PubMed2002
BACKGROUND: In order to decide on the feasibility of screening for disease by laboratory testing, a number of assessments need to be made. The necessary procedures and statistical calculations are required for an objective evaluation of the diagnostic accuracy of a laboratory test and of the potential for screening and of the data obtained from the screening. METHODS: The determination of reference intervals, is an essential first step in the evaluation of the performance of a laboratory test. The procedures and calculations are necessary for the determination of the sensitivity, specificity, diagnostic predictive value and receiver-operating characteristic curve of a test and their respective applications in laboratory screening for disease. The additional statistical tools include odds ratios and likelihood ratios for assessing the performance of a laboratory test. CONCLUSION: These tools are used in the development of an appropriate clinical decision limit in a defined population of subjects and under defined conditions for acceptable screening.
Clinical Laboratory Techniques↗