How do treatment preferences of patients with multiple myeloma compare with those of family caregivers and physicians? evidence from a discrete choice experiment in China

Treatment decisions in multiple myeloma (MM) inherently require balancing multiple competing dimensions. In China, the alignment of preferences among patients, family caregivers, and physicians on these dimensions remains unexplored. A multicenter discrete choice experiment was conducted among 369 participants (143 patients, 129 family caregivers, and 97 hematologists) recruited from two tertiary hospitals in China. Five treatment attributes were examined, namely progression-free survival (PFS), risk of serious infection, fatigue, pain, and monthly medical cost. Preference weights and marginal willingness-to-pay estimates were obtained using mixed logit models, relative attribute importance was calculated, latent class analysis was employed to explore preference heterogeneity, and scenario analyses were conducted to illustrate model-predicted choice probabilities under hypothetical treatment profiles. After dominance test screening, 126 patients, 121 family caregivers, and 83 hematologists were included in the analysis. PFS showed the highest relative importance among the three stakeholder groups, although its contribution differed across groups, with patients, family caregivers, and hematologists showing relative importance estimates of 28.78%, 37.69%, and 39.97%, respectively. Patients showed higher relative importance estimates for pain (27.45%) and monthly medical cost (22.55%) than family caregivers (20.80% and 15.92%) and hematologists (22.01% and 4.40%), respectively. The cost coefficient was not statistically significant in the hematologist model. Latent class analysis revealed three distinct preference profiles among patients, with the largest class (42.6%) characterized by a pain- and cost-focused pattern where PFS was not a significant driver. Scenario analyses provided supplementary illustrations of preference patterns under hypothetical treatment profiles. Differences in treatment preferences among patients, family caregivers, and hematologists highlight the importance of incorporating patient preferences into shared decision-making and tailoring treatment discussions to individual priorities. Not applicable.