A population-based diabetes register--development & applications.

Population-based registers for chronic disorders are invaluable for better understanding the epidemiology, aetiology and natural history of a particular disorder. Insulin-dependent diabetes mellitus is one example where registers have been used. The disorder often develops at a relatively young age continuing throughout life with adverse influences on health, frequently causing premature mortality. It is well suited to systematised data collection and register development. In this paper, the rationale for registers is considered and methods used in development of a diabetes register in the Canterbury region of New Zealand (pop. one-third of a million) are described. A preliminary analysis of data using prevalence date of 1 January 1984, has shown that there are 1,148 (M = 574, F = 574) insulin-treated diabetic persons on the register. Ages range from two to 93 years (median = 52 years). Only 11.1 per cent of the diabetic population were aged under 20 years whereas 27 per cent were 65 or more years of age. Duration of diabetes was one to 58 years (median = 12 years). Only 28 per cent of cases had presented with diabetes at age under 20 years, most being diagnosed in adult life. The overall prevalence of insulin-treated diabetes was 3.3 per 1,000 population and peak prevalence was observed in those aged 60–69 years (7.5 per 1,000). Ongoing research with this register will document the natural history of the disorder, its morbidity and mortality, and will measure the use of secondary health care resources, particularly in-patient services.