Evaluating primary care doctors' evidence-based medicine skills in a busy clinical setting.

RATIONALE, AIMS AND OBJECTIVES: To date, primary care doctors' (PCDs) evidence-based medicine (EBM) skills have rarely been studied. We conducted a cross-sectional study to evaluate PCDs' practical EBM skills and to determine risk markers associated with these skills.

METHODS: The study sample consisted of 70 PCDs (70.7% response rate) practising in a busy urban setting from a large health maintenance organization. Participants were given a short validated questionnaire gauging attitudes, barriers, online medical resources utilization, as well as demographic and personal characteristics. Additionally, doctors completed an online and written exam evaluating their ability to formulate clinical questions, and retrieve medical information efficiently. Data analysis was performed using both bivariate and multivariate analysis (linear regression).

RESULTS: PCDs found it difficult to formulate clinical questions both in the written and online exam, mostly neglecting to mention the Patient and Comparison components of PICO (patient, intervention, comparison and outcome). Search strategies primarily dispensed with the use of MeSH terms, ignoring appropriate limits. Doctors final scores were low (score = 41.5/100, SD = 16.2). In bivariate analysis clinical experience was negatively correlated with the final score (r = -0.36, P = 0.01), and specialists' scores were significantly higher than general practitioners' scores (46.7/100 and 31.5/100 respectively, P < 0.001). In multivariate analysis, doctors specialization was the only statistically significant predictor of the final score (B = 12.74, P = 0.002), while controlling for participating in a prior EBM course.

CONCLUSIONS: This study emphasizes the need for enhancing PCDs practical EBM skills. Future research and interventions should focus on this population emphasizing the specific needs of subpopulations (i.e. general practitioners and doctors without previous EBM training).