What can universities do to reverse the decline in the numbers of doctors entering general practice?

For the past 4 years primary care specialties have been the most popular career choice of medical graduates in the United States (Roberts 1995). This follows a long period when worries were expressed about the paucity of applications for such positions (Coulson 1996). Over the same time period, numbers of applications to general practice training schemes and partnership vacancies in the United Kingdom have dwindled (Vaughn 1995; Wilson et al. 1995). Postgraduate specialty experiences, personal and economic factors are all important determinants of individual career choice (Hays 1993), but the undergraduate curriculum remains an important influence (NHS Executive 1995). Over the past decade, universities in the United States and Canada have accepted a lead role in attempting to direct undergraduates towards primary care with many initiatives designed specifically for that purpose. These schemes have had variable success rates. Are any of the lessons learned by these medical schools applicable to universities in countries which, like the UK, may be considering measures to increase the supply of medical graduates to staff a primary care led National Health Service? This is a difficult area to analyse as no consensus on the appropriate methodology exists. Dohn recently summarized the possible approaches (Dohn 1996) as: traditional review; a focus on specific issues/specialties; hypothesis testing or meta-analysis. He also outlined the advantages and disadvantages of these approaches and divided the literature into three historical phases which have examined personal characteristics (1950–70), medical education (1970-85) and experience of postgraduate pay and conditions (1985–1995). A recent study (Bland et al. 1995) accepted the difficulties of applying a formal meta-analysis described by Dohn and instead used a non-statistical method to appraise the literature until 1994. This paper focused on choice of primary care as a career. This found 307 references in Medline, Psychinfo, Sociology abstracts, ERIC and Dissertation abstracts. Seventy-three of these papers had sufficiently high methodological and validity scores to be included in the detailed analysis, with the most useful studies appearing since 1987. A similar search strategy conducted by the authors reviewing the literature to June 1996 failed to add significant new information. The meta-analysis found that the majority of students start their medical course with an orientation towards primary care. This is especially true for students who are older, female or married and those with a broad range of life and educational experiences before they enter medical school. The specific features of curricula which were found to be likely to increase the numbers of students entering primary care were mandatory family practice attachments and primary care experiences lasting 6 weeks or more. Other strategies such as early exposure to general practice attachments or employing large numbers of general practice academic staff in universities were not shown to have any effect. So how does the current practice of UK medical schools compare with these findings? Recent studies of departments of general practice (Association of Teachers of General Practice 1990; Robinson et al. 1994) provide an insight. Almost all universities now provide a compulsory attachment to general practice but in many it is only 2–4 weeks. This is likely to prove inadequate in promoting an orientation to primary care. Recent pressures from the General Medical Council (GMC 1993) and support from revised funding arrangements (Strategic Increment For Teaching (SIFT) in the UK) may address this problem by encouraging medical schools to provide a longer period of experience in general practice. In addition, making the attachment to general practice an explicit career experience may help. In Glasgow both students and their tutors asked for one of the objectives of the 4 week final year attachments to be “assessment of general practice as a career”, somewhat to the surprise of the academic department (Morrison & Murray 1996). Information about admissions policies in the UK is often restricted to members of the admissions panel although universities are increasingly making this a transparent process (Ismail et al. 1995). Admissions officers who wish to ensure adequate numbers of primary care doctors in years to come might note the profile of the students more likely to enter primary care described above, and select a wider range of entrants than the predominantly young with science backgrounds. For example, they might encourage younger applicants to spend a ‘gap year’ elsewhere before entering university to provide them with a wider perspective than their recent school experiences. Since broader characteristics also correlate with other desirable attributes of entrants (high course completion rates and lower litigation rates after qualification) this method of selection has additional advantages (Rolfe et al. 1995). The recommendations to medical schools who wish to reverse the decline in numbers of doctors entering general practice based upon the existing literature are to broaden the range of entrants, make general practice attachments mandatory and at least 6 weeks long. Finally, given the paucity of recent research to explain the decline in entrants to general practice we would suggest that this area should be given a priority by researchers. We should like to thank Felicity Grainger, Librarian, Glasgow University for helping us perform the literature search upon which this article is based.