Developing a district diabetic register.

To compile a district wide diabetic register of all diabetic patients registered with general practitioners in the catchment area of a single district general hospital and to compare different approaches to identifying patients. Information for a register was obtained from general practitioners' practice registers, the Prescription Pricing Authority, and hospital diabetic clinic records. Catchment area of an inner London district general hospital with a large diabetic clinic. All patients with a diagnosis of diabetes resident in or attending general practitioners or hospital clinics in the district or its catchment area. Prevalence of diabetes, population of patients elicited by different approaches, proportion attending the local district general hospital, cost of using prescription returns for identifying diabetic patients. 4674 patients with diabetes were identified from all sources of information, which corresponds to a mean of 22.4 patients per general practitioner and the prevalence of known diabetes of 1.17%. 39.4% of patients identified had Prescription Pricing Authority returns and 42.8% of patients appeared on practices' diabetic registers. Only 56.5% of patients identified attended the district general hospital. For practices where all sources of information were available, practice registers included 60.4% of all patients, and prescription returns and the clinic register identified 64.9% and 40.6% respectively. The cost of using prescription returns to identify patients not detected in other ways was 6.37 pounds per patient. The task of developing district diabetic registers may prove, even in one cross sectional attempt, a major task in many inner city health districts.