Tailoring implementation interventions of different order in infection prevention and control: A cascadic logic model (IPC-CASCADE).

Implementation interventions in infection prevention and control (IPC) differ by recipients. The two target groups are healthcare workers directly involved in patient care ("frontline") and IPC professionals as proxy agents, that is, implementation support practitioners. While both types of implementation interventions aim to promote compliance with clinical interventions to prevent healthcare-associated infections (HAI), their tailoring may be vastly different, for example, due to different behavioural outcomes. Additionally, IPC teams, as recipients of empowering tailored interventions, are under-researched. To overcome this gap and improve conceptual clarity, we proposed a cascadic logic model for tailored IPC interventions (IPC-CASCADE). In the model, we distinguished between interventions by IPC professionals targeting clinicians and those targeting IPC professionals (first- and second-order implementation interventions, respectively). Tailoring implies selecting behaviour change techniques matched to prospectively-assessed determinants of either clinician compliance (in first-order interventions) or interventions by IPC professionals for frontline workers (in second-order interventions). This interventional cascade is embedded in the prevailing healthcare system. IPC-CASCADE is horizontally structured over time and vertically structured by hierarchy or leadership roles. IPC-CASCADE aims to highlight the potential of increasing the impact of tailored interventions by IPC professionals for clinicians (to improve their compliance) via tailored interventions for IPC professionals (to improve their work as proxy agents). It underlines the links that IPC professionals define between macro contexts (healthcare and hospitals) and frontline workers in HAI prevention. It is specific, i.e., "tailored" to IPC, and expected to assist implementation science to better conceptualise tailoring.