Is fasting blood glucose a reliable parameter for screening for diabetes in hypertension?

BACKGROUND The aim of this study was to evaluate the sensitivity and specificity of a combination of fasting plasma glucose (FPG) and glycosylated hemoglobin (HbA(1c)) for screening for diabetes in hypertensive patients. The oral glucose tolerance test (OGTT) served as a gold standard for the detection of diabetes. METHODS The cross-sectional study was performed in the Hypertension Unit of the Department of Emergency Medicine in the General Hospital of Vienna between January 1999 and July 2001. The FPG, HbA(1c), and OGTT were performed in 152 hypertensive patients (77 male and 75 female) to identify those individuals with diabetes. RESULTS A total of 33 patients were identified as diabetic based on the results of the OGTT. Diabetes was detected in 25 (16%) of 152 patients using an FPG > or =7.0 mmol/L. In addition, HbA(1c) was measured in the remaining 127 (84%) patients with an FPG <7.0 mmol/L. In these patients HbA(1c) > or =6.1% showed a sensitivity of 100% and a specificity of 75%. Cost analysis of both approaches (FPG + HbA(1c) versus OGTT in all patients) demonstrated a cost sparing effect of 31.03 $/patient (31.66$/patient) in favor of the combined use of FPG and HbA(1c). CONCLUSION The combination of FPG and HbA(1c) is a reasonable alternative to the generally recommended OGTT for the screening of diabetes, as diabetes was correctly diagnosed in all patients by this stepwise procedure.