Diabetes and pregnancy: National trends over a 15 year period

Aims: To examine time trends in national perinatal outcomes in pregnancies complicated by pre-existing type 1 (T1DM) or type 2 (T2DM) diabetes. Methods: We analysed episode level data on all obstetric inpatient delivery events (live or stillbirth) between 1st April 1998 to 31st March 2013 (n=813,777). Mothers with T1DM (n=3229) and T2DM (n=1454) were identified from the national diabetes database (Scottish Care Information-Diabetes) and perinatal outcomes compared in T1DM, T2DM and those without diabetes (non-DM). Results: Numbers of pregnancies complicated by diabetes increased significantly by 44% in T1DM and 90% in T2DM across the 15 years examined, to rates of 1 in 210 and 1 in 504 deliveries respectively. Compared to non-DM, delivery occurred 2.6 weeks earlier (T1DM, 36.7± 2.3 weeks) and 2 weeks earlier (T2DM, 37.3± 2.4 weeks), reducing significantly in both T1DM (from 36.7 weeks to 36.4 weeks: p=0.03) and T2DM (38 weeks to 37.1 weeks: p<0.001) across the time period. Proportions of preterm delivery were markedly increased in women with diabetes (35.3% T1DM; 21.9% T2DM; 6.1% non-DM: p<0.0001) and these proportions increased with time for both groups (p<0.005). Proportions of elective (ELCS: 29.4% T1DM; 30.5% T2DM; 9.6% non-DM) and emergency (EMCS: 38.3% T1DM; 29.1% T2DM; 14.6% non-DM) caesarean section were greatly increased in women with diabetes and increased over time apart from stable rates of EMCS in T1DM. Gestational age, sex and parity adjusted Z-score birthweight were higher in T1DM (1.33 + 1.34: p<0.001) and increased over time from 1.22 to 1.47 (p<0.001). Birthweight was also increased in T2DM (0.92+ 1.33: p<0.001) but did not alter with time. There were 64 perinatal deaths in offspring of mothers with T1DM and 38 to mothers with T2DM, representing perinatal mortality rates of 19.6 [15.1-25] and 25.8 [18.2-35.4] per 1000 births respectively and rates 3.3 and 4.3 times those observed in the background population (p<0.001). Stillbirth rates were 3.6 fold higher in mothers with T1DM and 5.3 fold higher in mothers with T2DM (p<0.001). Both perinatal mortality and stillbirth rates showed no significant fall over time despite small falls in the background population. Conclusion: Women with diabetes are having increased intervention in pregnancy (earlier delivery, increased caesarean section rates) but despite this higher birth weight. Improvements in rates of stillbirth seen in the general population are not being reflected in changes in stillbirth or perinatal mortality in our population with diabetes.