BACKGROUND: Laboratory measurements are an integral part of epidemiological studies in cardiovascular disease. Standardization and quality assurance is of utmost importance in the context of multicentre studies. METHODS: We evaluated a simple and cost-effective method of quality assurance for measurement of total cholesterol, high density lipoprotein (HDL) cholesterol and triglycerides in a study involving 10 centres. Three methods for quality assessment were used for the study that involved measurement of cholesterol, triglycerides and HDL cholesterol and included internal quality control, external quality control and 10% repeat analysis in addition to a uniform standardized protocol developed for the 10 centres. External quality control material was prepared and circulated by the coordinating laboratory. RESULTS: External quality control material was distributed 20 times during the study. The mean variance index suggested a substantial improvement in the performance of participating laboratories over a period of time for cholesterol and triglycerides. This was also evident in the improvement in per cent technical error as a measure of bias and a higher correlation between replicates of samples analysed in the coordinating laboratory and the participating centres for cholesterol, triglycerides and HDL cholesterol. CONCLUSION: A cost-effective quality assurance model for laboratory measurement using local capacities was developed and implemented in a multicentre epidemiology study. Such a programme would be useful for developing countries where cost-cutting is important.
BACKGROUND: Laboratory measurements are an integral part of epidemiological studies in cardiovascular disease. Standardization and quality assurance is of utmost importance in the context of multicentre studies. METHODS: We evaluated a simple and cost-effective method of quality assurance for measurement of total cholesterol, high density lipoprotein (HDL) cholesterol and triglycerides in a study involving 10 centres. Three methods for quality assessment were used for the study that involved measurement of cholesterol, triglycerides and HDL cholesterol and included internal quality control, external quality control and 10% repeat analysis in addition to a uniform standardized protocol developed for the 10 centres. External quality control material was prepared and circulated by the coordinating laboratory. RESULTS: External quality control material was distributed 20 times during the study. The mean variance index suggested a substantial improvement in the performance of participating laboratories over a period of time for cholesterol and triglycerides. This was also evident in the improvement in per cent technical error as a measure of bias and a higher correlation between replicates of samples analysed in the coordinating laboratory and the participating centres for cholesterol, triglycerides and HDL cholesterol. CONCLUSION: A cost-effective quality assurance model for laboratory measurement using local capacities was developed and implemented in a multicentre epidemiology study. Such a programme would be useful for developing countries where cost-cutting is important.