The trial, an embedded mixed-methods cluster randomized controlled trial carried out from 2021 to 2023, enrolled 141 village clinics and their doctors, of whom 128, or 90.8 percent, completed follow-up. Villages were randomized into one control condition and four intervention groups. All participating village doctors received general clinical training, but the intervention arms differed in what came next: one group gained telemedicine access and training through the kiosk; a second received telemedicine plus demand-side marketing aimed at attracting patients; a third received telemedicine plus monetary incentives tied to platform use; and a fourth received the full package of telemedicine, marketing, and incentives together. Village doctors used the kiosk to create digital patient profiles and to contact upper-level hospital physicians through one-click consultations or formal referrals.

The trial’s primary outcomes were telemedicine usage, primary care utilization, and clinic revenue, measured through administrative platform records and structured surveys. The quantitative data were complemented by in-depth qualitative interviews with 20 village doctors, designed to probe why the technology was or was not embraced. This embedded mixed-methods design allowed the researchers to pair statistical estimates of program effects with a granular, ground-level account of the behavioral barriers that shaped those estimates, an approach increasingly favored for implementation trials in complex health systems.

The headline finding is a stark gap between engagement and meaningful use. Among the intervention groups, 81.9 percent of doctors logged into the telemedicine platform at least once, suggesting that the technology itself was accessible and that doctors were willing to explore it. Yet only 14.9 percent ever conducted a one-click consultation with an upper-level physician. Logging in, in other words, did not translate into clinical practice. The kiosk became a tool that village doctors opened but rarely used for the purpose that mattered most: connecting their patients to hospital expertise in real time.

The combination of demand-side marketing and monetary incentives produced marginal, statistically borderline increases in platform logins, with an average treatment effect of 0.71 (95 percent confidence interval: -0.02 to 1.44; p = 0.056), and in the creation of patient files, with an effect of 3.15 (95 percent CI: -0.47 to 6.77; p = 0.088). Neither estimate reached conventional thresholds of statistical significance, and, crucially, neither the combined package nor any individual component succeeded in increasing the number of telemedicine consultations that doctors conducted. For the trial’s investigators, this null result on consultations was the study’s most consequential and most instructive finding, exposing the limits of financial and promotional levers when the underlying obstacle is attitudinal rather than material.

Where the interventions did work was in patient flow and income. Demand-side marketing alone increased primary care utilization, with an average treatment effect of 0.40 (95 percent CI: 0.03 to 0.76; p = 0.033), and the combined marketing-plus-incentive package produced a larger effect of 0.73 (95 percent CI: 0.28 to 1.17; p = 0.001). Net medical income rose across multiple arms: marketing alone increased it by an average treatment effect of 12.63 (95 percent CI: 1.83 to 23.43; p = 0.022), incentives alone by 12.53 (95 percent CI: 3.17 to 21.90; p = 0.009), and the combined interventions by 12.91 (95 percent CI: 6.32 to 19.49; p < 0.001). These gains indicate that the kiosk program, even without heavy teleconsultation traffic, made village clinics busier and more financially viable, a meaningful outcome in a health system where weak village-level care drives patients toward crowded county hospitals.

The qualitative interviews illuminated the mechanisms behind the numbers. Village doctors expressed mistrust of online care, doubting both the clinical reliability of remote consultations and their acceptance among patients who often prefer the familiarity of face-to-face interaction with a provider they know personally. Perhaps most strikingly, doctors reported a preference for instant messaging as an alternative channel, suggesting that when rural clinicians do reach out to hospital colleagues, they gravitate toward the informal, low-friction tools already embedded in their social and professional networks rather than toward the purpose-built telemedicine platform. Any digital health strategy for rural settings, the authors conclude, must account for these relational dynamics and integrate with the communication tools clinicians actually use.

The study’s design merits attention from anyone weighing its implications. By randomizing villages rather than individual doctors, the trial guarded against contamination between intervention and control conditions, an essential safeguard in close-knit rural communities where doctors talk to one another. The embedded qualitative component ensured that the behavioral insights explaining the null teleconsultation result were generated systematically rather than left to speculation. High retention, with more than 90 percent of enrolled clinics completing follow-up, strengthens the credibility of the effect estimates, although the modest margins on some outcomes, including the borderline login and file-creation effects, warrant caution in interpretation.

The broader significance of the trial extends well beyond China. Digital health initiatives targeting rural and underserved populations have proliferated worldwide, frequently justified by the assumption that connectivity and equipment are the binding constraints on telemedicine adoption. This study demonstrates that assumption can be wrong: the binding constraint may be trust, in the technology, in remote providers, and in whether online interactions can replicate the relational foundation of rural primary care. Marketing to patients and paying doctors moved the needle on utilization and income, suggesting demand and supply can be stimulated, but neither lever converted logged-in doctors into consulting doctors. For policymakers and implementers, the lesson is that investment in digital infrastructure must be paired with strategies that build clinical confidence, align telemedicine with existing communication habits, and address the deep-seated distrust that keeps rural patients and providers tethered to traditional channels.

The findings also carry a constructive message about the economics of rural clinics. Even when the flagship telemedicine function went largely unused, the kiosk program measurably increased the number of patients seeking care at village clinics and the net income those clinics earned, with the largest and most statistically robust gains occurring where marketing and incentives were combined. In fragmented rural health systems, redirecting even modest patient volume back to village-level care can improve continuity, reduce unnecessary hospital congestion, and shore up the financial base of frontline providers. Future iterations of the kiosk program, the trial suggests, should retain the demand- and supply-side supports that proved effective while redesigning the teleconsultation workflow around trust-building measures and integration with the instant messaging platforms that village doctors already prefer. As countries everywhere race to digitize primary care, this trial stands as a rigorous, cautionary, and ultimately practical guide: technology can open the door, but only trust walks rural clinicians and their patients through it.

Subject of Research: Evaluation of a digital health kiosk program on telemedicine use, primary care utilization, and clinic income in rural Chinese village clinics

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

Article References: Cheng, W., Zhang, Z., Du, Y., Dong, X., Liu, Y., Zhou, Q., Jing, F., Xu, Z., Xie, Y., Wu, D., Xue, H., Hoelzer, S., Yip, W., Tian, J., Ma, X., Sylvia, S., & Tang, W. (2026). The evaluation of a digital health kiosk on health care delivery in rural China: embedded mixed-methods cluster randomized controlled trial. BMC Medicine. https://doi.org/10.1186/s12916-026-05155-6

Image Credits: AI Generated

DOI: 10.1186/s12916-026-05155-6

Keywords: digital health, telemedicine, rural health, primary care, village doctors, cluster randomized controlled trial, monetary incentives, demand-side marketing, healthcare utilization, healthcare access, China, BMC Medicine

In early 2020, the Centers for Medicare & Medicaid Services (CMS) implemented pivotal policy changes in response to the public health emergency. These reforms granted telemedicine visits payment parity with in-person appointments, removed geographic restrictions that previously limited eligibility, and waived out-of-pocket cost-sharing for patients. These flexibilities, originally enacted as temporary measures, transformed healthcare delivery by eliminating many of the practical and economic barriers hindering widespread telehealth adoption. As the pandemic waned, legislative bodies have debated whether to embed these changes into permanent policy frameworks, amid unresolved questions about telemedicine’s long-term effects on healthcare systems.

The investigative team, led by Dr. John N. Mafi, Associate Professor-in-Residence at UCLA’s Division of General Internal Medicine and Health Services Research, sought to systematically quantify telemedicine’s influence across a representative cohort of more than three million adults. By leveraging the MedInsight database — a robust repository containing multi-payer claims data — the researchers examined records of individuals continuously enrolled in Medicare fee-for-service, Medicare Advantage, dual-eligible Medicaid plans, and commercial insurance over a five-year horizon from January 1, 2019, to December 31, 2023. This extensive dataset enabled population-level insights encompassing diverse insurance structures and demographic groups.

Contrary to expectations forecasting a surge in medical visits and spending due to increased telehealth availability, the results indicate that telemedicine’s expansion neither generated statistically significant increases nor decreases in overall ambulatory care utilization or expenditures. The analysis revealed a nominal 2.4% decline in visits and a marginal 0.5% reduction in spending, but these trends “crossed the null” threshold meaning the observed changes could plausibly be attributed to chance variation rather than true systemic effects. Crucially, the confidence intervals effectively ruled out substantial increases or decreases, implying any telemedicine-driven shifts in healthcare consumption or costs were modest at best.

Subgroup analyses probing demographic and geographic heterogeneity yielded similarly nuanced findings. Urban populations manifested a slight decrease in visits and spending, while rural counterparts experienced a small uptick in both metrics. Spending among Medicaid beneficiaries and dually eligible individuals exhibited minor non-significant reductions, paralleling similar patterns within Medicare Advantage enrollees and socially vulnerable cohorts. Conversely, commercially insured patients and the least socially vulnerable groups also displayed nominal spending increases without statistical significance. Collectively, these patterns suggest that telemedicine’s integration acts more as a substitute modality rather than an additive service amplifying healthcare utilization.

These findings carry substantial implications amid ongoing policy debates. Advocates heralded telemedicine as a promising solution to longstanding disparities in healthcare access, especially for populations facing geographic or logistical barriers. However, the data imply that telemedicine may predominantly replace in-person visits instead of broadening the overall reach of ambulatory care services, tempering expectations that it will drastically close access gaps. Simultaneously, the absence of escalating expenditures alleviates fears that telehealth expansion could nonsustainably inflate healthcare costs on a national scale.

Notwithstanding these insights, the study acknowledges several limitations inherent to its observational design and aggregated methodology. The expansive dataset, while representative, cannot capture granular individual-level variations or isolate causal relationships definitively. Moreover, uninsured populations were excluded, potentially skewing generalizability. The temporal scope concluding in late 2023 reflects a relatively early phase in telemedicine’s evolving utilization trajectory, underscoring a need for continued longitudinal research to ascertain enduring impacts on clinical quality, health outcomes, and cost containment.

Senior co-author Dr. Katherine Kahn, Distinguished Professor of Medicine at the David Geffen School of Medicine at UCLA and senior natural scientist at RAND Corporation, emphasizes the preliminary nature of these findings. She notes that telemedicine adoption appears to have stabilized into a new equilibrium following its precipitous rise during the pandemic’s acute phase, but stresses the necessity for vigilant policymaker surveillance and further study. Ongoing research should interrogate differential effects across diverse populations to ensure telehealth initiatives equitably enhance healthcare delivery without exacerbating existing disparities.

The study’s multidisciplinary authorship team combined expertise from UCLA, RAND, MedInsight (Milliman Inc.), Virginia Tech, and University of Michigan, reflecting its comprehensive analytic approach. Funding support stemmed from Arnold Ventures alongside significant backing by the National Institutes of Health and the National Institute on Aging, underscoring the study’s scientific rigor and societal relevance.

In summation, this extensive analysis compellingly suggests that telemedicine’s rapid integration into U.S. ambulatory care systems following the COVID-19 pandemic did not precipitate the feared surge in visit volume or medical spending through 2023. While telehealth has undeniably transformed healthcare delivery modalities, its effect thus far appears to be substitutive rather than additive, stabilizing rather than escalating utilization patterns. These findings provide a critical evidence base as policy frameworks governing telemedicine flexibilities approach a 2027 expiration, facilitating informed debate on how best to balance innovation, access, and cost containment in healthcare’s digital future.

Subject of Research: People

Article Title: Telemedicine Adoption, US Ambulatory Visits, and Total Medical Spending, 2019-2023

News Publication Date: 11-May-2026

Web References:

• https://jamanetwork.com/journals/jamanetworkopen/fullarticle/10.1001/jamanetworkopen.2026.11835

References:
Mafi, J.N., Kahn, K., et al. (2026). Telemedicine Adoption, US Ambulatory Visits, and Total Medical Spending, 2019-2023. JAMA Network Open. DOI: 10.1001/jamanetworkopen.2026.11835

Keywords: Telemedicine, Healthcare utilization, Medical spending, Ambulatory care, COVID-19 pandemic, Medicare, Medicaid, Health policy, Telehealth adoption, Health disparities