Intervention to Reduce Unnecessary Glucose Tolerance Testing in Pregnant Women.

Abstract

Gestational diabetes mellitus (GDM) can be diagnosed in pregnant women by increased fasting plasma glucose alone, which eliminates the need for performing a 75 g oral glucose tolerance test (OGTT). If whole blood glucose meters are used to triage fasting samples in order to decide whether to give the glucose drink, a cutpoint with appropriate sensitivity and specificity for elevated fasting plasma glucose is needed. The number of GDM diagnoses by increased fasting plasma glucose alone was determined from specimens collected and tested at core laboratories in urban hospitals, rural health centers, and from specimens collected at patient phlebotomy service centers (PSCs) for plasma testing at a central laboratory. The number of glucose drinks avoided was counted after implementing the diagnostic cutoff of ≥95 mg/dL (5.3 mmol/L) at urban hospitals and rural health centers, which have on-site plasma testing, and after selecting a PSC meter fasting venous whole blood glucose cutpoint after calculating sensitivity and specificity for plasma glucose ≥95 mg/dL (5.3 mmol/L) using logistic regression. Among 4850 OGTTs, there were 1315 GDM diagnoses annually, of which 409 were from increased fasting plasma glucose. Ninety-one percent of OGTTs were performed at PSCs. If a fasting plasma glucose cutpoint of ≥95 mg/dL (5.3 mmol/L) was implemented at urban hospitals and rural health centers and a meter fasting venous whole blood glucose cutpoint of ≥108 mg/dL (6.0 mmol/L) (25% sensitivity, 99.9% specificity) was implemented at PSCs, the drink would be appropriately avoided by 145 patients/year, and inappropriately avoided by 3 patients/year. After implementing these cutpoints, the drink was appropriately avoided in 91 patients during a 36-week period, with none inappropriately avoiding it. Modifying fasting glucose cutpoints reduced unnecessary diagnostic OGTTs in pregnant women.