Optimizing quality of cancer care using outcome information

This thesis focuses on the central role of healthcare outcomes in optimizing the quality of
cancer care. The overall aim is to advance both the understanding and the practical
application of outcome-based quality assessment in oncology. Part I explores how outcome information can be used for valid and reliable hospital comparisons, thereby stimulating continuous quality improvement, while Part II examines how outcome predictions can support personalised care and improve shared decision-making. Together, these approaches are essential for achieving outcome-driven, value-based cancer care.

In Chapter 2, we review the current landscape of cancer outcome benchmarking in Europe, focusing on quality indicators and the methodology of case-mix adjustment models. In Chapter 3, we develop a case-mix adjustment model for an important outcome indicator in the NBCA: complications after surgery. Since case-mix adjustment alone does not ensure the validity and reliability of an indicator, we introduce a structured framework to evaluate quality indicators based on feasibility, discriminative ability, validity, and reliability (Chapter 4), and apply it to both breast cancer (a high-incidence cancer) and oral cavity cancer (a low-incidence cancer; Chapter 5).

In Chapter 6, we externally validate the updated PREDICT tool (version 3.1) for supporting clinical decision-making in Dutch and Swedish breast cancer patients, with a focus on lobular breast cancer and younger patients, two groups for whom outcomes are harder to predict and more uncertain. In Chapter 7, we compare version 3.1 of the PREDICT tool with the previous version (2.2), using data from the Dutch population and 36 clinically relevant subgroups. Finally, in Chapter 8, we develop models to predict trends in health-related quality of life (HRQoL) after breast cancer surgery and reconstruction, using 15 different HRQoL outcomes, aiming to better support patients and clinicians in managing expectations during SDM.