Critical Appraisal of Diagnostic Test Studies

Introduction

• Diagnostic test studies are studies that reveal the accuracy of a test of interest compared to a reference standard for the diagnosis of a condition.

  • test - a lab test, a physical exam maneuver, a history question, etc.

  • test of interest - a new or experimental test

  • reference standard - the established, certain method of diagnosing a condition

• These studies give us statistics like sensitivity, specificity, predictive values and likelihood ratios so that we can better understand the clinical implications of a positive or negative diagnostic test.

• The study designs that are used to evaluate diagnostic test accuracy are cohort studies and case-control studies.

  • Sometimes, you will find randomized controlled trials about diagnostic tests - these are used to study the effect of the routine use of these tests in clinical care, but as an intervention, i.e. one group gets the test and the other does not.  This is a different way to study diagnostic tests which does not yield the statistics above, and would be done after the basic accuracy studies are completed.

    • example - Beta natriuretic peptide was studied initially for its accuracy in diagnosing congestive heart failure with cohort studies.  After its predictive values and likelihood ratios showed that it was a reasonable test to use to diagnose CHF, researchers did a randomized clinical trial to assess whether physicians should use it routinely to manage people suspected of CHF by comparing a group of patients whose physicians used BNP levels to a group whose physicians did not.

Video Tutorials