A 25-year Experience in Quality Assurance: A Methodological Review of Laboratory Quality Control and Quality Assurance in the Tehran Lipid and Glucose Study (TLGS).
Quality control (QC) and quality assurance (QA) are critical for validating laboratory data in longitudinal studies. The Tehran Lipid and Glucose Study (TLGS), the first large cohort study in Iran, was designed to investigate the incidence of and risk factors for cardiovascular diseases. This article reviews the methodological quality control system used for biochemical testing in the TLGS, with an emphasis on its design, implementation, evaluation, and data stability across study phases.
In this methodological review, internal and external quality-control protocols used in phases 2 to 7 of the TLGS were analyzed and compared with international standards, including those of the Cardiovascular Health Study and the Atherosclerosis Risk in Communities study. Analytical precision data (coefficient of variation [CV%]) for glucose, creatinine, total cholesterol, triglycerides, HDL-C, ALP, ALT, and AST were extracted from a central database, and temporal trends were evaluated. Internal quality control was performed daily using two levels of control sera (normal and abnormal), and interlaboratory accuracy was assessed through participation in an external evaluation program.
The mean CV for all tests across the study phases was less than 5% (glucose, 1.95%; creatinine, 3.10%; cholesterol, 2.05%; triglycerides, 2.16%; HDL-C, 3.33%; ALP, 2.10%; ALT, 3.00%; and AST, 3.35%). External quality-control results received an "Excellent" rating for all parameters. The temporal trend in CVs demonstrated stability and a gradual improvement in precision across phases. Levey-Jennings charts, Westgard rules, and weighted regression methods were effective in detecting systematic deviations and enabling prompt corrective action.
Establishing a multilevel QC/QA system in the TLGS ensured the stability and accuracy of biochemical data in a large national cohort. The findings indicate that the accuracy of biochemical testing was consistent with international standards and that the resulting data are highly reliable for epidemiological and clinical analyses. This experience may serve as a local model for other longitudinal studies in developing countries.
In this methodological review, internal and external quality-control protocols used in phases 2 to 7 of the TLGS were analyzed and compared with international standards, including those of the Cardiovascular Health Study and the Atherosclerosis Risk in Communities study. Analytical precision data (coefficient of variation [CV%]) for glucose, creatinine, total cholesterol, triglycerides, HDL-C, ALP, ALT, and AST were extracted from a central database, and temporal trends were evaluated. Internal quality control was performed daily using two levels of control sera (normal and abnormal), and interlaboratory accuracy was assessed through participation in an external evaluation program.
The mean CV for all tests across the study phases was less than 5% (glucose, 1.95%; creatinine, 3.10%; cholesterol, 2.05%; triglycerides, 2.16%; HDL-C, 3.33%; ALP, 2.10%; ALT, 3.00%; and AST, 3.35%). External quality-control results received an "Excellent" rating for all parameters. The temporal trend in CVs demonstrated stability and a gradual improvement in precision across phases. Levey-Jennings charts, Westgard rules, and weighted regression methods were effective in detecting systematic deviations and enabling prompt corrective action.
Establishing a multilevel QC/QA system in the TLGS ensured the stability and accuracy of biochemical data in a large national cohort. The findings indicate that the accuracy of biochemical testing was consistent with international standards and that the resulting data are highly reliable for epidemiological and clinical analyses. This experience may serve as a local model for other longitudinal studies in developing countries.
Authors
Hedayati Hedayati, Tohidi Tohidi, Azimzadeh Azimzadeh, Momenan Momenan, Daneshpour Daneshpour, Zarkesh Zarkesh, Pakizehkar Pakizehkar, Hosseinzadeh Hosseinzadeh, Azizi Azizi
View on Pubmed