Tailored interventions to overcome identified barriers to change: effects on professional practice and health care outcomes

Abstract

Background: In the previous version of this review, the effectiveness of interventions tailored to barriers to change was found to be uncertain.

Objectives: To assess the effectiveness of interventions tailored to address identified barriers to change on professional practice or patient outcomes.

Search strategy: For this update, in addition to the EPOC Register and pending files, we searched the following databases without language restrictions, from inception until August 2007: MEDLINE, EMBASE, CINAHL, BNI and HMIC. We searched the National Research Register to November 2007. We undertook further searches to October 2009 to identify potentially eligible published or ongoing trials.

Selection criteria: Randomised controlled trials (RCTs) of interventions tailored to address prospectively identified barriers to change that reported objectively measured professional practice or healthcare outcomes in which at least one group received an intervention designed to address prospectively identified barriers to change.

Data collection and analysis: Two reviewers independently assessed quality and extracted data. We undertook quantitative and qualitative analyses. The quantitative analyses had two elements.1. We carried out a meta-regression to compare interventions tailored to address identified barriers to change with either no interventions or an intervention(s) not tailored to the barriers.2. We carried out heterogeneity analyses to investigate sources of differences in the effectiveness of interventions. These included the effects of: risk of bias, concealment of allocation, rigour of barrier analysis, use of theory, complexity of interventions, and the reported presence of administrative constraints.

Main results: We included 26 studies comparing an intervention tailored to address identified barriers to change to no intervention or an intervention(s) not tailored to the barriers. The effect sizes of these studies varied both across and within studies.Twelve studies provided enough data to be included in the quantitative analysis. A meta-regression model was fitted adjusting for baseline odds by fitting it as a covariate, to obtain the pooled odds ratio of 1.54 (95% CI, 1.16 to 2.01) from Bayesian analysis and 1.52 (95% CI, 1.27 to 1.82, P < 0.001) from classical analysis. The heterogeneity analyses found that no study attributes investigated were significantly associated with effectiveness of the interventions.

Authors' conclusions: Interventions tailored to prospectively identified barriers are more likely to improve professional practice than no intervention or dissemination of guidelines. However, the methods used to identify barriers and tailor interventions to address them need further development. Research is required to determine the effectiveness of tailored interventions in comparison with other interventions.

Figures

Figure 1

Meta-regression plotThe log odds ratio at follow-up is plotted against the log odds ratio at baseline, with each circle representing one study in the analysis, and the red line indicating the pooled estimated follow-up log odds ratio for each value of the baseline log odds ratio. Circle size is relative to the standard error of the log odds ratio. The result from the classical analysis can be read from the graph in that when the intervention and control groups are equal at baseline (i.e. when the baseline odds ratio=0) then the estimated log odds ratio from the plot is 0.42. When exponentiated this gives the pooled effect size of 1.52.