Alternative type 2 diabetes screening tests may reduce the number of U.S. adults with undiagnosed diabetes

Aim To evaluate the U.S. population‐level impact of two alternatives for initial type 2 diabetes screening [opportunistic random plasma glucose (RPG) > 6.7 mmol/l and a 1‐h 50‐g glucose challenge test (GCT) > 8.9 mmol/l] compared with American Diabetes Association (ADA)‐recommended tests. Methods Using a sample ( n = 1471) from the National Health and Nutrition Examination Survey (NHANES) 2013–2014 that represented 145 million U.S. adults at high risk for developing type 2 diabetes, we simulated a two‐test screening process. We compared ADA‐recommended screening tests [fasting plasma glucose (FPG), 2‐h 75‐g oral glucose tolerance test (OGTT), HbA 1c ] vs. initial screening with opportunistic RPG or GCT (followed by FPG, OGTT or HbA 1c ). After simulation, participants were entered into an individual‐level Monte Carlo‐based Markov lifetime outcomes model. Primary outcomes were representative number of U.S. adults correctly identified with type 2 diabetes, societal lifetime costs and quality‐adjusted life years (QALYs). Results In NHANES 2013–2014, 100 individuals had undiagnosed diabetes [weighted estimate: 8.4 million, standard error ( se ): 1.1 million]. Among ADA‐recommended screening tests, FPG followed by OGTT (FPG–OGTT) was most sensitive, identifying 35 individuals with undiagnosed diabetes (weighted estimate: 3.2 million, se : 0.9 million). Four alternative screening strategies performed superior to FPG–OGTT, with RPG followed by OGTT being the most sensitive overall, identifying 72 individuals with undiagnosed diabetes (weighted estimate: 6.1 million, se : 1.0 million). There was no increase in average lifetime costs and comparable QALYs. Conclusions Initial screening using opportunistic RPG or a GCT may identify more U.S. adults with type 2 diabetes without increasing societal costs.