Diabetes registers: a grassroots approach.

To compile a register of diabetic patients within the catchment area of a district general hospital and evaluate the characteristics of the population using aggregated data from a general practice audit. Cross sectional study. Practices identified all known diabetics and completed a questionnaire from information in each patient's medical record. Practices affiliated to a district audit group in south east England. Number of participating practices; prevalence of diabetes and its complications; and sex distribution of patients, age at diagnosis and review, treatment, and metabolic control. Comparisons with similar data from other studies. 41 out of 43 practices participated, and 2574 diabetic patients were identified (prevalence 1.18%). 52.4% of patients were male. The mean age was 61.6 years. 32% of patients were treated with insulin, 51.5% with oral hypoglycaemic agents, 16.5% with diet alone. The mean random blood glucose concentration was 10.4 mmol/l and glycosylated haemoglobin 10.1%. 8% had proteinuria, 7% a history of myocardial infarction, 5% a history of stroke, and 2% a diabetes related amputation. These proportions were not significantly different from those found in studies performed by different methods in Poole, Islington, Powys, Trowbridge, and Southall. It is feasible to compile a register of diabetic patients in a district and evaluate their characteristics by using only general practice sources.