Ethnic differences in cause specific mortality among hospitalised patients with diabetes: a linkage study in New Zealand.

STUDY OBJECTIVE: To describe patterns of excess mortality among patients with diabetes in three ethnic groups. DESIGN: A linkage study of national hospital discharge records to death records. SETTING: New Zealand. PARTICIPANTS: The study included 74 847 patients (11,268 Māori, 5730 Pacific, and 57,849 non-Māori/non-Pacific) aged over 25 years with a hospital discharge diagnosis of diabetes between 1988 and 2001. By the end of 2001, 29,295 (39%) of the cohort had died. Based on the underlying cause of death, standardised mortality ratios (SMRs) (95% confidence intervals) were calculated for each ethnic group and sex. MAIN RESULTS: Comparing the mortality patterns of patients with diabetes to the general population of the same ethnic group, adjusting for age and calendar period, all cause SMRs were higher for Māori women and men: 3.80 (95% CI: 3.64 to 3.97) and 3.44 (95%CI: 3.30 to 3.58) than for Pacific (men: 2.41 (95%CI: 2.21 to 2.61); women: 2.23 (95%CI: 2.06 to 2.41)) and non-Māori/non-Pacific (men: 2.98 (95%CI: 2.93 to 3.04); women: 2.99 (95%CI: 2.93 to 3.04)) people. SMRs were significantly raised for several causes of death, including cardiovascular disease and many site specific cancers. CONCLUSIONS: The pattern of excess mortality among Māori with diabetes may relate to severity of disease. This needs further investigation, as the excess mortality may be amenable to intervention.