Understanding effects in reviews of implementation interventions using the Theoretical Domains Framework

Behavioural theory can be used to better understand the effects of behaviour change interventions targeting healthcare professional behaviour to improve quality of care. However, the explicit use of theory is rarely reported despite interventions inevitably involving at least an implicit idea of what factors to target to implement change. There is a quality of care gap in the post-fracture investigation (bone mineral density (BMD) scanning) and management (bisphosphonate prescription) of patients at risk of osteoporosis. We aimed to use the Theoretical Domains Framework (TDF) within a systematic review of interventions to improve quality of care in post-fracture investigation. Our objectives were to explore which theoretical factors the interventions in the review may have been targeting and how this might be related to the size of the effect on rates of BMD scanning and osteoporosis treatment with bisphosphonate medication.

A behavioural scientist and a clinician independently coded TDF domains in intervention and control groups. Quantitative analyses explored the relationship between intervention effect size and total number of domains targeted, and as number of different domains targeted.

Nine randomised controlled trials (RCTs) (10 interventions) were analysed. The five theoretical domains most frequently coded as being targeted by the interventions in the review included "memory, attention and decision processes", "knowledge", "environmental context and resources", "social influences" and "beliefs about consequences". Each intervention targeted a combination of at least four of these five domains. Analyses identified an inverse relationship between both number of times and number of different domains coded and the effect size for BMD scanning but not for bisphosphonate prescription, suggesting that the more domains the intervention targeted, the lower the observed effect size.

When explicit use of theory to inform interventions is absent, it is possible to retrospectively identify the likely targeted factors using theoretical frameworks such as the TDF. In osteoporosis management, this suggested that several likely determinants of healthcare professional behaviour appear not yet to have been considered in implementation interventions. This approach may serve as a useful basis for using theory-based frameworks such as the TDF to retrospectively identify targeted factors within systematic reviews of implementation interventions in other implementation contexts.