Use of prediction equations for reviewing measurement results in the clinical laboratory

Abstract

In the clinical laboratory, one of the most objective ways to perform the final review of measurement results is the use of the so-called plausibility control (i.e., set of procedures used to decide if a measurement result is valid or not according to established clinical and biological criteria). The present study is focused on the estimation of several prediction equations derived from pairs of biological quantities having a pathophysiological relationship and statistically correlated to detect objectively doubtful results in the context of plausibility control. These prediction intervals, that may be used alone or combined with other procedures involved in the plausibility control, are a very useful tool for the improvement of the final review of the laboratory results.