Introduction to Longitudinal Research in General Practice

Abstract

General practice should be studied in a way which appropriately reflects its basic qualities. One fundamental characteristic of patient care in general practice is the continuity of the doctor (practice)-patient relationship. In order to take full account of continuity of care processes and outcomes of patient care should be analysed longitudinally. Longitudinal research in general practice ranges from studies of a single or a few diseases to studies of the whole spectrum of health problems in general practice. Longitudinal research can provide valuable epidemiological information about the course over time of common health problems in general practice and essential insights into the provision of care for the problems which are presented. The recording and analysis of patients’ reasons for seeking care, doctors’ diagnoses and the activities of care in relation to a succession of doctor-patient contacts for the same health problem raises many practical and methodological problems. The concept of an episode has been introduced for structuring the registration and analysis of medical care. However the concept of a general practice episode of care is still ambiguous and needs refinement if it is to become a predominant tool in clinical practice and research. Sometimes the analysis of data from longitudinal studies of a single disease is hampered by lack of information on co-morbidity. Global recording of the content of all doctor-patient contacts in the practice population may, to some extent, solve the problem. However, even the recording of all contacts in general practice (i.e. general practice care) will still miss those concurrent external interventions (e.g. hospital care) which are beyond the direct influence of the practice but of importance in total patient care. Again this reminds us to state clearly the purpose of studies, and to remember the limits of the inference which can be drawn from them. General practice studies of health problems and care based on large scale recording are still lacking in most countries. With the dissemination of computers in general practice it is tempting to compile and subsequently analyse quantities of data from patient records in general practice. Experience tells us that the validity and reliability of such data is questionable. Instead of choosing (easy) ways of data collection academic general practice should develop primary care classifications and instruments for reliable longitudinal registration of problems and activity in general practice. If such instruments are disseminated on a large scale they should be a considerable help in the daily work of general practitioners. For example, managing information on each patient over time can be a problem when the electronic record has been in use for some years. Overview of problems and care can be lost as the number of screens filled with case-note text increases. If a registration instrument intended for research can also be a help in daily practice its introduction into existing practice computer systems will undoubtably be much easier. Most researchers feel the need for including patient perspectives in longitudinal studies of general practice care. Consequently there is a need to link general practice epidemiology with consultation research (doctor/patient interaction). No well established methodologies exist and they should be developed within the discipline itself with strong input from non medical researchers. The transition from cross-sectional studies to longitudinal studies in general practice is now taking off. Let us hope that we will witness a strong international development of methodologies in the next few years.