From predictions to practice: value of prediction models for personalised sarcoma care

Abstract

This thesis aimed to enhance personalised care for soft-tissue sarcoma patients by 1) evaluating the value of prediction models in treatment and follow-up decision making, and 2) assessing quality of life challenges to better align treatments with patients’ needs and preferences. Both approaches show significant potential for enhancing personalised care. Prediction models, in particular, have proven effective in helping clinicians evaluate the added value of adjuvant treatment (e.g., chemotherapy) and communicate this information during clinical consultations— an advantage also appreciated by patients. However, even the best-designed interventions require thoughtful implementation to achieve their full potential. To maximize the effectiveness of prediction models, it is essential to promote multidisciplinary research that integrates clinical context, implementation science, and statistical modelling. This thesis provides valuable initial insights into the application and evaluation of RPMs to support (shared) decision-making and personalise care.