Evaluation of Rapid Diagnostic Test Performance

Abstract

Rapid diagnostic tests are used for the determination of binary qualitative results not uniquely in non-hospital-based but also in hospital-based tests. Principally in developing countries, rapid diagnostic tests are the primary option since tests to be used in medical laboratories are discarded due to the higher cost. The test’s performance is evaluated to assure that the chance of results to be false is clinically acceptable. Therefore, the diagnostic accuracy of results (diagnostic sensitivity and specificity) is assessed to guarantee the safety of post-clinical decision. The statistical approach requires that representative samplings of the populations of infected and healthy individuals are tested. The area under the receiver operating characteristic curve is a complementary measurement using the same samplings. It represents the diagnostic accuracy in a single outcome. When samplings with known diagnostic are unavailable, samplings with known outcomes from a comparative test are used to determine the agreement of results. However, this approach is secondary, due to diagnostic accuracy to be unmeasurable. The seronegative period is another critical measurement allowing to determine an individual biological bias during a period where results of an infected individual are false negatives due to seroconversion. The claimed requirements should be defined for diagnostic accuracy and agreement outcomes. A spreadsheet is used to estimate the results considering the absolute value and the 95% confidence interval.