Development and proof-of-concept of a complex intervention to support appropriate imaging for musculoskeletal pain:
Nicole Lindner1, Kristina Buch2, Konrad Hierasimowicz2
1Department of Primary Care, Philipps-Universität Marburg, Karl-von-Frisch-Str. 4, 35043, Marburg, Germany. nicole.lindner@uni-marburg.de.
Background:
Inappropriate diagnostic imaging for musculoskeletal pain is common and causes patient harm and unnecessary costs. Existing interventions frequently target single conditions and stakeholders and show limited integration into routine consultations. This study aimed to develop and test a complex intervention to de-implement low-value care by supporting appropriate imaging decisions for musculoskeletal pain in primary care.
Methods:
We developed Betti ("Better Imaging") following the Medical Research Council framework (MRC) for complex interventions. The development was both theory- and evidence-based and included: specification and iterative refinement of a programme theory (logic model); a comprehensive literature review; qualitative interviews with patients; structured expert feedback; and a proof-of-concept test with general practitioners and patients. The Behaviour-Change-Wheel informed intervention components. We addressed early implementation considerations throughout the development process.
Results:
Findings from the literature review revealed a wide range of intervention approaches and components. Overall effectiveness was mixed. Reductions in imaging were reported more frequently in multicomponent interventions that included both physician-facing components and patient-facing materials. Qualitative interviews with people with musculoskeletal pain highlighted that expectations of primary care consultations are highly individual and shaped by context (e.g., prior experiences). Expert feedback emphasised communication and reassurance to support imaging decisions. Informed by these findings, we developed Betti, multicomponent intervention comprising: (1) a multimedia training module for general practitioners, (2) a clinical decision support system based on guideline recommendations across musculoskeletal pain, and (3) multimedia patient information materials. In the proof-of-concept test, patients and general practitioners perceived Betti as well structured and supportive. However, implementation challenges emerged: patients were not directed to the materials, indicating that Betti was not integrated into the consultation as intended. These findings led us to refine the programme theory, explicitly positioning consultation-integrated delivery and physician-mediated handover explicitly as essential for our program theory.
Conclusions:
Betti is a theory- and evidence-based, stakeholder-developed intervention. Early findings show high acceptability but underscore consultation-integrated delivery as critical. The study adds transferable implementation insights for de-implementing low-value imaging by specifying mechanisms, determinants, and strategy choices beyond tool use, informing further refinement and future feasibility and effectiveness-implementation evaluation.
