Gulab Fatima Rani, Hazrat Bilal, Shumaila Jabeen, Abid Sohail Taj
HbA1c is an approved test for the diagnosis and monitoring of diabetes. Several assays are in use worldwide based on chromatography, cation-exchange, enzymatic and immunological principles. High Performance Liquid Chromatography (HPLC) is regarded as the gold standard method while other assays including enzymatic assays (EA) are shown to have high efficiency and comparable results. In this study, the HbA1c levels were measured using HPLC and EA and compared for concordance and agreement. HbA1c levels were measured on blood samples of 1014 participants regardless of age, gender and type of diabetes. Samples were run on both assays as per manufacturer guidelines and standard operating procedures. Data analysed using GraphPad Prism for Pearson's correlation, simple linear regression, measures of diagnostic accuracy and Bland Altman analysis. Participants with mean age of 53.49 ± 13.94 year and 50.89% females were included. Mean HbA1c was slightly higher for HPLC overall and in females but showed a high correlation with correlation coefficient of 0.97. Correlation between HbA1c on two assays was strong positive for total samples and diabetic samples whereas moderate to weak in non-diabetics and pre-diabetics. HbA1c mean difference was within the acceptable range of bias for all samples hence suggestive of a good agreement between two assays. Diagnostic accuracy measures were slightly lower in HbA1c measured on EA in pre-diabetic and diabetic range HbA1c. Our findings show that the HbA1c levels measured using both assays show a strong association and agreement of bias. However, Further studies are recommended for comparison of pre-diabetic HbA1c range between assays while taking interference factors into account.