Identifying behaviour change techniques in 287 randomized controlled trials of audit and feedback interventions targeting practice change among healthcare professionals

Abstract

Abstract Background: Audit and feedback (A&F) is among the most widely used implementation strategies, providing healthcare professionals with summaries of their practice performance to prompt behaviour change and optimize care. Wide variability in effectiveness of A&F has spurred efforts to explore why some A&F interventions are more effective than others. Unpacking the variability of the content of A&F interventions in terms of their component behaviours change techniques (BCTs) may help advance our understanding of how A&F works best. This study aimed to systematically specify BCTs in A&F interventions targeting healthcare professional practice change. Methods: We conducted a directed content analysis of intervention descriptions in 287 randomized trials included in an ongoing Cochrane systematic review update of A&F interventions (searched up to June 2020). Three trained researchers identified and categorized BCTs in all trial arms (treatment & control/comparator) using the 93-item BCT Taxonomy version 1. The original BCT definitions and examples in the taxonomy were adapted to include A&F-specific decision rules and examples. Two additional BCTs (‘Education (unspecified)’ and ‘Feedback (unspecified)’) were added, such that 95 BCTs were considered for coding. Results: In total, 48/95 BCTs (50%) were identified across 360 treatment arms at least once (mean=5.2, SD=2.8, range=1-29 per treatment arm). The most common BCTs were ‘Feedback on behaviour’ (present 89% of the time; e.g., feedback on drug prescribing), ‘Instruction on how to perform the behaviour’ (71%; e.g., issuing a clinical guideline), ‘Social comparison’ (52%; e.g., feedback on performance of peers), ‘Credible source’ (41%; e.g., endorsements from respected professional body), and ‘Education (unspecified)’ (31%; e.g., giving a lecture to staff). The 287 control/comparator arms contained on average 3.0 BCTs (SD=2.4, range=1-15), of which the most common were identical to those identified in treatment arms. Conclusions: A&F interventions to improve healthcare professional practice include a moderate range of BCTs, focusing predominantly on providing behavioural feedback, sharing guidelines, peer comparison data, education, and leveraging credible sources. We encourage the use of our A&F-specific list of BCTs to improve knowledge of what is being delivered in A&F interventions. Our study provides a basis for exploring which BCTs are associated with intervention effectiveness. Trial registrations: N/A