Evaluating the impact of an evidence-based medicine educational intervention on primary care doctors' attitudes, knowledge and clinical behaviour: a controlled trial and before and after study.

RATIONALE, AIMS AND OBJECTIVES: Traditional continuing medical education programmes that offer passive learning have been shown to be poorly effective at changing doctors' clinical behaviour. A multifaceted evidence-based medicine (EBM) intervention was conducted at the largest health maintenance organization (HMO) in Israel, attempting to facilitate a change in doctors' attitudes, knowledge and clinical behaviour. No study thus far has examined the association between the teaching of EBM principles and doctors' clinical behaviour. This study evaluated the intervention programme through a controlled trial and before and after study. The objective of the evaluation is binary: first, to examine the impact of an educational intervention on family doctors' test ordering performance and drug utilization by their patients; and second, to assess the impact of the intervention on attitudes towards evidence-based practice and knowledge.

METHODS: Controlled trial and before and after study. Primary care clinics comprising similar patient characteristics were randomly allocated to the experimental or to the control group. Doctors in the experimental group participated in an EBM educational intervention, while the control group did not take part in the intervention. Clinicians' test ordering performance and their patients' drug utilization were derived from the HMO's database before intervention, after workshops and after intervention. Participants in the controlled trial consisted of 75 doctors and their 106 349 patients. The before and after study evaluated intervention doctors' (n = 70) EBM attitudes and knowledge through a validated questionnaire before and after workshops.

RESULTS: EBM workshops enhanced intervention doctors' EBM knowledge scores from 22.4/100 before workshops to 40.8/100 after workshops (P = 0.000). Doctors improved their ability to formulate clinical questions while enhancing their search strategy using Medline. In a linear regression model, two covariates, specialization (B = 12.59; P = 0.001) and habitually reading medical journals (B = 6.45; P = 0.052), best explained the variance in doctors' EBM knowledge scores, while controlling for pre-intervention scores (R(2) = 0.569; P = 0.000). Results from the controlled trial indicated that no statistically significant differences were found between intervention and control doctors' test ordering performances, and their patients' drug utilization.

CONCLUSIONS: The results of the study suggest that the intervention positively influenced attitudes and knowledge; however, no statistically significant impact was found on doctors' test ordering performance and on their patients' drug utilization. The intervention's inability to change doctors' clinical behaviour might be remedied by improving future interventions through adding additional facets to the educational intervention, such as social marketing techniques and personal feedback. A longer and more extensive intervention might be more effective but is extremely difficult to execute as we found in this study. Future larger-scale interventions must incorporate the intervention into the routines of the organization, thus minimizing barriers towards EBM implementation.