Interpretation and Analysis of Individual Diagnostic Tests and Performance

Diagnostic tests are performed daily by veterinary practitioners and for various pur-poses. For many tests, accuracy characteristics are available in the form of an esti-mated sensitivity (ability to detect affected animals) and specificity (ability to detect nonaffected animals). However, because gold standard tests are not available for many clinical diseases, the imperfect sensitivity and specificity can lead to misdiag-nosis especially in the absence of a clear diagnostic test prescription context. Clearly establishing the reasons of test prescription based on a prior knowledge of the test probability (pretest probability) is helpful to put test results in perspective (eg, using Bayes rule). This clinical approach also emphasizes the validation of diagnostic test accuracy results. Making informed decisions based on validated diagnostic tests ulti-mately serves to improve patient care and well-being and to tailor antimicrobial use, and would also improve the image of the veterinary practitioner as an added-value for diagnostic test interpretation.