Patient Preferences for Metastatic Colorectal Cancer Treatment: A Multi-method Approach Using Discrete Choice Experiments and Best-Worst Scaling.

Background: Treatment decisions for metastatic colorectal cancer (mCRC) require patients to balance survival benefits, health-related quality of life (HRQoL), and potential risks of side effects while also factoring in their own preferences for different treatment options. Despite growing interest, quantitative patient preferences are not yet integrated into health technology assessments (HTAs) for drug reimbursement recommendations.

Objectives: The Colorectal Cancer Canada's Patient Values Project aims to explore approaches to incorporate quantitative patient preferences into cancer treatment HTA decision-making processes. As a first step, we elicited the treatment preferences and risk tolerance of patients with mCRC in Canada using a multi-method approach.

Methods: We developed a preference survey that included two discrete choice experiments (DCEs) and case 1 best-worst scaling (BWS-1) to estimate preferences for mCRC treatments. DCE1 included change in HRQoL and overall survival as attributes, and treatment attributes in DCE2 were administration and frequency, side effects (nausea, pain, diarrhea), and progression-free survival. The BWS-1 included 25 attributes of potential mCRC treatment side effects based on the cancer-specific quality-of-life questionnaire. The survey was administered across Canada to patients with mCRC aged ≥ 18 years with a self-reported diagnosis of mCRC through patient organizations, cancer centers, and an online panel. Data were analyzed using mixed logit and latent class models (DCEs) and count-based analysis (BWS-1).

Results: Overall, 127 patients with mCRC completed the full survey (n = 143 fully completed DCE1, n = 108 fully completed DCE2, n = 127 fully completed BWS-1). Relative preferences for the treatment attributes in the study were consistent with the expectation that better clinical outcomes were preferred over worse clinical outcomes. In DCE1, patients valued both overall survival (24 vs. 12 months) and HRQoL (improvement to 90 vs. worsens to 50 out of 100) as almost equally important. In DCE2, patients preferred better outcomes (longer progression-free survival and no side effects) over worse outcomes, with a disutility for oral capsules/pills compared with intravenous infusions. Significant preference heterogeneity was observed depending on experiences with CRC treatments, treatment side effects, and health status. In the BWS-1, "need help with eating, dressing, washing yourself or using the toilet", "vomiting", and "pain" were ranked as the least and "need to rest", "trouble doing strenuous activities", and "feel tired" as the most tolerable side effects.

Conclusions: This study highlights the value of a multi-method approach in comprehensively assessing treatment preferences and risk tolerance in mCRC. By triangulating multiple preference-elicitation methods, our findings offer a more robust foundation for integrating patient perspectives into Canada's HTA framework. These results will inform the next step of the Colorectal Cancer Canada's Patient Values Project, which aims to explore approaches to explicitly incorporate patient preferences alongside clinical and economic evidence into the cancer treatment HTA decision-making process in Canada.