Systematic review of evidence-based medicine tests for family physician residents

Evidence-based medicine (EBM) is increasingly important in resident education, and reliable and valid tests of competence for family medicine residents are needed.

MEDLINE, PsycINFO, ERIC, ERC, and the Research & Development Resource Base (University of Toronto) were searched from inception to June 2014 to identify competence tests of family medicine and general practice residents in EBM. Two authors independently assessed all titles, abstracts, and full texts and abstracted data.

Three EBM courses were evaluated by the Fresno test. Seven other authors designed EBM interventions and individual tests to evaluate them. Content validity was assessed by nine studies, construct validity by five, face validity by three, and concurrent validity by one. Internal reliability was reported by seven studies, inter-rater by four, item difficult and item discrimination by two, and intra-rater by one. Eight studies reported that knowledge scores increased significantly after the intervention.

Content validity and internal reliability were the most frequently assessed measures. The basic EBM activities of identifying Population, Intervention, Comparison, Outcome, and Study Design (PICOS questions) and computing sensitivity, specificity, and number needed to treat (NNT) are unlikely to change. However, guidelines are often used in EBM courses, and they are updated regularly, which will involve new clinical scenarios, PICOS questions, and statistical computations. The Fresno test has been evaluated with three groups of family medicine residents, has the best documentation of validity and reliability, and is the best candidate for future development. Evaluation tools also need to be developed to measure if care received by patients is EBM.