Randomized trial of the effect of research design and publication characteristics on physician change

Prim Care Companion J Clin Psychiatry. 2006;8(1):32-8. doi: 10.4088/pcc.v08n0105.

Abstract

Background: The primary barrier to translation of research into practice relates to physician use of research. If we are to succeed at translating research into practice, we must understand to which research characteristics and publication formats practitioners attend.

Objective: To determine which characteristics of research design (sample characteristics, study design) and publication (type of publication) are most influential on the acquisition of knowledge and change in behavior of family practitioners.

Method: This randomized clinical trial was conducted in family practice offices on the 305 family physicians who scored lowest on a survey of knowledge about management of major depressive disorder (MDD), panic disorder, and generalized anxiety disorder (GAD). Subjects were randomly assigned to receive 1 of 3 MDD abstracts differing in study site, 1 of 3 panic disorder abstracts differing in study design, and 1 of 3 GAD communications differing in format. The main outcome measures (knowledge and management strategies) were assessed immediately following the intervention and again 6 months later. Data for the intervention survey were gathered in November 2002.

Results: This study found significant increases in knowledge level and use of first-line agents with all interventions; however, knowledge declined again after 6 months for both panic disorder and GAD. The only statistically significant interoption difference was that the POEM (patient-oriented evidence that matters) was associated with better retention of knowledge of the treatment of GAD.

Conclusion: In conclusion, all interventions were associated with immediate increases in knowledge and use of first-line treatments. However, such gains were not retained for panic disorder and GAD. Except for better knowledge retention associated with POEM use, no consistent differences between intervention options were noted.