Methodological Contributions to Guideline Development in Oncology

Abstract

Clinical practice guidelines are used to standardize care provided by clinicians and are based on systematic review of available evidence. Due to the rapid advancement in therapies and need for more up to date guidance, living guidelines were developed to provide continuously updated, methodologically robust recommendations that keep pace with rapidly emerging evidence. In oncology, where disease burden is high and new data on therapies, and technologies evolve quickly, the living guideline model offers a promising approach for ensuring that guidance remains current, trustworthy, and implementable. Despite the importance of clinical guidelines and the growing interest in the living guideline model, important methodological questions remain regarding how best to evaluate guideline adherence, address implementation challenges, engage guideline panels, and integrate emerging tools such as artificial intelligence and automation in the development process. This thesis contributes to the methodological foundations of guideline development in oncology through four interconnected studies. The first chapter focuses on examining adherence to oncology clinical practice guidelines and its impact on patient outcomes with the use of a systematic review and meta-analysis. By quantifying adherence patterns and associated health effects, this chapter highlights the need for more evaluation studies on the uptake of guidelines and the importance of improving guideline implementation strategies to improve guideline adherence and patient’s health outcomes. The second chapter evaluates the Panel Subgroup (PSG) method as a structured approach to engaging multidisciplinary panels during guideline development. Findings demonstrate how PSGs can improve efficiency, inclusiveness, and methodological rigor which are elements critical to sustaining guideline processes over time. The third chapter explores barriers and facilitators influencing adoption of the living guideline framework within oncology by employing a mixed-methods design. This work deepens understanding of the contextual and organizational factors that shape successful development of living guidelines. The fourth chapter assessed current frameworks, methods, and applications of artificial intelligence and automation in living guideline development. This chapter maps the emerging methodological landscape and identifies opportunities for integrating AI and automation to support evidence surveillance, workflow optimization, and decision-making. Together, these four studies offer complementary methodological insights that advance the science and practice of guideline development in oncology. This thesis provides evidence, and practical considerations aimed at supporting organizations and guideline developers as they design, implement, and sustain guideline processes in a rapidly evolving clinical environment.