Patient Preferences for Cancer Care Within Hub-and-Spoke Networks: A Discrete Choice Experiment.

Abstract

Hub-and-spoke cancer networks are increasingly used to extend specialty expertise to rural populations; however, patient preferences for how surgical and nonsurgical cancer care should be delivered within these networks are poorly defined. We quantified rural patient preferences and identified distinct preference phenotypes to inform network design. A cross-sectional discrete choice experiment was administered to adult patients at a rural primary care clinic affiliated with an academic cancer center (November 2023 to May 2024). Respondents completed 30 forced-choice tasks evaluating trade-offs among travel time, provider specialization, visit type (in-person vs telemedicine), expert case review (direct, tumor board, or quality oversight), and surgical team experience. Effects-coded conditional logit models generated preference weights. Latent class analysis segmented respondents into preference phenotypes. Among 171 respondents (median age 52 years; 43% Black or people of color), surgical team experience (34%) and surgeon specialization (25%) were most important for surgical care. Fifty-three percent of respondents prioritized a specialist surgeon and a high-volume team and were willing to travel; 39% prioritized in-person visits and accepted indirect case review; and 8% prioritized limiting travel. For nonsurgical care, visit type (35%), travel (27%), and provider type (27%) were the most important factors. Four phenotypes emerged: preference for specialist oncologist (29%), direct expert review (28%), in-person visits (28%), and limiting travel (18%). Rural patients exhibit heterogeneous and treatment-specific preference phenotypes within hub-and-spoke cancer networks. Preference elicitation may enable networks to align specialist deployment, telemedicine infrastructure, multidisciplinary case review, and resource allocation with patient priorities.