Comparison of high performance liquid chromatography and turbidimetric inhibition immunoassay methods for measurement of hemoglobin A1c

Abstract

Aim: Hemoglobin A1c is a valuable parameter for the diagnosis and follow-up of its diabetes mellitus since its biological variation is low, does not require preparation before the test, is not affected by acute stress, and has high preanalytical stability. HbA1c measurement by HPLC has been determined as the reference method by National Glycohemoglobin Standardization Program (NGSP) in USA; after that The International Federation of Clinical Chemistry (IFCC) defined another reference method which could be related with NGSP. In our study, we aim to compare the two NGSP-certified methods of HbA1c, which are high-performance liquid chromatography (HPLC) and turbidimetric inhibition immunoassay (TINIA).
 Material and Method: HbA1c levels of the patients were measured using two HPLC and one TINIA method in three different hospitals (Lab A, Lab B (Both are HPLC), and Lab C (TINIA), in which Lab A was served as a reference). Because of the lower precision values of LabB, we firstly conducted a method comparison study of 40 volunteers (Group 1). After that, corrective and preventive activities carried out and the precision values in LabB reached the desired range. Following this, another method comparison study consisting of 60 new volunteers (Group 2) was conducted. The statistical flow of this study complied with Clinical Laboratory Standards Institute (CLSI) EP09-A3; Precision studies, Blant-Altman and Passing Bablok regression analysis were performed.
 Results: The percentage of the mean difference between the two HPLC methods (LabA and LabB) was 3.1%. After corrective and preventive actions had been taken, the mean difference between the two HPLC methods decreased to 2.0%. A decrease in systematic bias was found in our study. Two HPLC methods can be used interchangeably in both Group 1 and Group 2. In Group 1; 95% CI of intercept and slope were found as (-1.41 to -0.30) and (1.03 to 1.22), respectively. In Group 2; 95% CI of intercept and slope were found as (-1.33 to -0.31) and (1.01 to 1.17), respectively. HPLC and TINIA methods could not be used interchangeable without affecting patient results and outcome in both Group 1 and Group 2.
 Conclusion: Our study concluded that TINIA and HPLC methods could not be used interchangeably without affecting patient results and outcome. Because of the methodology that clinical laboratories are used to, clinicians and clinical biochemists should collaborate on managing diabetes mellitus regarding diagnosis, treatment, and follow-up.