Diagnostic accuracy of H b A 1c in diabetes between E astern and W estern

Background In 2010, the A merican D iabetes A ssociation recommended the use of H b A 1c as a diagnostic criterion for diabetes. However, H b A 1c is not an accepted diagnostic tool for diabetes in E astern A sia, because genetic differences compromise the standardization of the diagnostic cut‐off point. Objectives This study evaluated differences in the use of H b A 1c for diagnosing diabetes in E astern and W estern populations and investigated whether H b A 1c cut‐off point of ≥ 6·5% is diagnostic of diabetes in patients from E astern A sia. Methods Literature was obtained from MEDLINE , EMBASE and C ochrane databases. The pooled sensitivity and specificity of each H b A 1c cut‐off point were extracted and compared between W estern and E astern populations. Differences in the cut‐off point for diagnosing diabetes in each region were compared by examining differences in the area under summary receiver operating characteristic ( SROC ) curves .Results Twelve publications from E astern countries ( n  =   59 735) and 13 from W estern countries ( n  =   22 954) were included in the analysis. Areas under SROC curves in the E astern and W estern groups were 0·9331 and 0·9120, respectively ( P  =   0·98). The cut‐off point of the highest Y ouden index was 6·0%. At the H b A 1c cut‐off point of 6·5%, the pooled sensitivity and specificity were 58·7% and 98·4% for E astern countries and 65·5% and 98·1% for W estern countries, respectively. Conclusions H b A 1c exhibits the same diagnostic value for diabetes in E astern and W estern populations. In both populations, H b A 1c levels > 6·0% identify the population at high risk of diabetes, and H b A 1c  >  6·5% is diagnostic of clinically established diabetes.