ESTRO 2024 - Abstract Book

S2765

Interdisciplinary - Health economics & health services research

ESTRO 2024

was 52% (95% CI: 51%-53%) and the actual RTU was 40% (95% CI: 37%-44%). Only 4.4% of patients aged 80+ received 70Gy. The completion rate for a 70Gy course of radiotherapy for patients aged 80+ was 92%. The ED presentation rate was similar for all age groups.

Conclusion:

For patients aged 80+ years, the actual RTU was less than the optimal RTU. Further research is required to determine reasons other than excess toxicity for the low receipt of curative intent schedules, and for all age groups, reasons for the underutilisation of RT.

Keywords: Radiotherapy utilisation, Older patients, Cancer

304

Mini-Oral

A research-based implementation strategy for AI in radiotherapy: can implementation science help?

Rachelle Swart 1 , Liesbeth Boersma 1 , Rianne Fijten 1 , Wouter van Elmpt 1 , Paul Cremers 1 , Maria Jacobs 2

1 Maastricht University Medical Centre+, Department of Radiation Oncology (Maastro), GROW School for Oncology and Reproduction, Maastricht, Netherlands. 2 Tilburg University, Tilburg School of Economics and Management, Tilburg, Netherlands

Purpose/Objective:

In recent years, there has been significant interest in artificial intelligence (AI) and its role in radiotherapy (RT) (1 – 3). AI is being developed, applied, and evaluated for automation and optimization of the workflow and has also the ability to develop predictive models to support personalized treatment choices (4 – 6). Recommendations for the clinical implementation of AI in RT exist (2) in literature, with a focus on the commissioning, clinical implementation and daily use of the AI model and the daily quality assurance (QA). However, the broad application of AI in RT remains limited (1, 7 – 9) due to significant implementation challenges (1 – 3,7 – 9). Validated insights from implementation science hold the promise of addressing these challenges in RT. The purpose of this study is twofold: 1) to formulate a research-based implementation strategy that improves the clinical implementation of AI in RT and 2) to assess the acceptability, appropriateness and feasibility of the proposed implementation strategy in RT.

Material/Methods:

We developed a research-based implementation strategy, initially optimized for a large academic radiotherapy department in the Netherlands. The strategy focuses on implementing two types of AI: AI for increasing efficiency and quality of care; and AI for developing prediction models to support personalised treatment choices. The strategy was built following the following steps: a) Identifying stakeholders and conducting stakeholder analysis, categorizing them by influence, interest, and relation characteristics; b) Identifying barriers and facilitators through