The use of HbA1c improved diagnosis of Diabetes Mellitus among at-risk individuals in the community setting in Hong Kong

Abstract

INTRODUCTION: HbA1c was adopted as a diagnostic criterion for the detection of diabetes mellitus (DM) by the American Diabetic Association (ADA) in 2010 and the World Health Organization (WHO) in 2011. This study aims to evaluate the change in prevalence of newly diagnosed DM when HbA1c is used together with plasma glucose tests for diagnosing DM among at-risk patients in a general-out-patient-clinic (GOPC) in Hong Kong. METHOD: A retrospective cohort study on anonymous patients was conducted at Cheung Sha Wan Jockey Club GOPC from January 2009 to July 2011. 106 patients who had available paired results of HbA1c and fasting plasma glucose between January to December 2009 and without known diagnosis of DM before the test period were identified by the Clinical Data Analysis and Reporting System (CDARS) and included in the data analysis. The prevalence of newly diagnosed DM was determined by both HbA1c and conventional plasma glucose criteria based on available laboratory results up to July 2011. RESULTS: 44 patients (41.5%) were newly diagnosed to have DM during the study period. Plasma glucose criteria identified 23 of the new diabetic patients (52.3%) whereas HbA1c identified a total of 34 patients (77.3%); 21 patients (47.7%) would not have been diagnosed to have DM if only plasma glucose tests were used; the prevalence of DM among this group of patients increased from 21.7% to 41.5% when both criteria were used instead of plasma glucose tests only. Based on the paired HbA1c-fasting glucose results in 2009, HbA1c identified 8 out of the 9 new diabetic cases confirmed by fasting glucose tests, giving a sensitivity of 89%. The negative predictive value of HbA1c was 99% (85/86) when compared against fasting glucose tests as the diagnostic standard. DISCUSSION: Despite limitations in methodology and sample size, our results demonstrated how the use of HbA1c as a diagnostic criterion for DM in addition to conventional plasma glucose criteria significantly improved the detection of DM among at-risk patients in the clinical setting in Hong Kong with good sensitivity and negative predictive value. Larger scale prospective study is warranted to further delineate the actual impact of its use.