Comparison of healthcare utilization and outcomes by gestational diabetes diagnostic criteria

Objective: To compare healthcare utilization and outcomes using the Carpenter-Coustan (CC) criteria vs. the National Diabetes Data Group (NDDG) criteria for gestational diabetes mellitus (GDM). Methods: This is a retrospective cohort study. Prior to 8/21/2013, patients were classified as "GDM by CC" if they met criteria. After 8/21/2013, patients were classified as "GDM by NDDG" if they met criteria and "Meeting CC non-GDM" if they met CC, but failed to reach NDDG criteria. "Non-GDM" women did not meet any criteria for GDM. Records were reviewed after delivery. Results: There was a 41% reduction in GDM diagnosed using NDDG compared to CC (P = 0.01). There was no significant difference in triage visits, ultrasounds for growth or hospital admissions. Women classified as "Meeting CC non-GDM" were more likely to have preeclampsia than "GDM by CC" women [OR 11.11 (2.7, 50.0), P = 0.0006]. Newborns of mothers "Meeting CC non-GDM" were more likely to be admitted to neonatal intensive care units than "GDM by CC" [OR 6.25 (1.7, 33.3), P = 0.006], "GDM by NDDG" [OR 5.56 (1.3, 33.3), P = 0.018] and "Non-GDM" newborns [OR 6.47 (2.6, 14.8), P = 0.0003]. Conclusion: Using the NDDG criteria may increase healthcare costs because while it decreases the number of patients being diagnosed with GDM, it may also increase maternal and neonatal complications without changing maternal healthcare utilization.