Common Barriers to Implementation Across Contexts: Evidence to inform the selection of implementation strategies

Background: The implementation of evidence-informed interventions is required to strengthen health systems and improve health outcomes. Identifying implementation barriers underpins evidence-based approaches to do so, but this process is considered complex and time consuming in practice. Understanding whether certain barriers are more common across contexts may guide the development of more effective implementation strategies when comprehensive primary data collection to assess barriers is not feasible. Methods: We conducted a pooled analysis of barrier data from studies that quantitatively assessed implementation barriers using the Theoretical Domains Framework (TDF) survey. To assess barrier frequency aligned to each TDF domain, we calculated the proportion of studies where domain scores were less than four on a standardised 5-point Likert scale. To describe their "strength" we pooled data across studies and reported mean domains scores (lower domain scores represent stronger perceived barriers). Subgroup analyses using pooled domains scores were undertaken to examine differences by population, intervention and geographic characteristics. Results: Data from 42 studies published 2012 to 2024, with a combined sample of 9,809 participants were included in the analysis. In 10 of 14 TDF domains, both mean and median scores were 4 or below, indicating that they were typically perceived as barriers. Four domains had a mean score of 4 or below in >80% of all studies that assessed them - "reinforcement", "environmental context and resources", "social influences" and "behavioural regulation". TDF domains with the lowest scores (representing the strongest barriers), were "environmental context and resource"; "behavioural regulation" and "social influences." Few differences (four of 42 statistical comparisons) were found between TDF domain scores and population, intervention and geographic factors. Conclusions: This study identified a set of barriers that appear to be common, and consistent in their perceived strength across a range of population groups, intervention types and geographic localities. The findings provide a basis for those undertaking efforts to improve implementation of evidence-informed health interventions to anticipate types of barriers they may encounter and so, likely strategies that may be needed to address these. This may be beneficial in resource contexts where primary data collection for more comprehensive barrier assessments may not be feasible.