From CXR to ultrasound: a multi-framework approach to advancing De-implementation in acute care

Efforts such as Choosing Wisely have raised awareness about low-value care, yet entrenched practices persist, particularly in acute care settings where clinician behavior is shaped by workflow defaults, training norms, and institutional policies. Routine chest radiography (CXR) after central venous catheter (CVC) placement continues despite strong evidence supporting point-of-care ultrasound (POCUS) as a safe and efficient alternative. We applied a multi-framework, theory-informed approach to diagnose behavioral drivers sustaining CXR use and to design targeted de-implementation strategies.

We conducted focus groups and interviews with emergency and critical care clinicians within a single integrated academic health system. Transcripts were deductively coded using the Consolidated Framework for Implementation Research (CFIR) and mapped to the COM-B model (capability, opportunity, motivation). A causal map was developed to represent behavioral drivers. Using the Behaviour Change Wheel (BCW), we selected aligned intervention functions and specified mechanisms of action using the Behavior Change Technique Taxonomy (BCTT). Strategies were organized within an Implementation Research Logic Model (IRLM).

Three interrelated themes emerged: (1) variable self-efficacy for POCUS confirmation (capability), (2) entrenched organizational norms and workflow defaults favoring CXR (opportunity), and (3) behavioral inertia reinforced by habit and perceived safety (motivation). These determinants informed seven candidate strategies, including competency-based ultrasound training, audit and feedback with peer comparison, policy supported clinical algorithms, workflow redesign, peer champions, and structured adaptation cycles. Fourteen behavior change techniques were specified to clarify mechanisms of action. The IRLM linked determinants, strategies, mechanisms, and anticipated outcomes, providing a transparent pathway from qualitative diagnosis to strategy specification.

Integrating CFIR, COM-B, BCW, and BCTT enabled systematic translation of contextual and behavioral insights into mechanism-based de-implementation strategies. This structured approach enhances transparency in strategy specification and offers a replicable roadmap for reducing low-value care in complex clinical environments. Future work will pilot and evaluate these strategies to support broader de-implementation efforts.