Abstract 061: Comparison Of Diagnostic Methods For Diabetes Diagnosis In Patients With Heart Failure: A National Health And Nutrition Examination Survey Analysis

Abstract

Introduction: The use of hemoglobin A1c (HbA1c) to screen for diabetes in patients with cardiovascular disease has shown low sensitivity, and the currently recommended cutoff point may leave many patients undiagnosed. The objective of the study was to examine the degree of agreement among current diagnostic methods for diabetes, including HbA1c, fasting plasma glucose (FPG), and two-hour plasma glucose (2hPG) after oral glucose tolerance test (OGTT) among community-dwelling heart failure (HF) patients. Methods: We included 237 HF participants (20+ years) without history of diabetes or diabetes medication use, using data from the 2005-2016 National Health and Nutrition Examination Survey cycles. Diagnostic criteria for diabetes were based on the American Diabetes Association (ADA) guidelines: (1) HbA1c ≥ 6.5%, (2) FPG ≥ 126 mg/dL, and (3) 2hPG after OGTT ≥ 200 mg/dL. The sensitivity, specificity, and Receiver Operating Characteristic (ROC) curve for HbA1c and FPG were examined against the OGTT (FPG ≥ 126 mg/dL and/or 2hPG ≥ 200 mg/dL) as the reference method. Results: Included participants had a mean age of 65±1 (SE) years, 45.5% were female, and 12.5% were non-Hispanic black. Among included participants, 50 patients were newly diagnosed with diabetes based on ADA guidelines. Hemoglobin A1c alone identified only 27.0% of patients, whereas the OGTT identified 90.1% of patients. The area under the ROC curve, assessing HbA1c performance for the diagnosis of diabetes defined by OGTT, was 0.76 (95% CI, 0.67-0.84). Sensitivity and specificity using the HbA1c cutoff at ≥ 6.5% were 19.1% and 97.5%, respectively. The Youden’s J statistic for HbA1c was maximized at 5.8-5.9% (69.9-64.3% sensitivity and 63.9-70.1% specificity). Conclusion: Our data show that identification of diabetes among HF patients was inadequate on the basis of HbA1c screening alone. Our findings suggest diabetes screening on the basis of the OGTT or an HbA1c cutoff point of 5.8-5.9%.