General practice postal surveys: a questionnaire too far?BMJ 1996; 313 doi: http://dx.doi.org/10.1136/bmj.313.7059.732 (Published 21 September 1996) Cite this as: BMJ 1996;313:732
- a Department of Primary Health Care, School of Health Sciences, Medical School, University of Newcastle upon Tyne, Newcastle upon Tyne NE2 4HH
- Correspondence to: Dr Kaner.
- Accepted 29 July 1996
A primary care led NHS, driven by evidence based practice, needs to build on a firm foundation of research in primary care. As researchers are making increasing use of questionnaire surveys to assess general practitioners' views and attitudes, so response rates to questionnaire surveys among general practitioners are dropping. The reasons include lack of perceived relevance of the research and lack of information and feedback about it, and researchers need to be more aware of the realities of everyday practice. Approaches that might reverse this trend include monitoring all research activities going on in an area to ensure that practices are not overused, giving general practitioners incentives to participate, and improving the relevance of research and the quality of questionnaires.
A strong research culture in general practice is necessary to enable general practitioners and primary health care teams to meet the evolving medical needs of their practice populations.1 In the new primary health care led NHS, with the focus on evidence based practice, there is an increased need for research and development. As the balance continues to shift from hospital care to primary health care and community services, the search for information from general practitioners is likely to continue to increase.2 Yet, just when general practice has its greatest opportunity to seize the high ground of policy, there is widespread concern over job satisfaction, morale, autonomy, workload, bureaucracy, recruitment, and retention.3
For reasons of costs and generalisability of research findings, large scale postal surveys of general practitioners are often carried out to obtain information and to ascertain attitudes about current health issues. Many studies, however, testify to the fact that general practitioners are poor responders to such surveys.4 5 6 7 Indeed, as early as 1978 a study of professional responses to health studies between 1961 and 1977 reported that probably the most worrying finding was the drop in response rate over time from doctors, particularly general practitioners.4 Recent evidence suggests that this trend is continuing.8
A postal survey in 1995-6 of general practitioners in the Midlands produced an initial response rate of 32% (N Heather, personal communication). Telephone follow up of the non-respondents revealed that 25% felt they were too busy to complete research questionnaires and another 13% never answered questionnaires or routinely threw them away. One practitioner returned the uncompleted questionnaire in its prepaid envelope, enclosing an invoice for £5 to cover his “administrative costs.” This may become a more common phenomenon in our market driven health service and would further increase the costs of medical research.
Why general practitioners do not respond
Studies which have investigated general practitioners who do not respond to postal surveys have characterised them as older, more experienced, less well qualified, and often singlehanded4 and possibly those who feel under more stress.7 If these are the general practitioners who tend not to respond to surveys then surely a valuable body of expertise is being missed in reported surveys, aside from the specific issue of generalisability of findings to the whole population of general practitioners.
There are many reasons why general practitioners do not respond to surveys. These include being swamped by the volume of questionnaires arriving on their desks2 4 5; resenting interference in their activities by outside researchers,9 particularly academic general practitioners10 11 12; the length of questionnaires4 or the time taken to participate in research13; having qualms about encroachment on patients' or their own confidentiality14 or more generally being opposed to research methods used15; being uninterested in or disliking the topic being researched or the threatening nature of some subjects4; being insufficiently involved in the subject being researched16; fearing that research activities disrupt the general practice workload14; and lack of financial incentives.13
There is also strong evidence that general practitioners are reluctant to participate in a survey if they have not been given adequate information beforehand.17 18 19 They are also less likely to participate in future surveys if they are given insufficient feedback.2 17 19 The lack of feedback may have led to general practitioners' increasing cynicism about the impact of their views on service planning and provision.20 Since the 1990 contract general practitioners have been found to have an increased workload,21 which has had effects upon their stress levels, job satisfaction, and mental health.22
Ways of encouraging participation
How can we encourage general practitioners to participate in research? Murphy et al outline that careful and appropriate negotiation of agreement with identified stakeholders may increase response rates and enhance data quality.18 One suggestion to improve general practitioner recruitment into surveys is to have personal contact between general practitioners and researchers either by telephone or a personal visit.9 Although having a known and respected medical colleague on the research team can increase a study's legitimacy (or credibility) and boost response rates,23 Ward found that general practitioners were less influenced by names and the track record of the research team than by the research topic itself and personal approaches used to obtain consent.19 Assurances about confidentiality, both for patients and for the general practitioner, and about the maintenance of general practitioner autonomy are also important.23 Lastly, Ward reported that general practitioners need substantially more information than they usually get before a researcher can engage their help and they also need to be given feedback on the outcomes of studies in which they have participated.19
One of the most exciting developments in recent years has been the establishment of research networks in primary care. At least a dozen now exist throughout the country, some formally constituted, like the Northern and Wessex research networks (NoReN and WReN), and others which consist of practitioners who have a long tradition of meeting for support and to exchange ideas. Whether they operate formally or informally, they are developing a research culture and commitment to evidence based research, dissolving the boundaries of research and practice.24
Researchers need to understand the pressures on general practitioners and question how much research material they send out. In Scotland there have been moves to create a database of research to inform those about to launch a survey which general practitioners are already or have recently been the subjects of surveys.2 This approach may go some way to reducing the sheer number of questionnaires that general practitioners receive each week. Moreover, several university departments of general practice and primary health care have policies of monitoring the practices they include in studies to avoid overusing particular practices.
If the government wants general practitioners to carry out and participate in research it should acknowledge its value and help general practitioners to set aside time for research activities. Accreditation (and payment) exists for education and training activities in general practice, but no such incentive exists for participation in research. Howie has commented that “If research is to thrive there must be a climate of opinion in which research is an expected, valued and rewarded activity.”25 General practitioners need tangible incentives to participate in research.
As general practitioners develop critical appraisal skills for evaluating journal articles, they should become better able to differentiate good from poor questionnaires. Providing them with guidance on this and developing a simple scoring system for rating the quality of questionnaires would be positive steps researchers could take, perhaps with the encouragement of the Royal College of General Practitioners.
Finally, it needs to be said that the routine discarding of questionnaires without answering them may lead to a weakening of general practitioners' power to influence service planning and provision. For as the planners and purchasers begin to look for alternative sources of information in primary health care, general practitioners may find that their voice is being heard less over the enthusiastic clamour of other members of the primary health care team and patients armed with the Patient's Charter.
Funding EFSK is supported by a grant from the Alcohol Education and Research Council.
Conflict of interest None.