Variations in death certification practices distort international comparisons of mortality from diabetes.

Background: Israel is ranked second among OECD countries in diabetes mortality despite good performance on diabetes care measures. This study assessed whether variations in death certification practices could explain differences in diabetes mortality rates between countries, using a comparison between the USA and Israel as an example. Methods: Multiple cause mortality data for Israel and the USA were analyzed. The proportions of cases with diabetes coded as the underlying cause of death (UCOD), of all certificates with diabetes listed as one of the multiple causes of death (MCOD), were calculated by age-group, and compared between the USA and Israel, with emphasis on cases in which cardiovascular events were reported in part I of the certificate. Results: The diabetes UCOD/MCOD ratio was higher in Israel for all age groups. The differences in proportions were larger when cardiovascular events were reported in part I. Diabetes mortality rate ratio between the countries would be 49% lower if the UCOD/MCOD ratios in US data were applied to the Israeli data. Conclusions: Half of the difference in the reported diabetes mortality rate between the USA and Israel is explained by different coding practices. International comparisons could be improved by using multiple cause data or by clarifying guidelines regarding certification of diabetes deaths.