(327) Evaluation of “low threshold” early diabetes screening practices at MAHEC

Thursday, January 30, 2025

10:30 AM – 12:30 PM

Submitting Author and Presenting Author(s)

RP

Rajesh Paladugu, MD

Fellow
Mountain Area Health Education Center (MAHEC)
Asheville, NC, United States

Coauthor(s)

Lisa M. Foglia, MD

Fellowship Director
Mountain Area Health Education Center (MAHEC)
Asheville, North Carolina, United States
TC

Traci Carson, PhD

Director of Analytics and Scholarship
Mountain Area Health Education Center (MAHEC)
Asheville, North Carolina, United States
KG

Katherine Grette, MD

Faculty Physician
Mountain Area Health Education Center (MAHEC)
Asheville, North Carolina, United States

Objective: This study aims to evaluate if a stepwise approach to early gestational diabetes (GDM)screening, using A1c followed by fasting plasma glucose (FPG), is more predictive of early and overall treatment as compared to an approach using A1c alone. Additionally, if this stepwise approach was more predictive of early insulin use than the algorithm utilized in Sweet Success (SS) guidelines.

Study Design: This was a retrospective cohort study of pregnant patients undergoing GDM screening in the first trimester between 2011-2023 at a single institution. Patients underwent stepwise testing with an initial A1C measurement followed by FPG if A1C was between 5.0 and 6.4. Patients were treated as GDM if both tests were elevated. Patients were then classified into four groups based on A1C and FPG results (table 1). The primary outcomes were the need for insulin at any point in pregnancy and gestational age at initiation.

Results: 887 patients met criteria for inclusion (Table 1). Demographic characteristics were similar across all groups. Patients in group 2 were more likely than patients in group 1 to require insulin during pregnancy (32 v 4.2%, p < 0.0001) and required earlier initiation (20 v 32 wks, p < 0.0001). Additionally, the stepwise algorithm had a higher specificity for identification of patients who would not require insulin prior to 28 weeks as compared to the SS algorithm (95 v 88%) though the sensitivity was lower (48 v 65%). Negative predictive values were similar (97 v 98%). PPVs for both algorithms were low (32 v 22%).

Conclusion: Stepwise diabetic screening using a lower A1c threshold followed by FPG was associated with overall insulin use and with earlier initiation. However, the overall number of patients who were captured using this technique was low. The poor PPV for both algorithms for insulin initiation prior to 28 weeks when patients would be captured by traditional GDM screening casts doubt on the clinical utility of universal early screening. This is consistent with the results of a recent RCT on early treatment of GDM and revised ACOG guidelines.