“Tailoring” of evidence-based messages: worth the effort?

Background: Marketing theory suggests that for effective knowledge translation, we should tailor the way we present scientific evidence to take account of the beliefs, attitudes and motivation of our intended audience. The evidence on how to tailor evidence-based messages is limited. This study, in Swedish general practice, sought to test the impact of a systematic approach to message tailoring.

Aim: Using ACE inhibitor prescribing for hypertension, to test the hypothesis that evidence-based drug information provided to GPs in “tailored” format (that is, adjusted according to their initial attitudes and motivation) would have more influence on prescribing behaviour than evidence based drug information provided with no tailoring, and to identify the key drivers influencing their prescribing decision.

Study design: Three-phase study comprising exploratory focus groups, cross-sectional survey of attitudes and randomised controlled trial (RCT).

Methods: Focus groups with 16 GPs explored attitudes to evidence-based medicine and drug prescribing. These findings fed into the design and refinement of a postal survey to 368 GPs to generate quantitative data on these attitudes. In the RCT, 14 medical information officers providing drug information were pair-wise matched and randomized into an intervention or control group. The intervention group was trained to provide evidence-based drug information tailored by motivational interviewing with the GP and also based on key findings from the focus group study. The control group provided standard evidence-based drug information without tailoring. In total, 991 GPs were included in the study (408 intervention and 583 control). The primary outcome measure was change in proportion of ACE inhibitor prescribed relative to the sum of ACE inhibitors and Angiotensin II receptor blockers, during 0–3 and 4–6 months after the intervention. These data were analyzed with multiple linear regression, by intention-to-treat and per protocol.

Results: Focus group data indicated strongly that the primary influence on GPs’ drug prescribing decisions was an expectation of “prompt and pragmatic benefit”. The survey indicated a range of views on evidence-based medicine and information from the pharmaceutical industry, with most GPs (especially women) perceiving the information from industry as excessive and the quality of public information as high and useful. The category “prompt and pragmatic benefit” was used to influence the training of medical information officers in the intervention group, so that they focused primarily on this element when providing information to GPs. The RCT demonstrated statistically significant improvements in the primary outcome measure in both groups following the intervention, but no statistically significant differences between the groups.

General conclusions and implications: Whilst it is possible to tailor evidence-based information to GPs and train medical information officers to deliver a personalized information service, the impact on prescribing behaviour may be no better than when standard information is provided. This study calls into question the recommendation to tailor information.

A discussion on the benefit of tailoring will be invited from the audience.