To the Editor: In 1997 an analysis of the medical workforce in North Queensland showed that Townsville's 125 000 people were served by about 180 GPs providing about 120 full-time-equivalent (FTE) GP workloads. A doctor–population ratio (DPR) of about 1: 1000 was evidence of some sort that the community was reasonably well served, at least according to benchmarks of the time. As a result, the "area of need" status was removed from many North Queensland centres, a decision reinforced by similar findings from the Australian Medical Workforce Advisory Committee (AMWAC) report released soon after.1 Further, the AMWAC report was the basis of a decision not to increase the number of GP vocational training places in the region, even though there is spare capacity and the program is one of the more successful in terms of retention of rural GPs.2

Despite persistent claims to the present time that there is no shortage of GPs in regional centres, in 2002 the number of FTE GPs in this community appears to have fallen to about 95, despite strong population growth to about 150 000 people, resulting in a DPR of around 1: 1500. The city's two extended-hours clinics have closed, very few practices direct bill, new residents have trouble getting an appointment in any general practice, and fewer GPs are providing after-hours care. The recent Access Economics report, commissioned by the AMA, indicates that this trend is evident elsewhere.

Anecdotal evidence suggests that it is not necessarily the total number of GPs that is changing, but rather their work patterns. The increasing proportion of female graduates probably reduces the available FTE workforce, and more male graduates are now opting for a lifestyle that better balances clinical work with family responsibilities and interests outside of medicine.

Medical workforce research is an interesting, yet risky, academic business, beset by many complex issues relating to definitions, data sources and the measurement and interpretation of DPRs.2 Despite our best endeavours and close proximity, the report conducted by me and my colleagues in 1997 was almost certainly incorrect within a very short time. I am inclined to think that GPs have a better sense of what is happening on the ground than do the sifters of data. The biggest challenge to workforce analyses, and therefore to patient access to GP care, may be the societal changes in work patterns, not in raw numbers, and these issues need to be better understood if we are to make progress in managing workforce issues.