Evidence-based medicine in otolaryngology, Part 3: everyday probabilities: diagnostic tests with binary results

Otolaryngol Head Neck Surg. 2012 Aug;147(2):185-92. doi: 10.1177/0194599812447750. Epub 2012 May 15.

Abstract

Recent studies have demonstrated that the majority of physicians cannot accurately determine the predictive values of diagnostic tests. Physicians must understand the predictive probabilities associated with diagnostic testing in order to convey accurate information to patients, a key aspect of evidence-based practice. While sensitivity and specificity are widely understood, predictive values require a further understanding of conditional probabilities, pretest probabilities, and the prevalence of disease. Therefore, this third installment of the series "Evidence-Based Medicine in Otolaryngology" focuses on understanding the probabilities needed to accurately convey the results of dichotomous diagnostic tests in everyday practice.

MeSH terms

  • Diagnostic Tests, Routine / statistics & numerical data
  • Evidence-Based Medicine*
  • Humans
  • Otolaryngology / standards*
  • Otorhinolaryngologic Diseases / diagnosis*
  • Predictive Value of Tests
  • Probability
  • Sensitivity and Specificity
  • Uncertainty