Biases in diagnostic test studies: Impact on estimating sensitivity and specificity

Abstract

Diagnostic tests make it possible to determine whether a person has a disease or not. Before incorporating a new diagnostic test in the clinical setting, it is necessary to define its validity through its indicators of performance, sensitivity, and specificity. In these studies, like in any research, the results might not be reliable when there are biases during their execution. This article entails the discussion about the biases in diagnostic test studies that may cause inaccuracy in sensitivity and specificity. The main biases that affect the validity of these studies are: incorporation bias, partial and/or differential verification bias, an imperfect reference standard, a limited spectrum of the disease, and the ambiguous results of the test to be validated. In addition, examples of how these biases impact on the results of sensitivity and specificity are given in this paper.