Appraising and applying evidence about a diagnostic test during a performance-based assessment

BMC Med Educ. 2004 Oct 13;4:20. doi: 10.1186/1472-6920-4-20.


Background: The practice of Evidence-based Medicine requires that clinicians assess the validity of published research and then apply the results to patient care. We wanted to assess whether our soon-to-graduate medical students could appraise and apply research about a diagnostic test within a clinical context and to compare our students with peers trained at other institutions.

Methods: 4th year medical students who previously had demonstrated competency at probability revision and just starting first-year Internal Medicine residents were used for this research. Following an encounter with a simulated patient, subjects critically appraised a paper about an applicable diagnostic test and revised the patient's pretest probability given the test result.

Results: The medical students and residents demonstrated similar skills at critical appraisal, correctly answering 4.7 and 4.9, respectively, of 6 questions (p = 0.67). Only one out of 28 (3%) medical students and none of the 15 residents were able to correctly complete the probability revision task (p = 1.00).

Conclusions: This study found that most students completing medical school are able to appraise an article about a diagnostic test but few are able to apply the information from the article to a patient. These findings raise questions about the clinical usefulness of the EBM skills possessed by graduating medical students within the area of diagnostic testing.

Publication types

  • Comparative Study

MeSH terms

  • Bayes Theorem*
  • Clinical Clerkship / standards
  • Clinical Competence / statistics & numerical data*
  • Depressive Disorder, Major / diagnosis
  • Education, Medical, Undergraduate / methods*
  • Evidence-Based Medicine / education*
  • Evidence-Based Medicine / statistics & numerical data
  • Heart Failure / blood
  • Heart Failure / diagnosis
  • Hematologic Tests / statistics & numerical data
  • Hospitals, University
  • Humans
  • Internal Medicine / education*
  • Internship and Residency / standards
  • Iowa
  • Panic Disorder / diagnosis
  • Patient Care
  • Patient Simulation
  • Predictive Value of Tests*
  • Psychiatry / education*
  • Reproducibility of Results
  • Sensitivity and Specificity