The proportion of death certificates of diabetic patients that mentioned diabetes in Osaka District, Japan

To determine the accuracy of the stated causes of death in diabetic patients, we analyzed 503 death certificates of diabetic patients who had been followed up for 9.4 years on average at Osaka Medical Center for Adulthood Diseases and had died during the observation period from 1960 to 1984. Of the certificates examined, only 213 (42.4%) mentioned diabetes, either as the underlying cause or as a contributory condition. The proportion of certificates that mentioned diabetes was related to the underlying cause of death; it was 18.7% for malignant neoplasms, 40.5% for cerebrovascular disease and 46.9% for disease of the heart, while it was 81.8% for renal disease. The proportion was higher for females (48.7%) than for males (39.5%), and a significant difference was observed between the sexes in the proportion in the case of disease of the heart. The proportion was not related to any significant extent to the year of death or the age at death. However, it was related to fasting plasma glucose (FPG) level and type of treatment. It was 37.8% and 54.4% for patients with FPG < 200 mg/dl and FPG ≥ 200 mg/dl, respectively, 24.1% for those treated with diet, 51.4% for those treated with oral hypoglycemic agent and 80.7% for those treated with insulin. As a result, it was found that the proportion of death certificates that mentioned diabetes was related to the accompanying complications and the severity of diabetes. Based on these results, the actual number of deaths of diabetic patients was estimated to be at least 6.4 times higher than the number appearing in mortality statistics, since diabetes as the underlying causes of death accounted for 36.8% of all certificates that mentioned diabetes in our previous study in Osaka. This value corresponds to 42.4% of the observed number of deaths of diabetic patients in the present study.