Regional population-based study on pregnancy outcomes in women with diabetes mellitus in Japan

Aim: We compared pregnancy outcomes in diabetic women with the background population in Miyazaki, Japan. Methods: In 1998, we started the regional, population-based, peer-review conference to investigate the possible causes of perinatal deaths. For this purpose, at least one obstetrician and one neonatologist congregated from each institution (one tertiary and five secondary perinatal centers). A retrospective, population-based study was conducted in a total of 53 862 deliveries during 1999-2003. Among them, there were 248 perinatal deaths including six deaths in association with diabetes mellitus (DM). The number of diabetic pregnancies was estimated to be 381 during the study period. Perinatal mortality was compared between the diabetic women and background population by chi(2)-test. Results: The perinatal mortality rate in diabetic women was 15.7 per 1000 deliveries, which was compared with 4.6 per 1000 in the background population (P = 0.003; odds ratio: 3.5; 95% confidence interval: 1.5-7.9). Four of the six perinatal deaths in diabetes were sudden intrauterine demises after 37 weeks' gestation, and the others were neonatal deaths of congenital heart anomaly or extreme prematurity. Most deaths were attributable to either undiagnosed or insufficient perinatal management. Conclusions: In an unselected population in Japan, diabetic women have 3.5 times the reported risk of perinatal mortality of the general population. Further improvements in the diagnosis and management of DM during pregnancy are required.