Clinical Presentation and Outcomes of Diabetic Ketoacidosis in Pregnancy

OBJECTIVE: To examine the presentation, management, and outcomes of pregnancies complicated by diabetic ketoacidosis (DKA) in a contemporary obstetric population. METHODS: This is a case series of all admissions for DKA during pregnancy at a single Midwestern academic medical center over a 10-year period. Diabetic ketoacidosis was defined per the following diagnostic criteria: anion gap more than 12 mEq/L, pH less than 7.30 or bicarbonate less than 15 mEq/L, and elevated serum or urine ketones. Demographic information, clinical characteristics, and maternal and neonatal outcomes were assessed. Patient characteristics and clinical outcomes were compared between individuals with type 1 and those with type 2 diabetes mellitus. RESULTS: Between 2012 and 2021, there were 129 admissions for DKA in 103 pregnancies in 97 individuals. Most individuals (n=75, 77.3%) admitted for DKA during pregnancy had type 1 diabetes. The majority of admissions occurred in the third trimester (median gestational age 29 3/7 weeks). The most common precipitating factors were vomiting or gastrointestinal illness (38.0%), infection (25.6%), and insulin nonadherence (20.9%). Median glucose on admission was 252 mg/dL (interquartile range 181-343 mg/dL), and 21 patients (17.6%) were admitted with euglycemic DKA. Fifteen admissions (11.6%) were to the intensive care unit. Pregnancy loss was diagnosed during admission in six individuals (6.3%, 95% CI, 2.3-13.7%). Among pregnant individuals with at least one admission for DKA, the median gestational age at delivery was 34 6/7 weeks (interquartile range 33 2/7-36 3/7 weeks). Most neonates (85.7%, 95% CI, 76.8-92.2%) were admitted to the neonatal intensive care unit and required treatment for hypoglycemia. The cesarean delivery rate was 71.9%. Despite similar hemoglobin A 1C values before pregnancy and at admission, individuals with type 1 diabetes had higher serum glucose (median [interquartile range], 256 mg/dL [181-353 mg/dL] vs 216 mg/dL [136-258 mg/dL], P =.04) and higher serum ketones (3.78 mg/dL [2.13-5.50 mg/dL] vs 2.56 mg/dL [0.81-4.69 mg/dL] mg/dL, P =.03) on admission compared with those with type 2 diabetes. Individuals with type 2 diabetes required intravenous insulin therapy for a longer duration (55 hours [29.5-91.5 hours] vs 27 hours [19-38 hours], P =.004) and were hospitalized longer (5 days [4-9 days] vs 4 days [3-6 days], P =.004). CONCLUSION: Diabetic ketoacidosis occurred predominantly in pregnancies affected by type 1 diabetes. Individuals with type 1 diabetes presented with greater DKA severity but achieved clinical resolution more rapidly than those with type 2 diabetes. These results may provide a starting point for the development of interventions to decrease maternal and neonatal morbidity related to DKA in the modern obstetric population.