S1428
Interdisciplinary - Health economics & health services research
ESTRO 2026
into a comprehensive implementation plan by the end of Q4 2025, preceding clinical rollout in Q1 2026. References: 1. Reardon, C.M., et al., The Consolidated Framework for Implementation Research (CFIR) User Guide: a five- step guide for conducting implementation research using the framework. 20252. Reed, M.S., et al., Who's in and why? A typology of stakeholder analysis methods for natural resource management. 20093. Damschroder, L.J., C.M. Reardon, M.A.O. Widerquist, and J. Lowery, The updated Consolidated Framework for Implementation Research based on user feedback. 20224. Powell, B.J., et al., A refined compilation of implementation strategies: results from the Expert Recommendations for Implementing Change (ERIC) project. 20155. Shea, C.M., et al., Organizational readiness for implementing change: a psychometric assessment of a new measure. 2014 Keywords: Artificial Intelligence, Implementation Science Digital Poster Highlight 4398 Appraising the value of radiotherapy innovations: results from a modified TOPSIS framework integrating Delphi and AI - assisted evidence Gloria Brigiari 1 , Miet Vandemaele 2 , Ester Rosa 1 , Dario Gregori 1 , Ajay Aggarwal 3,4 , Yolande Lievens 2,5 1 Department of Cardiac, Thoracic, Vascular Sciences and Public Health, Unit of Biostatistics, Epidemiology and Public Health, University of Padova, Padova, Italy. 2 Department of Human Structure and Repair, Radiation Oncology, Ghent University, Ghent, Belgium. 3 Institute of Cancer Policy, King's College London, London, United Kingdom. 4 Faculty of Public Health and Policy, London School of Hygiene and Tropical Medicine, London, United Kingdom. 5 Radiation Oncology Department, Ghent University Hospital, Ghent, Belgium Purpose/Objective: Appraisal of new radiotherapy interventions requires methods that are transparent, reproducible, and able to balance multidimensional outcomes with the strength of supporting evidence. We propose a modified Technique for Order Preference by Similarity to Ideal Solution (TOPSIS) tailored for innovation appraisal in radiation oncology (RO). As a pilot investigation, we use three prostate cancer innovations: hypofractionation, high - dose - rate brachytherapy, and proton radiotherapy. Material/Methods: For each innovation, relevant literature was retrieved using the ESTRO–VBRO bibliometric pipeline (2), and interventions were classified with the VBRO categorization system (3). Essential endpoints
organisational readiness, and acceptance influencing AI-assistant implementation to inform a context- specific implementation plan supporting adoption and
workflow integration. Material/Methods:
A five-step CFIR (Consolidated Framework for Implementation Research)-guided approach structured this prospective qualitative study in a Dutch radiotherapy centre [1]. Key stakeholders were identified through a stakeholder analysis using an organogram and participated in semi-structured interviews exploring usability, acceptability, workflow integration, and contextual determinants [2]. Data were deductively coded to constructs from the updated CFIR using Atlas.ti, integrating COM-B (Capability, Opportunity, Motivation, Behaviour) and TDF (Theoretical Domains Framework) components [3]. The CFIR–ERIC (Expert Recommendations for Implementing Change) matching tool was applied to identify targeted implementation strategies [4]. Participants and the radiotherapy patient panel were asked to complete ORIC (Organisational Readiness for Implementing Change) and AI-acceptability surveys, respectively [5]. Analyses of both surveys are ongoing; current findings are based on interview data. Results: The stakeholder analysis identified 23 professionals across clinical, technical, and organisational roles. Within the Innovation domain, radiation oncologists and physician assistants (n=14) cited ease of use and clinical value as key facilitators, while concerns about overhyping (n=7) and regional dialect limitations (n=11) reflected issues of Evidence-Base, Source, and Adaptability. The Process domain highlighted a preference for iterative, low-burden evaluation (e.g., think-aloud sessions) and the role of local champions in supporting engagement (Reflecting and Evaluating, Engaging). Within the Outer Setting, stakeholders noted legal ambiguity across MDR, EU AI Act, and GDPR frameworks (Policies and Laws). The Inner Setting emphasised the need for a structured legal– quality assurance workflow to enable readiness and alignment (Culture – Learning-Centeredness, Mission Alignment). Implementation strategies included identifying and preparing clinical champions, promoting system adaptability to regional dialects, conducting thorough legal analysis before implementation, and facilitating QA integration to ensure compliant, sustainable adoption. Conclusion: Preliminary findings suggest that successful AI- assistant implementation in radiotherapy requires a staged, implementation science–based approach led by prepared clinical champions, promoting adaptability and legal readiness, and guided by transparent evaluation. Ongoing readiness and patient-acceptability surveys will refine these insights
Made with FlippingBook - Share PDF online