NGHIÊN CỨU TƯƠNG QUAN TUYẾN TÍNH GIỮA CÁC PHƯƠNG PHÁP ĐỊNH LƯỢNG LDL-CHOLESTEROL GIÁN TIẾP VỚI TRỰC TIẾP Ở NGƯỜI BỆNH ĐÁI THÁO ĐƯỜNG

Abstract

Background: Using the formulas to estimate the amount of LDL-cholesterol indirectly has been allowed and widely applied in laboratories, to reduce the time to answer test results and save costs for patients. Objectives: 1) To determine the correlation by linear regression of LDL-C between indirect and direct methods of quantification. 2) To determine the percentage difference of LDL-C between indirect and direct methods of quantification on diabetes patients. Materials and method: A cross-sectional descriptive study was carried out of 817 diabetic patients who were indicated for cholesterol, triglyceride, HDL-C and LDL-C testing. By the way, we estimated LDLC concentration by using the Friedewald and Vujovic formulas. Results: The correlation of LDL-C by Vujovic's

indirect quantitative method compared with LDL-C level by direct method was r=0.785, with linear equation y=0.696x+0.769 (p<0.05). The correlation of LDL-C by Friedewald's indirect quantitative method compared with LDL-C level by direct method was r=0.723, with linear equation y=0.56x+1.390 (p<0.05). The LDL-C bias between Vujovic indirect method with LDL-C in direct method was 4.38% (p<0.05). The bias LDL-C between Friedewald indirect method with LDL-C in direct method was 10.08% (p<0.05). Conclusions: In diabetes patients, there is a strong, positive linear correlation in LDL-C concentration between Vujovic and Friedewald’s indirect quantitative methods with direct quantification. The bias of LDL-C concentration in the Vujovic is lower than in Friedewald.