The well recognised global shortage of doctors1 reinforces the need for effective medical workforce planning. Simplistically, medical workforce planning aims to ensure an adequate supply of doctors with the appropriate skills to meet the current and projected future medical care needs of the population.2 This requires accurate information about the current medical workforce,3 but Australia lacks comprehensive, timely, nationally consistent and longitudinally linked data.3,4

The principal source of Australian medical workforce data is the annual Medical Labour Force Survey, administered by each state and territory medical board on behalf of the Australian Institute of Health and Welfare as part of the annual medical registration renewal process. The survey provides data on the number of medical practitioners, their age and sex, the type of work they do, their specialties, and the hours they work.4 However, the utility of these data is limited, as they are neither timely nor comprehensive. The latest workforce report, published in 2008, reports on the 2006 survey. Response rates ranged from 80% in Queensland to 29% in the Northern Territory, although no survey data were received from the NT in 2006, necessitating estimates based on responses to the 2007 survey and weighted to 2006 benchmark figures.4 The type of practitioners surveyed varies between jurisdictions, with Queensland surveying only general registrants and conditionally registered specialists, and Tasmania excluding conditional registrants who are overseas-trained doctors or interns. Furthermore, the current survey data are cross-sectional and not longitudinally linked, limiting the identification and monitoring of factors associated with trends in changing demographics of the medical workforce.2,3

National registration of the medical workforce and the associated National Minimum Data Set Project,5 scheduled for commencement in 2010,6 could be the ideal opportunity to move Australian medical workforce data collection into the 21st century. The introduction of web-based electronic data capture for both medical registration and workforce data would facilitate timely analysis and reporting. Making the workforce survey a compulsory component of registration (with a strong and defensible rationale for each requested data item) would enable development of a comprehensive national dataset. Additional voluntary-response questions could be included to explore the factors influencing work patterns identified in subsets of respondents. Linking the workforce data to each individual’s registration number (or another “unique identifier”5) would allow monitoring of flows into and out of the workforce,3 thereby facilitating projection of trends.

Expanding the workforce survey to include all general and conditionally registered medical practitioners and medical students would further enable informed modelling for workforce planning. Conditional registrations of overseas-trained doctors form a sizeable proportion of total registrations in some states (over 10% in Queensland and Western Australia in 20037) and should be included in any national data collection. The Australian Health Workforce Ministerial Council’s recent proposal to include students in the national register8 provides an avenue by which the Medical Schools Outcomes Database, which has collected demographic, educational and career-intention data from medical students across all Australian medical schools since 2006,9 could be longitudinally linked with graduate doctor data. The ways in which medical education factors (eg, school leaver versus undergraduate- and postgraduate-entry status, full fee-paying versus subsidised or bonded students, academic institution, and clinical placement experiences) influence career choices, retention and productivity could then be explored.

Such a proposal is especially important for general practice workforce planning. Australia does not have enough general practitioners, and this is unlikely to change unless more doctors choose a career in general practice.10 Suggested strategies to actively encourage students to consider working in general practice include longer and higher-quality general practice attachments during medical school and early postgraduate years, such as the Pre-vocational General Practice Placements Program.11 The impact of such initiatives, including the effect of increasing the educational demands placed on the GP supervisor workforce, could be monitored through a national, longitudinally linked electronic workforce survey.

Other trends affecting the supply of GPs could also be monitored. For example, the increasing feminisation of the general practice workforce necessitates greater understanding of the factors associated with female doctors’ work patterns.4 Overseas-trained doctors are relied on to provide general practice services in areas of workforce need,2 and it is important to identify the factors that influence their retention. An increasing number of GPs are restricting their clinical practice to areas of specific interest and expertise, effectively reducing the number of GPs providing undifferentiated first-contact clinical care. The actual numbers involved are not known, and while crude methods have been employed to estimate them,12 more data are required to inform not only workforce planners and policymakers, but also educational institutions seeking broad-based general practice exposure for increasing numbers of students, junior doctors and GP registrars.

We welcome the planned move to national medical registration in Australia in 2010 and the development of a national minimum dataset. We urge that the labour force survey be made a compulsory component of medical registration, with adaptability to explore identified trends, and that serious consideration be given to web-based collection of the data. This would facilitate a comprehensive, longitudinally linked national dataset for the timely analysis and reporting of medical workforce data — factors considered essential for informed policy development and evidence-based workforce planning.3