challenge research orthodoxies may have the importance of their work discounted by those inhabiting those orthodoxies, and receive fewer votes than their contributions might deserve, but this is already possible under the present system.

Simon Chapman PhD, FASSA · Gemma E Derrick PhD · Abby S Haynes MA · Wayne D Hall PhD, FASSA

Objective: To estimate the number of deaths among people released from prison in Australia in the 2007–08 financial year, within 4 weeks and 1 year of release.Design, participants and setting: Application of crude mortality rates for ex-prisoners (obtained from two independent, state-based record-linkage studies [New South Wales and Western Australia]) to a national estimate of the number and characteristics of people released from prison in 2007–08.Main outcome measures: Estimated number of deaths among adults released from Australian prisons in 2007–08, within 4 weeks and 1 year of release, classified by age, sex, Indigenous status and cause of death.Results: It was estimated that among people released from prison in 2007–08, between 449 (95% CI, 380–527) and 472 (95% CI, 438–507) died within 1 year of release. Of these, between 68 (95% CI, 56–82) and 138 (95% CI, 101–183) died within 4 weeks of release. Most of these deaths were not drug-related.Conclusion: The estimated annual number of deaths among recently released prisoners in Australia is considerably greater than the annual number of deaths in custody, highlighting the extreme vulnerability of this population on return to the community. There is an urgent need to establish a national system for routine monitoring of ex-prisoner mortality and to continue the duty of care beyond the prison walls.

Stuart A Kinner PhD · David B Preen BSc(Hons), PhD · Azar Kariminia BSc, MSc, PhD · Tony Butler PhD, MSc · Jessica Y Andrews BHSci/Comm(Hons) · Mark Stoové PhD · Matthew Law MA, MSc, PhD

Objective: To examine the frequency of deaths in low-mortality diagnosis-related groups (LM-DRGs) and the patient and hospital characteristics associated with them.Design, setting and patients: Retrospective cohort study of 2 400 089 discharge episodes for adults (> 18 years) from 122 Victorian public hospitals from 1 July 2006 to 30 June 2008.Main outcome measures: Frequency of episodes of death in LM-DRGs (defined as DRGs with mortality < 0.5% over the previous 3 years or < 0.5% in any of the previous 3 years); associations between characteristics of patients and hospitals with deaths in LM-DRGs.Results: There were 1 008 816 LM-DRG episodes with 0–15 LM-DRG deaths per hospital in the 2006–07 financial year and 0–20 deaths per hospital in the 2007–08 financial year. Increased age, level of comorbidity, being male, admission from a residential aged care facility, interhospital transfer, emergency admission and lower hospital volume were associated with an increased risk of death in LM-DRG episodes in both years. Metropolitan location and teaching/major provider status were not associated with LM-DRG deaths (P > 0.10). More than 40% of LM-DRG deaths were among patients aged 83 years or over, who had a length of stay of less than 1 day and had a medical DRG classification. Standardised mortality ratios (SMRs) that adjusted for the patient and hospital characteristics identified nine outlier hospitals with high frequencies of deaths in LM-DRGs in the 2006–07 and six in the 2007–08 financial year compared with 59 hospitals flagged by the death-in-LM-DRG indicator.Conclusions: The use of the LM-DRG indicator requires further investigation to test its validity. LM-DRG deaths are infrequent, making it difficult to identify temporal changes and outlier hospitals. Patient characteristics unrelated to quality of care increase the likelihood of death among LM-DRG patients. The SMR analysis showed that failure to adjust for these characteristics may result in unfair and inaccurate identification of outlier hospitals. The increased risk of death associated with interhospital transfer patients and low-volume hospitals requires further investigation.

Anna L Barker PhD, MPhty(Geriatrics), BPhty · Caroline A Brand MB BS, FRACP, MPH · Sue M Evans BN, GDip ClinEpi, PhD · Peter A Cameron MB BS, MD, FACEM · Damien J Jolley MSc(Epidemiology), MSc(Statistics), AStat

Neil Orr was born on 19 October 1919 on a property called “St Helens” in Toowoomba, Queensland, the only child of Scottish immigrants. Neil attended school at Toowoomba Grammar and studied medicine at the University of Sydney. At university, he stayed at Wesley College, where he was in the rowing team. Neil graduated in 1944 and served his internship at Royal Prince Alfred Hospital, Sydney. He then set sail for England as a ship’s surgeon on the HMS Bellerophon, which stopped on the way at Batavia to collect the Dutch gold that had been sent there for safety at the beginning of the First World War. He arrived in England in 1947, where he worked at several hospitals, including Hillingdon Hospital, Uxbridge, where he undertook most of his surgical training. In 1960, he was awarded Fellowship of the Royal College of Surgeons. In the early 1960s, Mona Vale Hospital opened on Sydney’s northern beaches and Neil was appointed its first Medical Superintendent. He helped set up the hospital and recruit staff. A house in the hospital grounds was built for the Superintendent, with a magnificent view over the ocean and golf course. Neil was a keen surf swimmer and swam regularly. Neil remained in this role until he retired in 1986. As Superintendent, Neil regularly visited all sections of the hospital and knew all who worked there, including nurses, pharmacists, and kitchen and ground staff. Although he no longer practised as a surgeon, he maintained his interest by assisting consultant surgeons at the hospital. In his later years, he helped his close friend Dr Josephine Wiseman in her nuclear medicine practice. She subsequently looked after him in his last few years when he was confined to a wheelchair. After a long illness, Neil died on 9 August 2010. James B Roche

James B Roche

No-fault compensation, based on the ethical principle of redistributive justice, should form a cornerstone of Australia’s immunisation strategy Australia has an enviable reputation for its publicly funded vaccine program — a program that has benefited Australian children and adults over many years. In 2010, the National Immunisation Program funded 12 vaccines, twice as many as a decade previously. To monitor outcomes from this program, the Australian Childhood Immunisation Register, which commenced data collection in 1996, provides a detailed record of vaccine uptake by children.1 Funding for the register and for incentives to general practitioners to improve vaccine uptake are part of the total budget for Australia’s vaccine program, estimated to exceed $400 million annually.2,3 One area for improvement in the vaccine program is monitoring of adverse events following immunisation (AEFI). Another would be the introduction of a no-fault compensation scheme for serious adverse events which can be confidently attributed to vaccination. An investigation into the unexpectedly high number of febrile convulsions in children aged less than 5 years after they had received the influenza vaccine in 2010 — in some cases, with devastating consequences4 — provided a forceful reminder that timely vaccine safety monitoring is needed in Australia.5 More active adverse event surveillance is certain to uncover more AEFI but many of these will only be coincidental, while others will be of a transient or relatively trivial nature. On rare occasions, a serious AEFI with long-term sequelae will be recognised. A decision will then need to be made on whether the vaccine was responsible for that serious event. The World Health Organization defines four categories of serious AEFI: hospital admission or prolongation of an existing hospital admission; permanent disability; any event that is life threatening; or death.6 Using these criteria, 8% (193/2396) of the AEFI reported by passive surveillance in Australia in 2009 were judged to be serious.7 However, unlike many countries where compensation schemes exist for adverse events attributed to a vaccine, Australia has no routine approach to making the assessment of attribution. Parents of children or adults who believe they deserve compensation for a serious adverse event that they attribute to a vaccine are therefore required to make their case through the adversarial legal system. This requires the demonstration that an individual or an organisation was at fault. However, fault is often difficult to demonstrate and an adverse event may be caused by vaccination through no fault of the vaccine manufacturer, the regulator or the person who administered the vaccine. We have previously argued that a Queensland child who developed transverse myelitis after receiving oral polio vaccine was an example of an adverse event following vaccination where no fault was attributable to any party.8,9 Despite detailed epidemiological evidence that was consistent in this case with the causal criteria for an AEFI promulgated by the Institute of Medicine of the National Academies in the United States,8 and despite laboratory evidence showing that the polio virus recovered from this child was similarly pathogenic to a polio virus that has been accepted as causing vaccine-associated paralytic polio,9 the polio expert committee concluded that the evidence was insufficient to support a causal relationship between the oral polio vaccine and transverse myelitis. As causality has not been accepted, this child has received no compensation. The general principles associated with this case raise a number of pertinent questions for Australia. First, should a child who may have been injured by a vaccine, which was endorsed and paid for by the community, be compensated by the community when the serious adverse event may be attributed to the vaccine? Second, what are the criteria for accepting an attributable relationship between receipt of the vaccine and a subsequent adverse event? Third, what is the best method for financing a compensation scheme? Each question may highlight a potential barrier to the implementation of a no-fault AEFI compensation scheme in Australia. By 2010, 19 countries around the world had implemented no-fault AEFI compensation, implicitly answering “yes” to the question of whether the community owes a duty of care to an individual injured by a vaccine.10 There is also a strong ethical argument for this position, based on the concept of redistributive justice. Any person who is injured while helping to protect the community — for instance, by contributing to herd immunity, such that there are sufficiently many people immunised to prevent widespread disease transmission within the community — should not bear the consequences of injury alone. In essence, the community owes a debt of gratitude to that person. Temporal association of an adverse event with receipt of a vaccine does not establish causality and the underlying notion of causation used in most compensation schemes is similar to that used in epidemiology.10 The World Health Organization has published guidelines on causality for an AEFI.11 An adverse event considered to be very likely or certainly due to a vaccine would comprise a “Clinical event with a plausible time relationship to vaccine administration, and which cannot be explained by concurrent disease or other drugs or chemicals”.11 To simplify and expedite determinations of causality in the US, a vaccine injury table is used to predetermine causality if a vaccine injury is included in the table.10 However, determining causation is a complex issue. Recognising this, most countries have a designated committee, comprising medical and legal members, which deliberates on the attributable relationship between receipt of the vaccine and subsequent adverse event.10 Concerns about funding a no-fault compensation scheme is another of the probable barriers to its implementation in Australia. Schemes are currently funded by one of four methods: a vaccine levy; compensation for AEFI as part of a much broader injury compensation scheme; specific AEFI compensation funded through general tax revenue; and funding in association with industry.10 Funding through a vaccine levy has been self-sustaining in the US. Despite compensation payments having been made to 2580 claimants since 1989, the compensation fund there has a surplus of about US$3 billion.12,13 No-fault vaccine-injury compensation programs are based on the premise that any adverse event attributable to vaccination is not due to the fault of a specific individual or organisation, but due to an unavoidable risk that is acknowledged as being associated with vaccines. Germany has been operating a no-fault AEFI compensation scheme for 50 years.10 France restricts its compensation to serious AEFI, since these are likely to have long-term implications for the injured party.10 Restricting compensation to events with long-term consequences, above a nominated clinical threshold, may be an acceptable model for Australia. We have previously argued that Australia should follow the lead of other advanced countries and implement a no-fault compensation scheme.14 We continue to argue that such a scheme, based on the ethical principle of redistributive justice, should form a cornerstone of Australia’s immunisation strategy. Disclaimer The views expressed are those of the authors and have not been endorsed by any institution or organisation with which the authors are affiliated or by any committees of which the authors are members.

Heath A Kelly BSc, MB BS, MPH · Clare Looker MB BS, MPH · David Isaacs MD, FRACP, FRCPCH

Medical academics Christine Jorm and Michael Frommer believe it is simply the right thing to do For vast amounts of performance data are collected in the Australian health system, many of which are released by agencies such as the Australian Institute of Health and Welfare, state and territory health departments and the Australian Bureau of Statistics. However, merely releasing performance data is different from publicly reporting such data. Public reporting incorporates interpretation and comparisons that make the data meaningful for the community. Public reporting of health care performance has three major uses: (i) to ensure accountability — the Australian public, which pays for health care, is entitled to assess the effectiveness and efficiency of health care and to press for change as needed; (ii) to stimulate action by health care providers that leads to improvements; and (iii) to give consumers information on which they can base their expectations of health services and individual health care choices. All of these uses are important, and the scrutiny afforded by public reporting of performance should therefore be welcomed. It is difficult to determine what types of reporting represent the best investment, because the consequences of public reporting are difficult to measure. The quantum of research on the effects of any complex health policy intervention is scant. Unsurprisingly then, evidence on the value of public reporting is currently limited.1,2 What have we learnt from the international experience? Research in the United Kingdom and the United States shows a thirst among consumers for publicly reported performance data, while pointing out the difficulty of producing data that people can readily use for making health care choices.2 In neither the UK nor the US has patient choice in itself been shown to have had a reliable influence on health care quality. However, when the public is made aware of poor health care service performance, calls for political action are common. This is a desirable outcome. Public reporting — particularly reporting of institutional performance — consistently stimulates health care providers to improve quality.1 For example, after the first publication of comparative cardiac surgery outcomes in New York State, some surgeons who recognised that they were outliers on performance scales voluntarily changed their scope of practice, or retired. More commonly, it is institutions that undergo performance evaluations, and they usually respond vigorously to improve care processes and patient outcomes. This is especially so when public reporting is framed by expectations that people understand and value, such as accompanying goals and targets.2 Does performance measurement corrupt the delivery of health care? It is secrecy that corrupts. The absence of public reporting generates suspicion and cynicism among both clinicians and the public. Notably, the Australian Medical Association submission on the National Health Reform Amendment (National Health Performance Authority) Bill 2011 advocated stronger investigative and disciplinary powers for the National Health Performance Authority and, importantly, a mandatory requirement to release reports. Opponents of public reporting highlight the potential for merely improving the data rather than the quality of care (for instance by finding more risk factors to include or reclassifying chairs as beds) and for neglect of high-risk patients, but evidence from the UK suggests that both are rare. The introduction of targets and public reporting of emergency department waiting-time data drove real reform.3 In a study of more than 27 000 patients who had cardiac surgery, risk-adjusted mortality fell after the introduction of public reporting, and there was no evidence that fewer high-risk patients were offered surgery.4 Some Australian clinical leaders choose to focus their opposition to public reporting on the impossibility of providing adequate risk adjustment — that is, of fully accounting for the differences between individuals and their circumstances. This is an argument that, when taken to extremes, would invalidate most of the evidence base for medical practice. However, reporting of process indicators can reduce the need for, and the debate over, sufficient risk adjustment. Rather than corrupting delivery, the Queensland experience — where there is timely return of data to health care providers, and performance measures are accompanied by reporting on the institutional responses to the data — has shown that public performance reporting can create a culture of improvement.5 Could the personal cost to the clinicians and managers be too great to justify implementation of public reporting? Public reporting affords protections and benefits, and should be welcomed as the price of public service and funding. The risk of unfair reproach will be mitigated by the use of reliable performance measures, timeliness, high-quality data, rigorous analysis, and integrity and sensitivity in reporting results. Review, comparison and criticism are essential components of professional practice. The National Open Disclosure Standard requires clinicians to share uncomfortable truths about adverse events with individual patients. Honesty is a central value of professionalism, and the wellbeing of patients must take precedence over professional self-interest. Public performance reporting promotes strong adherence to these values and therefore should be at the core of good health system governance.

Christine M Jorm MD, PhD, FANZCA · Michael S Frommer MB BS, MPH, FAFPHM

Health services experts Jeffrey Braithwaite and Russell Mannion doubt the value Against the idea that publicly reported performance measurement, like apple pie, parenthood and the national flag, deserves a warm, uncritical glow of universal support should be roundly rejected. Introducing any costly new initiative must always be analysed with a cool head for its risks and potential downside. Where is the Australian business case, or the international cost–benefit analysis or cost-effectiveness analysis, applied to Australia, that compels us to not only accept, but insist on its introduction? These have not been provided by proponents to date. It is a considerable logistical exercise to collect, process, analyse and distribute national data, and it is therefore very costly. Do the benefits outweigh the costs? Almost everyone will have their doubts. There is probably no clinician or manager who has not reported information to some system or other, and never heard about it again. Health care is renowned for “hoovering up” data, which get stuck in the system. Will a national performance measurement initiative fare any better? We prefer to maintain a healthy scepticism. Most experts in favour of performance measurement argue that the twin aims are to improve accountability and enhance the system’s performance.1 But health care’s enormous complexity must be acknowledged. There are many stakeholders, a complicated multiplicity of services and products delivered through many public and private providers contributing millions of encounters across acute, aged, primary and tertiary sectors, with increasing emphasis on prevention, promotion and community care. There is heavy political involvement in health care, and much media attention, both of which distort priorities. Against this heady mix of systems changeability, key challenges are to determine what should be measured, and how, and from whose perspective, while ensuring fairness and objectivity. These questions have not been answered satisfactorily. Even if they were, there are several other problems. One is how to solve technical issues about data quality and the effectiveness of measurement. For example, are the input data collected in the same way, are they gathered systematically by all providers or according to different coding and institutional rules, and what is the extent of gaming (ie, portraying or modifying data to one’s strategic advantage) by participants? How accurately do the data reflect actual performance, are apples being compared with apples and how good are the information systems or the data collection measures that produce the data? It is well known that it is extremely hard to measure performance because of difficulties in risk adjusting. Different risk adjustment models give rise to different outcomes.2 Another issue is whether this initiative will have the desired outcomes. There are many examples of performance measurement systems which, even when they have solved some of the main technical problems, have failed to have meaningful effects. The reporting data are ignored, argued over or politicised, improvement efforts founder, or targets and indicators have perverse effects.1 Indeed, public performance measures are not neutral assessments of performance, but can alter behaviour in unintended and dysfunctional ways. All this gives rise to the potential for Type 1 errors (higher performing organisations or services are assessed as underperforming) or Type 2 errors (lower performing organisations or services are assessed as adequately performing). Traditionally, the attention has been on avoiding Type 1 errors, but since high-profile inquiries at Bristol Royal Infirmary in the United Kingdom, King Edward Memorial Hospital in Western Australia, Campbelltown and Camden hospitals in New South Wales, and others,3 attention has shifted to avoiding Type 2 situations.4 It is very tricky to get the balance right. No performance measurement systems internationally claim to have done so. There are also timing and attribution issues that need to be resolved. Performance measurement systems are necessarily backward looking, as it takes time to assemble and disseminate data. By the time a problem is spotted, it may be too late to do anything about it. There is also the attribution problem: when good or bad performance is observed, is it causally related and assigned correctly? In systems where many things are changing simultaneously — and health is the par-excellence exemplar of this — this is an ongoing issue. So what can contribute to success? Apart from a good data collection system and agreed definitions, targets and indicators (none of which we have at this point), we need excellent partnerships between sectors, agencies and health departments; leadership, not politics; incentives to participate; really good communication of outcomes; fair media reporting of results; and well designed mechanisms to improve performance. It remains doubtful whether these can be readily achieved in Australia. All in all, performance measurement systems often have little impact on changing behaviour or improving performance.5 As that is the point of them, and until the fundamental problems we describe are sorted out, we respond with a resounding no to the proposition.

Jeffrey Braithwaite PhD, FACHSM, FAIM · Russell Mannion PhD, FRSA

To the Editor: The Health Practitioner Regulation National Law Act 2009 (Part 8, Sections 140 and 141) enshrines mandatory notification in the new national registration framework. As registered health practitioners, clinician-researchers are bound by the notification requirements. This raises the question of whether mandatory notification has implications for observational research in health services that is conducted by clinician-researchers. In particular, how likely is it that these requirements will lead to reclassification of one’s observations from “research data” to “notification evidence”? Three initial considerations are important here. First, the Act was designed to make health care safer for patients. Its intent is to limit incidents by ensuring clinicians are more open about and address inappropriate care. Second, serious incidents are rarely isolated, instantaneous and therefore easily observable disasters. When something goes seriously wrong, problems tend to be inherent in how teams practise, communicate and support one another over time. Third, observers may encounter instances of substandard care, but these become notifiable only when the threshold of unsafety is surpassed. This threshold is pegged to relatively high levels of severity, frequency and risk.1 In all, observational research can help clinicians to identify existing risks, but it is unlikely to become a source of notification. Human research ethics committees may also feel obliged to acknowledge and consider the possibility of incident notifications arising from observational research. However, it would not be wise to regard the risk of such notification as detracting from a study’s potential for obtaining ethics approval. The situation calls for specification of: how the observers will deal with incidents if and when observed the observers’ understanding of the definition and threshold of notification how the definition of notification is likely to bear on the study how the design of the study affects the likelihood of notification (eg, does the researcher seek to identify care irregularities or track these irregularities over time?) a projection of the relevant service’s vulnerabilities to notification, and plans for addressing and resolving existing vulnerabilities. In addition, mandatory notification does not mean that observational research will be more difficult to “sell” to ethics committees and frontline clinicians. The aim is generally to stimulate learning and raise awareness of problems. Our experience is that if the research is designed with frontline clinicians, and they contribute to its implementation, analysis and publication,2 it attracts considerable interest and support.3 Frontline staff know that the best way to understand the complexities inherent in their everyday work is through observation. Such research takes seriously their specific and unique circumstances, enabling them to actively participate as analysts and improvers of their own practice. Observation encourages reflection, and this means they become aware of and can proactively resolve their own vulnerabilities. Ultimately, the priority for clinician-researchers involved in observational research in health services, as for patients, is to reduce risks and prevent incidents.

Rick A M Iedema · Donella A Piper

To the Editor: I was extremely surprised to note that there was no mention of the Australian Council on Healthcare Standards’ (ACHS) extensive national clinical database (http://www.achs.org.au/ClinicalIndicators) in the recent article by Evans and colleagues on clinical-quality registries.1 Since 1993, commencing with a small set of generic indicators that were of limited value, the ACHS has been collecting clinical data through its clinical indicator (CI) program as part of its accreditation process. Now, with over 300 CIs in use and around 670 health care organisations (HCOs), including some from New Zealand, contributing data, the scope of the ACHS national clinical database is unique in the world. Its longevity makes it one of very few national programs that can display longer term trends in processes and outcomes — both desirable and undesirable — of medical care. The clinical indicators, which form the basis of the database, were all developed in conjunction with providers of medical care, namely, the medical colleges. The validity, reliability and effectiveness of the indicators were established (and published) subsequent to introduction of the more specific indicators into the accreditation process.2 In addition to providing aggregate and peer-comparative data to the contributing HCOs, the ACHS publishes aggregate data with detailed analyses on an annual basis.3 As with similar databases, problems have arisen regarding maintenance of currency and relevance, timelines for reporting, and other issues. However, it seems extraordinary that this valuable aid to determining the quality of health care standards — envied by many countries — could be ignored by Evans et al, as it was similarly ignored by the authors of a recent special MJA supplement that reported on gathering patient-centred health care data aimed at improving care.4

Brian T Collopy, AM

CorrectionsIncorrect estimate of number of lives saved by the use of surgical checklists: In “Is money spent on quality improvement better spent on clinical care?” in the 20 June 2011 issue of the Journal (Med J Aust 2011; 194: 641), the order of magnitude of the number of lives potentially saved in Australia by surgical checklists is incorrect. The potential saving is thousands of lives rather than “tens of thousands of lives”. The html and pdf versions of this article were corrected when published online on 20 June 2011.

William B Runciman

We need to balance the technical challenges of hospital standardised mortality ratios with the need to improve care processes That “sunshine is the best disinfectant” is attributed to the distinguished American jurist Louis Brandeis, who spent much of his career supporting individual rights in the context of corporate and monopoly power.1 State and federal governments, in conjunction with bodies such as the Australian Commission on Safety and Quality in Health Care and the Australian Institute of Health and Welfare, are developing indicators for the safety and quality of Australian hospital care, presumably with the intention of using them to illuminate some of the inner workings of our hospital system for the benefit of both health care providers and the community at large. The challenges posed by such a program are illustrated in this issue of the Journal by Scott and colleagues2 and Gallagher and Krumholz.3 Health care professionals face a basic dilemma. The vast majority of us are hard working, conscientious and altruistic. Yet the hospitals in which we work collectively expose patients to substantial risks over and above those posed by their clinical conditions, and improving hospital safety and quality is work that ultimately can only be done by hospital staff. The United Kingdom Department of Health recently summed up the task of a safety and quality indicator such as the hospital standardised mortality ratio (HSMR): A high HSMR is a trigger to ask hard questions. Good hospitals monitor their HSMRs actively and seek to understand where performance may be falling short and action should not stop until the clinical leaders and the Board at the hospital are satisfied that the issues have been effectively dealt with.4 The challenge is that exposure to a safety and quality indicator will trigger hard questions about the technical qualities of the indicator rather than underlying care processes, and opportunities for improvement will be lost. The technical issues canvassed2,3 include the value of coded administrative data as a source for risk adjustment; analysis and interpretation of data from small hospitals; and problems related to classifying and coding hospital palliative care provision. All require detailed work, such as is in progress for HSMRs.5 Furthermore, the Australian Commission on Safety and Quality in Health Care is going through a rigorous development process for at least 16 measures.6 A dilemma in indicator development is whether to accept the merely good or await the perfect. Discussion about technical matters should not divert attention from the basic questions about what available indicators such as relative hospital mortality actually measure, and how they, and other indicators, should be used and distributed. Mortality indicators compare observed mortality against a risk-adjusted expected mortality derived from a broader reference source. Differences are silent as to cause. They are simply a prompt for hard questioning. Process measures of adherence to evidence-based care pathways are well established quality indicators that provide immediate feedback on areas for improvement. Mortality measures do not correlate well with process measures.2 Does this undermine the validity of mortality as a safety and quality indicator? Only if process measures are taken as the gold standard of hospital safety and quality. But what if a measure of getting evidence-based steps right (a process measure) is not directly related to a measure of what happens when things go wrong? What if hospital mortality measures are at a tangent to process measures, rather than directly consequential?7 Then the hard questions provoked by an elevated mortality rate of any kind might do well to begin with examining the capacity to rescue patients when things go wrong,8 as they will in even the best-organised departments. Scott and colleagues2 provide excellent practical advice on how institutions might use indicators of various kinds to provoke hard questions. They are wary, however, of providing hospital staff, or the public at large, with comparative information on hospital outcomes. The efficacy of comparative and public reporting as a prompt to improving hospital safety and quality cannot be settled by rigorous scientific means. It is just not possible to carry out randomised double-blind trials of accurate versus inaccurate dummy comparative reporting. The extent of public reporting is a public policy issue in which the risks to institutional reputation and morale have to be balanced against a need for accountability and the opportunity for informed choice by current and future users of health care institutions.9 Gallagher and Krumholz note that Australia has lagged behind other countries in publicly reporting hospital outcomes.3 As Brandeis might have argued, one of the most important virtues of public reporting is that it makes it harder for vested interests, whether they are in government or institutions, to suppress unwelcome information. No doubt the debate has just begun.

David I Ben-Tovim PhD, MB BS, FRANZCP

Objective: The development of an effective therapeutic equivalence program (TEP) through the collaborative support of medical staff, using the principles of disinvestment.Design and setting: A TEP was introduced at Southern Health, a metropolitan health service in Melbourne, in the 2006–07 financial year. Therapeutic classes were selected for the TEP by stakeholder consensus, and a preferred medication for each class was selected on the basis of cost considerations and therapeutic equivalence. New patients were commenced on preferred medicines, but patients receiving another medicine from a therapeutic class included in the program were not automatically switched to the preferred medicine. For the first 4 years of the program, prescribing patterns were monitored, and savings achieved (due to lower prices for and increased use of preferred medicines) were calculated on a monthly basis.Main outcome measures: Prescribing trends for preferred medicines, as a measure of acceptance of the TEP, and savings produced by the program.Results: Over the 4-year study period, 11 therapeutic classes were targeted. The use of all preferred medicines increased once they become part of the TEP and a total of $3.16 million was saved. The annual savings increased each year, and the rate of increase was six times that of the increase in patient separations.Conclusions: The TEP at Southern Health resulted in significant savings. It showed that, by using a collaborative and evidence-based approach, the principles of disinvestment can be applied to use of medicines.

Ian Larmour BPharm, FSHP, MSc · Silvana Pignataro BPharm(Hons) · Kerryn L Barned BPharm(Hons) · Stav Mantas BPharm(Hons) · Melvyn G Korman MB BS, PhD, FRACP

Objectives: To assess the utility of Australian health care incident reporting systems and determine the depth of information available within a typical system.Design and setting: Incidents relating to patient misidentification occurring between 2004 and 2008 were selected from a sample extracted from a number of Australian health services’ incident reporting systems using a manual search function.Main outcome measures: Incident type, aetiology (error type) and recovery (error-detection mechanism). Analyses were performed to determine category saturation.Results: All 487 selected incidents could be classified according to incident type. The most prevalent incident type was medication being administered to the wrong patient (25.7%, 125), followed by incidents where a procedure was performed on the wrong patient (15.2%, 74) and incidents where an order for pathology or medical imaging was mislabelled (7.0%, 34). Category saturation was achieved quickly, with about half the total number of incident types identified in the first 13.5% of the incidents. All 43 incident types were classified within 76.2% of the dataset. Fifty-two incident reports (10.7%) included sufficient information to classify specific incident aetiology, and 288 reports (59.1%) had sufficient detailed information to classify a specific incident recovery mechanism.Conclusions: Incident reporting systems enable the classification of the surface features of an incident and identify common incident types. However, current systems provide little useful information on the underlying aetiology or incident recovery functions. Our study highlights several limitations of incident reporting systems, and provides guidance for improving the use of such systems in quality and safety improvement.

Matthew J W Thomas PhD · Timothy J Schultz PhD · Natalie Hannaford DipAppSc(Nurs) · William B Runciman PhD

Physician Alasdair Millar believes that QI siphons resources for little return and threatens health YES Quality improvement (QI) is now a dominating influence in health care. The overall aims are good: to ensure patient safety, avoid errors and achieve optimum health “outcomes”. Who would dare criticise? I argue that clinical QI practice is failing to satisfy its ostensible aims and is a threat to health. QI in hospitals has a wide range of functions, such as developing clinical pathways, accreditation, maintaining professional competence, reporting sentinel events and reviewing morbidity and mortality, and behind each of these is a bureaucracy. Thus, aggregate costs are high. Every hospital has a quality unit, and there are state and federal quality organisations under various names with duplicated or overlapping roles. The cost of government-funded QI across Australasia could easily exceed seven figures, without including the unknown costs of individual projects. The marginal cost of QI is unmeasured, but must also be high. Modern hospitals operate at a high degree of safety, with error rates of individual types (eg, prescribing or surgical errors) being less than 2%.1 Improvement is possible, but is necessarily affected by the law of diminishing marginal returns. Since QI is funded from within the health budget, funds are siphoned from direct patient care, compromising clinical outcomes. Silo budgeting negates direct effects, but indirect competition certainly exists. For example, the “SQuIRe” project cost the Western Australian Health Department $24 million over 3 years.2 Its funding was recently extended,3 while clinical services were under threat.4 The New Zealand government has recently introduced a maternity QI program,5 but will fund it at only four “demonstration sites” while simultaneously requiring all health boards to decrease expenditure in real terms. Hospital QI is dominated by policies generated by government organisations and other external influences, such as the need to maintain accreditation or to implement standardised processes. In response to avoidable hospital deaths, the New South Wales Government created two new state agencies and expanded another,6 at what is likely to be substantial cost. The roles of such agencies are presented to the public in a favourable light, but have the effect of restricting the capacity of individual hospitals to analyse and solve their own problems. Whether patients obtain any benefit from centrally or politically derived policies is moot. There is little evidence of clinical benefit from QI, and some evidence of failure. Mandatory reporting of sentinel events was introduced as a means of reducing such events,7 but the trends to date are constant (Victoria) or the opposite (Western Australia and New Zealand). One large study showed improvements in process measures but not in clinical outcomes,8 and, in the United Kingdom, efforts to improve doctors’ working hours has been “spectacularly unsuccessful”.9 The SQuIRe website contains no outcome data, and those of the three new NSW agencies mentioned above contain no record of patient benefit. Other authors have also questioned whether QI is beneficial.10,11 High cost with poor outcomes inevitably means low cost-effectiveness. The QI industry has no requirement to demonstrate cost-effectiveness (as is mandatory in other areas), and there is no justification for this. A detailed study was unable to show cost-effectiveness of QI or that it produces net savings.12 QI is largely process based. Outcomes are measured using methods inferior to those providing the highest level of clinical evidence. The methods are appealing because the end points are easily measured quickly, using small samples,13 but the data suffer from bias, unintended consequences and lack relevance to clinical outcomes.10,13 There is a division between QI and clinicians, for whom the main threat to patient safety lies in budget constraints restricting access to proven treatments, and for whom improved clinical outcomes derive from applying clinical trial evidence. Involvement in QI projects, in addition to that required for registration purposes, consumes time and effort for little tangible gain. In summary, clinical QI needs to be reviewed because it is possible that it involves substantial expenditure for little return, and thus compromises clinical outcomes. This review should include the methods and benefits of hospital accreditation. QI projects and policies should be conditional on a satisfactory prospective business case, based on clinical end points. National and state bodies need to delegate QI efforts to individual hospitals, where local problems can be owned by those most immediately affected.

J Alasdair Millar PhD, FRACP, FRCP