Stankevich et al. (1) demonstrate that between 69% and 89% of first HbA1c analyses performed in the southwest region of France are performed as screening tests: only 11% to 31% are performed to monitor patients with known diabetes. Of the screening tests performed, 4.1% were diagnostic of diabetes (HbA1c ≥6.5%) and 21.3% were diagnostic of prediabetes (HbA1c 5.7–6.4%) (1). The prevalence of treated diabetes in the Aquitaine region has been reported to be 3.9% (2), and the prevalences of diagnosed and undiagnosed diabetes have been reported to be 4.6% and 2.7%, respectively, in French adults 20–79 years of age (3). Their findings (1), like ours (4), suggest that HbA1c testing is being used to target individuals at high risk for prediabetes and previously undiagnosed diabetes.
The data from our study further suggest that HbA1c testing is more likely to be performed in overweight or obese individuals with hypertension and dyslipidemia and in those with glucose test results that suggest a diagnosis of intermediate hyperglycemia. Impaired fasting glucose (100–126 mg/dL) has been reported to be sensitive (76%) but not specific (59%), and HbA1c of 5.7–6.4% has been reported to be less sensitive (36%) but more specific (81%), for identifying individuals at risk for progression to diabetes (5). Initial screening with fasting glucose will identify as many as 44% with prediabetes but will result in many false positive results, whereas initial screening with HbA1c will identify only 21% with prediabetes and will result in many false negative results (5). Initial screening with fasting or even random glucose levels and subsequent confirmation of the diagnosis of prediabetes based on HbA1c testing should be further explored as a pragmatic and potentially cost-effective approach to diagnosing prediabetes and diabetes in clinical practice.
Article Information
Duality of Interest. No potential conflicts of interest relevant to this article were reported.