General practice research: in the big league at last?

Medical research, like football, can be played many different ways, and in general practice the code of choice is definitely the survey! A flick through the research papers published in the Medical Journal of Australia between 1997 and 2001 will reveal that nearly half of those with GPs as primary authors were based on questionnaire surveys, compared with nearly a fifth of papers authored by physicians and about a quarter of those by surgeons. This phenomenon is not resticted to this Journal — a review of the medical literature between 1980 and 1999 showed that 41% of Australian general practice research was purely descriptive, comprising mainly surveys of GPs' views.1

Why this preponderance of surveys? One of the reasons is historical. From 1990 to 1999, after the Commonwealth Government introduced structural changes to general practice, the government-funded General Practice Evaluation Program (GPEP) allocated over $13 million as the main source of funding for general practice research.2 The criteria for receiving and the short-term nature of the funding favoured descriptive studies such as surveys. Thus, 90% of all 248 GPEP projects funded between 1990 and 1999 were descriptive studies, while 13% were interventional studies.2

Other reasons relate to the nature of general practice itself. Research expertise and capacity are less evolved in general practice than in most other disciplines. Moreover, general practice does not lend itself easily to the methods of randomised controlled trials.3 Our patients often have problems that are multiple, multidimensional (with psychosocial as well as biological components), and undifferentiated, lacking set diagnostic criteria. Our interventions may be correspondingly complex, often involving non-drug therapies.3

A (very appropriate!) survey of British GPs supports the need for relevant research: at least 61% of GPs surveyed considered that clinical topics, such as dealing with chronic or acute illness, patient behaviour and treatment, should be priority areas for general practice research.4 In this issue of the Journal (page 74), Queensland GPs echo this belief.5 Yet, the research being performed does not reflect this "wish list". More than half of all GPEP-funded projects focused predominantly on service organisation and supply, education, training and research methods.6 While these are important issues to tackle, how immediately relevant are they to the nucleus of general practice activity, the doctor–patient encounter?

In December 2001, general practice research funding appeared to join the big league. A new program of National Health and Medical Research Council Project Grants to support primary healthcare research arose as part of the Primary Health Care Research, Evaluation and Development (PHC-RED) Strategy funded by the Commonwealth Government (with $50 million over five years), replacing GPEP.7 However, it is still unclear how much of this will be spent on actual research projects. Results of the assessment of grant applications for the new funding (to start in 2003) are not yet available, and our attempts to determine the exact funding allocation from government sources have hit a bureaucratic wall. It seems GP researchers, like football teams, are at the mercy of the sponsors, who hold the purse strings.

Also at question is the extent to which the new program will meet the research needs of general practice. Priorities have been set for the allocation of the grants (Box), but many of these appear to reflect existing government policy rather than the priorities of "bag-carrying" GPs. The PHC-RED standing committee of the ministerial advisory body, the General Practice Partnership Advisory Council, has recommended that the next stage of the priority-setting process for the PHC-RED Strategy should have more emphasis on clinical research (Professor Max Kamien, Department of General Practice, University of Western Australia, personal communication). We would argue that, if the ultimate purpose of medical research is to improve the health of the population, such research must:

• pose research questions relevant to daily practice; and

• use appropriate and rigorous study methods to answer these questions.

How can general practice research meet these criteria? A day's consultation may bring three patients with vertigo, but different histories, social contexts, values and expectations, and not much by way of physical signs. Few data exist on the predictive value of particular clinical features in the general practice population, particularly those features which are "red flags" for serious conditions, warranting early intervention.9 We need research that documents the natural history of the undifferentiated presentations in general practice. This may be achieved, for instance, through cohort or case–control studies; the large samples required should be possible through collaborations involving Divisions and research networks, as espoused by Gunn10 (page 63) and van Weel11 (page 62) in this issue of the Journal. The process will be facilitated by the increased computerisation of general practice, allowing large-scale data collection.

In assessing interventions, observational studies,12 cluster randomisation (where the unit of randomisation may be the practice, not the patient)13 and pragmatic trials (that measure the effectiveness of a treatment in routine practice rather than explanatory trials that measure treatment efficacy in ideal conditions)14 may also have a place. NHMRC assessors of applications for the new primary-care grants will need to avoid perpetuating poorer-quality research (as in many questionnaire surveys), yet appreciate that methods that are rigorous and appropriate to general practice are still evolving.15

Not all GPs will want to be active researchers, but there should be funding and infrastructure to train and support those who do. This forms an integral part of the PHC-RED Strategy (for instance, through the new NHMRC fellowships and scholarships and the Researcher Development Program),7 and should build on the recommendations of the General Practice Strategy Review Group in 1998,16 including master's or doctoral programs, dedicated career pathways, research "mentors", and research networks. Practical issues, such as adequate remuneration and locum provision while GPs take time off from practice for research, also need to be managed.5

GPs need to be engaged in research throughout their careers, with fostering of critical appraisal and research skills from medical school onward. Clinical audits (now a requirement of the Royal Australian College of General Practitioners Quality Assurance and Continuing Medical Education Program17), coupled with the possibilities of information technology, will also encourage us to evaluate our practice and refine our craft, even if they do not constitute "true" research.

General practice research may have reached the big league, but let's not forget the endgame. The point of promoting the evidence base for general practice is to improve the health of our patients. Increasing general practice research culture and capacity is crucial to this process, but should not be an end in itself. The gap between evidence and practice is more likely to be bridged if the gap between researcher and clinician is bridged, with greater ownership of the research agenda by clinicians5,18 rather than government committees. Whether the stated priorities of the new Project Grants will be discordant with clinical priorities remains to be seen. GPs are, however, pragmatists at heart: we will be more likely to practise with reference to an evidence base if the research underlying it is directly relevant to our practice. Our patients expect no less and neither should we.