Abstract

Abstract Objective To compare the cost and time to diagnosis associated with several screening strategies for diabetes in women with histories of gestational diabetes mellitus (GDM)

Abstract Research Design and Methods We simulated screening for diabetes with fasting plasma glucose (FPG), a 2-hour oral glucose tolerance test (OGTT), and a hemoglobin A1c (HbA1c) annually, every 2 years, and every 3 years over a period of 12 years. We assumed that women had negative screening tests 6 weeks after delivery, progressed to diabetes at 8% per year, and that each positive FPG and HbA1c was followed by a confirmatory FPG. For each strategy, we calculated the cost per case detected, cost per woman screened, the percent of cases detected, and the time elapsed with undiagnosed diabetes. In sensitivity analyses, we considered inclusion of indirect costs, impact of imperfect adherence to screening strategies, exclusion of confirmatory tests, and lower rates of progression to diabetes.

Abstract Results When annual, biannual, or every 3 year screening strategies were employed, OGTT resulted in lower costs per case detected than FPG or HbA1c. Testing every 3 years resulted in lower costs per case detected than more frequent testing. These patterns persisted in sensitivity analyses, except FPG resulted in lower cost per case detected than OGTT assuming annual screening and inclusion of indirect costs, or assuming annual screening without a confirmatory FPG.

Abstract Conclusions Screening every 3 years with OGTT results in the lowest cost per case of diabetes detected.