The evaluation of a digital health kiosk on health care delivery in rural China: embedded mixed-methods cluster randomized controlled trial

Rural primary health care (PHC) in China faces persistent challenges, including limited workforce capacity, fragmented information systems, and patient distrust of village-level care. Although digital health may strengthen PHC delivery, evidence on effective implementation strategies remains limited. We evaluated the effects of a multicomponent provider-to-provider digital health kiosk program including telemedicine platform access, training, demand-side marketing, and monetary incentives on provider engagement, primary care utilization, and clinic income. We conducted an embedded mixed-methods cluster randomized controlled trial from 2021 to 2023. Quantitative data included administrative platform records and surveys. 141 village clinics and doctors were enrolled, with 128 (90.8%) completing follow-up. Villages were randomized into 1 control and 4 intervention groups. Village doctors received (1) general clinical training only (control), (2) telemedicine access and training, (3) telemedicine plus demand-side marketing, (4) telemedicine plus monetary incentives, or (5) telemedicine plus both marketing and incentives. Village doctors used the kiosk to create patient profiles and contact upper-level hospital physicians through one-click or referral consultations. Primary outcomes were telemedicine usage, primary care utilization, and clinic revenue. Qualitative interviews were conducted with 20 village doctors. Among intervention groups, 81.9% of doctors logged into the platform, but only 14.9% conducted one-click consultations. Combined marketing and incentives marginally increased logins (ATE = 0.71; 95% CI: -0.02, 1.44; p = 0.056) and file creation (ATE = 3.15; 95% CI: -0.47, 6.77; p = 0.088) but did not increase consultations. Marketing alone (ATE = 0.40; 95% CI: 0.03,0.76; p = 0.033) and combined interventions (ATE = 0.73; 95% CI: 0.28, 1.17; p = 0.001) improved primary care utilization. Net medical income increased with marketing (ATE = 12.63; 95% CI: 1.83, 23.43; p = 0.022), incentives (ATE = 12.53; 95% CI: 3.17, 21.90; p = 0.009), and combined interventions (ATE = 12.91; 95% CI: 6.32, 19.49; p < 0.001). Qualitative findings highlighted mistrust of online care and preference for instant messaging as an alternative platform. Marketing and monetary incentives improved utilization and income but did not sustain telemedicine consultations. Trust and integration with social networks are essential for sustained adoption in rural settings. ChiCTR2100053872.