Preferences heterogeneity and willingness to pay for oral treatment services among Chinese adults: a latent class discrete choice experiment

To examine preference heterogeneity and willingness to pay (WTP) for key attributes of oral treatment services among Chinese adults. A discrete choice experiment (DCE) was administered through face-to-face surveys among Chinese adults aged 18 years or older. Surveys were conducted from December 1, 2025, to February 1, 2026, using multistage sampling across eastern, central, and western China. Each choice task included two hypothetical treatment alternatives and an opt-out option. Data were analyzed using mixed logit models, latent class models, and multinomial logistic regression. Sensitivity analyses were conducted to assess the robustness of the main findings. A total of 1,316 participants were included. Higher treatment cost reduced choice utility ( β = − 0.001, P < 0.001), whereas lower pain, shorter healing duration, fewer visits, favorable dietary-recovery levels, preferred treatment settings, and higher dentist qualification increased preference (all P < 0.05). Willingness To Pay (WTP) estimates were highest for dentist qualification, reaching 589.14 RMB for chief physicians, which was interpreted as a marginal trade-off value within the experimental design. The latent class model identified five preference groups: rapid dietary-recovery-oriented (27.46%), dietary-recovery-time-sensitive (10.53%), pain-outcome-oriented (24.02%), provider-qualification-oriented (16.15%), and cost-burden-oriented (21.85%). Class membership was associated with education, self-rated oral health, income in some comparisons, and sex. Chinese adults showed marked heterogeneity in stated preferences for oral treatment services. Treatment experience, recovery, care quality, and affordability jointly shaped oral treatment choices. These findings suggest that oral healthcare services should be designed around patient-valued attribute combinations and tailored to preference-based subgroups. WTP estimates should be interpreted cautiously because they were derived from hypothetical choices within the experimental cost range.