Introduction:
Pregnancy increases the risk of diabetic ketoacidosis (DKA) due to enhanced lipolysis and ketone body production in women with pre-existing diabetes. DKA is life-threatening to the fetus, with reported demise rates of 15-60%, while maternal death is rare (1). There are no published case series on the triggers, and outcomes of DKA during pregnancy in Australia.
Aim: To review the pregnancy and neonatal outcomes in women with type one diabetes (T1DM) who develop DKA in pregnancy.
Methods: We performed a retrospective observational cohort study over ten years from January 2013 to January 2023 of pregnant women with T1DM admitted to Monash Health, a service delivering over 10,000 babies per year. Women were identified using ICD-coding and data were extracted from Electronic Medical Records (EMR), Scanned Medical Records (SMR) and the Birthing Outcomes System (BOS). The diagnosis of DKA was made clinically and biochemically (pH < 7.3, base excess >12, bicarbonate <15mmol/L ketones ≥0.6mmol/L).
Results: Over the 10 year period, 31 cases were identified as possible DKA. Only 24 satisfied criteria, with 7 diagnosed as starvation ketosis. At baseline 16 women (66.7%) were managed with basal bolus insulin with 5 women (20.8%) managed via continuous subcutaneous insulin infusion (CSII). At presentation with DKA, mean HbA1c was 8.4%, glucose 19mmol and ketones 3.8mmol/L. Gestation ranged from 5-36 weeks (mean 25, SD 9). The common triggers for DKA were nausea and vomiting n=8 (33.3%) and intercurrent illness n=7 (29.2%). DKA was treated with a standard insulin infusion in 20 women (83.3%) and five women (20.8%) required ICU admission. Notably, there were no maternal or fetal fatalities. Mean gestation at delivery was 34.9+/-2.3 weeks, one neonate (4.2%) was delivered due to DKA at 36 weeks gestation.
Conclusion:
This study highlights that while women with T1DM constitute a minority birthing at a large public health service, clinicians should be alert to their heightened risk of DKA during pregnancy. DKA precipitated delivery in only one case and there were no cases of fetal demise, in contrast to the rates suggested by previous literature. Factors that may have contributed to these findings include management in a specialist tertiary obstetric/diabetes pregnancy service, routine review with diabetes education focusing on sick day plans, early recognition and intervention, widespread use of continuous glucose monitoring and CSII. Women with T1DM should receive pre-pregnancy ketone and sick day management advice, with nausea, vomiting, and intercurrent illness emphasised as common triggers.