Village doctor-led integrated care for older adults with diabetes in rural China: a cluster-randomized pilot trial Authors: Qingqing Zhang, Chuanmeng Zhang, Jing Zheng, Hong Xu, Si Sun, Suyun Jiang, Peng Gao, Yucheng Wu, Ming Chu Published: 2026 ABSTRACT BackgroundIn rural China, diabetes management is still far from optimal owing to limited resources, low compliance with blood glucose monitoring, and restricted access to medical guidance.ObjectiveTo evaluate the effectiveness of a village doctor-led intervention that combines financial incentives and telemedicine support to improve diabetes management among older adults with type 2 diabetes in rural China.MethodsA cluster randomized pilot study was conducted in all six village clinics of Baima Township, which were randomly assigned in a 1:1 ratio to either the usual care group or the integrated care group (combining financial incentives and telemedicine support). The primary outcome was the change in HbA1c levels from baseline to 3 months.ResultsA total of 215 patients completed the study and were included in the final analysis (108 in the usual care group and 107 in the integrated care group). Over the 3-month intervention, the integrated care group had a greater reduction in HbA1c than the usual care group (− 0.42 ± 1.04% vs. −0.23 ± 1.12%), although the difference was not statistically significant. A significant decreasing trend in fasting glucose was observed across the 13-week period (β = −0.077, 95% CI − 0.109 to − 0.046; P < 0.001), with greater reductions in the integrated care group (β = −0.047, 95% CI − 0.087 to − 0.007; P = 0.023). Postprandial glucose levels also showed a modest downward trend, although the between-group difference was not statistically significant. The mean monthly adherence to blood glucose monitoring was significantly higher in the integrated care group than in the usual care group (87.60% vs. 76.47%; mean difference 11.17%, 95% CI 5.10% to 17.23%; P < 0.001).ConclusionsThis pilot trial showed that a village doctor–led intervention combining small non-cash incentives with telemedicine support improved adherence to blood glucose monitoring among older adults with type 2 diabetes in rural China. Although no significant change in HbA1c was observed over 3 months, favorable trends in other outcomes suggest potential benefits. These findings suggest that such a model may represent a potentially feasible approach for improving diabetes management in rural primary care settings, which warrants confirmation in larger studies.Trial registrationThis study was registered at the Chinese Clinical Trial Registry (Registration number: ChiCTR2400088221, registered on 2024-08-13, https://www.medicalresearch.org.cn/login).Supplementary InformationThe online version contains supplementary material available at 10.1186/s12877-026-07703-4. Introduction Diabetes mellitus among older adults is a major global public health challenge. Worldwide, approximately 142 million individuals aged 65 years or older are living with diabetes, with China accounting for about 35.5 million cases, representing the highest number globally [1, 2]. In China, the prevalence of diabetes rises significantly with age, reaching 28.8% among individuals aged 60–69 years and 31.8% among those aged 70 years and older, compared to 12.8% across all age groups [3]. While urban areas benefit from better healthcare resources, leading to improved glycemic control, rural regions face substantial challenges, including limited healthcare infrastructure, low adherence to self-management practices, and inadequate access to professional diabetes care. Recognizing these disparities, the National Basic Public Health Service (BPHS) program was launched in 2009 to provide essential preventive and public health services to all residents [4, 5]. Under this program, primary health care institutions are responsible for delivering services such as health record management, health education, vaccination, and chronic disease management for conditions including hypertension and type 2 diabetes (T2D). Village doctors working in village clinics serve as frontline primary care providers in rural areas and play a central role in implementing these services. They are responsible for providing basic medical care, conducting regular follow-up for patients with chronic diseases, and supporting health promotion activities within rural communities. Historically evolved from the “barefoot doctor”, village doctors constitute the foundation of rural primary healthcare in China [6–8]. According to national regulations, village doctors are required to obtain either a village doctor certificate or a rural assistant general practitioner certificate before providing medical services. Despite their critical role in rural primary healthcare, the village doctor workforce faces several challenges, including an aging workforce and difficulties integrating evidence-based practices into routine care [9–12]. Given limited opportunities for professional development and a lack of standardized training, village doctors may face challenges in adopting updated clinical knowledge and delivering consistent, high-quality diabetes care [13]. To address these workforce limitations, innovative strategies are needed to support village doctors in delivering effective chronic disease management. Financial incentives have been increasingly recognized as an effective strategy to improve adherence to chronic disease management interventions. Previous studies indicate that financial incentives can effectively motivate patients to engage in health-promoting behaviors, including regular blood glucose monitoring, medication adherence, and lifestyle modifications [14, 15]. However, despite their potential, evidence on the effectiveness of financial incentives in rural primary care settings in China remains limited. Meanwhile, telemedicine provides an opportunity to expand access to higher-level medical expertise without increasing the burden on village doctors [16, 17]. We conducted a cluster randomized pilot study to evaluate the effectiveness of a village doctor-led intervention that combines financial incentives with telemedicine support. This study aimed to improve glycemic control and adherence to blood glucose monitoring among older adults with T2D in rural China. Methods Study design This open, cluster randomized controlled trial was conducted in six village clinics in Baima Township, Gaogang District, Taizhou City, Jiangsu Province, China. Each village clinic was staffed by one village doctor responsible for the management and follow-up of enrolled patients. The primary outcome was the change in HbA1c levels from baseline to 3 months. This study was registered at the Chinese Clinical Trial Registry (ChiCTR2400088221) on 13 August 2024. The study was conducted and reported in accordance with the CONSORT guidelines for pilot and feasibility trials. Ethical approval was obtained from the Ethics Committee of Taizhou People’s Hospital (Approval number: KY 2024-096-01). Study participants and recruitment The inclusion criteria for the study were: (1) Participants with T2D enrolled in Jiangsu Province’s Basic Public Health Service; (2) Aged 65 years or older; (3) Receiving residence and care at Baima Township Health Center; (4) Willing to participate. The exclusion criteria were: (1) Presence of severe comorbidities (e.g., coronary heart disease, stroke, psychiatric disorders, malignancies); (2) Long-term residence outside the area or for other reasons. Eligible participants were recruited by village doctors through home visits or telephone interviews. All participants enrolled in the evaluation provided written informed consent during the baseline assessment. Sample size Prior evidence suggests that telehealth may improve HbA1c, with observed differences of 0.20% to 1.43% compared to the control group [18–20]. We hypothesized that the integrated care group would achieve a 0.5% absolute reduction in HbA1c compared with the usual care group. Sample size was calculated using PASS (24.0.3) based on a cluster-randomized design and GEE methods. Assuming two groups, an average cluster size of 35, an intra-cluster correlation coefficient (ICC) of 0.01, a two-sided alpha of 0.05, and 80% power, the required sample size was 3 clusters per group (6 in total), with 210 participants overall. Randomization and masking The six village clinics were cluster-randomized in a 1:1 ratio to either the usual care group or the integrated care group. The randomization sequence was generated by a statistician using a computer-based randomization algorithm. Due to the cluster-based design and the nature of the intervention, blinding of study participants, village doctors, and research personnel responsible for clinical outcome data collection was not feasible. Procedures A protocol was implemented to improve adherence to blood glucose monitoring and glycemic control through the combination of low-cost financial incentives and telemedicine. The study design is illustrated in Fig. 1A. The Xingwang Plan—Regional Chronic Disease Digital Coordination Platform supported primary diabetes management in Gaogang District, Taizhou, by facilitating telemedicine services. Fig. 1Study design and participant flow. A Schematic comparison of usual care versus the integrated care model. In the usual care group, patients received standard care from village doctors. In the integrated care group, a village doctor-led model was implemented, incorporating real-time financial incentives and telemedicine support to improve diabetes management. B CONSORT flow diagram illustrating the cluster-randomized allocation of six village clinics, patient screening and enrollment, exclusions, and completion rates for the usual care and integrated care groups Study design and participant flow. A Schematic comparison of usual care versus the integrated care model. In the usual care group, patients received standard care from village doctors. In the integrated care group, a village doctor-led model was implemented, incorporating real-time financial incentives and telemedicine support to improve diabetes management. B CONSORT flow diagram illustrating the cluster-randomized allocation of six village clinics, patient screening and enrollment, exclusions, and completion rates for the usual care and integrated care groups All participants were informed that they could receive free weekly blood glucose testing at village clinics and that village doctors would provide lifestyle counseling for patients with suboptimal glycemic control. In the integrated care group, patients were additionally informed that small financial incentives would be offered if their blood glucose levels were well controlled. They were also informed that free teleconsultations with upper-level physicians would be arranged if blood glucose remained poorly controlled despite guidance from village doctors, with further treatment recommendations provided accordingly. Village doctors in the integrated care group received standardized training in blood glucose measurement, medication adherence counseling, lifestyle modification guidance, and the use of telemedicine services through the platform. Following an initial in-person training, interactive onsite training and certification were conducted by the investigators. During the 3-month study period, village doctors provided weekly blood glucose monitoring for each patient. Patients whose blood glucose levels were within the target range—defined according to the Chinese Guidelines for the Prevention and Treatment of Diabetes (2024 Edition) [21] as fasting glucose of 4.4–7.0 mmol/L and 2-hour postprandial glucose ≤ 10.0 mmol/L—were eligible to receive non-cash rewards (e.g., toothpaste, soap, or a towel). For patients with uncontrolled blood glucose, village doctors provided health coaching on lifestyle modification and medication adherence. If blood glucose remained poorly controlled, village doctors initiated video-based teleconsultations with endocrinologists to obtain further treatment guidance. Teleconsultation involved three-way video communication among patients, village doctors, and endocrinologists to discuss treatment strategies, medication adjustment, and lifestyle management for patients with poorly controlled glucose levels. During teleconsultation, endocrinologists were able to access patients’ clinical information through this digital platform, including medication use, blood glucose monitoring records, and annual health examination results. In contrast, village doctors in the usual care group received only standard training in blood glucose measurement, medication adherence, and lifestyle intervention, without access to the telemedicine system. They continued to provide routine diabetes management according to the National BPHS program, including blood glucose monitoring, lifestyle counseling, medication adjustment for patients with abnormal glucose levels, and regular follow-up. All village doctors received a fixed stipend of 5 RMB for each completed follow-up session that included blood glucose testing and behavioral counseling. Outcomes The primary outcome of this study was the mean change in HbA1c levels from baseline to month 3. The secondary outcomes included: (1) the average monthly adherence to blood glucose monitoring during the 3-month intervention period, and (2) the mean weekly fasting and postprandial blood glucose levels across the 13-week study period. Monthly adherence was defined as the proportion of weeks in each month in which the patient completed at least one blood glucose test relative to the total number of weeks in that month (i.e., four weeks in months one and two, and five weeks in month three). Data collection For HbA1c assessment, venous blood samples were collected at baseline and at month 3 by trained nursing staff at local village clinics. All samples were transported under standardized conditions to Taizhou Zhilian Medical Laboratory, where HbA1c levels were analyzed using high-performance liquid chromatography. Blood glucose data (fasting or postprandial) were collected by village doctors using finger-stick blood glucose meters. These devices were automatically connected to the platform via Bluetooth, and blood glucose data were uploaded to the corresponding patient record on the platform during each measurement. Blood glucose monitoring data were exported monthly by a data administrator, who was blinded to group allocation during data extraction and processing. Monthly adherence to blood glucose monitoring was defined as the proportion of weeks in each month in which the patient completed one fingerstick blood glucose measurement, divided by the total number of weeks in that month (i.e., four weeks in months one and two, and five weeks in month three). Completion of weekly blood glucose measurements was determined based on the records available on the platform. Statistical analysis The analysis followed the modified intention-to-treat principle. Data were presented as mean ± standard deviation for continuous variables and frequency (percentage) for categorical variables. Differences in baseline demographic and clinical characteristics between the two groups were assessed using independent sample t-tests for continuous variables and chi-square tests for categorical variables. HbA1c changes between groups were compared using independent t-tests. Adjusted analyses were performed using ANCOVA, controlling for age, gender, baseline HbA1c, and other relevant covariates. Weekly fasting and postprandial blood glucose levels were analyzed using weighted least squares regression. A generalized estimating equation linear model with exchangeable correlation structure was used to analyze differences in blood glucose monitoring adherence between groups. The model applied an identity link function and assumed independence with unequal variances for the covariance structure of the three observations. Data analysis was performed using the “geepack package” in R, adjusting for group, follow-up month, and covariates (age, duration of diabetes, gender, educational level, marital status, presence of hypertension and oral antidiabetic medication use, and insulin use). Statistical significance was set at two-tailed P < 0.05, and all analyses were performed using the R software (version 4.4.1). Results Study participants The study was conducted between August 2024 and November 2024. The study schedule according to the SPIRIT guidelines is presented in Table 1. Between August 8 and 15, 2024, a total of 373 eligible participants were screened and enrolled from all six village clinics in Baima Township, Gaogang District, Taizhou City, Jiangsu Province. These six clinics were randomized in a 1:1 ratio to either the integrated care group or the usual care group. Among them, 115 patients were enrolled in the usual care group (48, 22, and 45 from each clinic, respectively), and 111 patients were enrolled in the integrated care group (32, 33, and 46, respectively). During follow-up, 7 participants in the usual care group and 4 in the integrated care group withdrew. Ultimately, 108 participants in the usual care group and 107 in the integrated care group were included in the final analysis (Fig. 1B). Table 1 SPIRIT time schedule table Table 1 SPIRIT time schedule table SPIRIT time schedule table Baseline characteristics The baseline characteristics of the 215 participants are summarized in Table 2. The mean age was 72.6 ± 5.6 years, and the median duration of diabetes was 88 months (IQR: 17–144), with no significant differences between the usual care and integrated care groups (P > 0.05). Most participants were female (53.0%), married (68.4%), and enrolled in the New Rural Cooperative Medical Scheme (88.8%), with comparable distributions between groups. Hypertension was present in 58.1% of participants and was similarly distributed across groups. At baseline, 76.3% of participants used oral antidiabetic medications, with a significantly higher proportion in the usual care group than in the integrated care group (84.3% vs. 68.2%, P = 0.009). Therefore, oral antidiabetic medication use was included as a covariate in the adjusted analyses of the primary outcomes. Insulin use was low in both groups (11.1% vs. 5.6%), with no significant difference (P = 0.226). Table 2Baseline characteristics of study participants by groupAll(n = 215)Usual care(n = 108)Integrated care(n = 107)P valueAge (y)72.60 ± 5.6372.95 ± 5.7672.25 ± 5.490.362Duration of Diabetes (m)88 (17–144)77 (19–135)94 (16–154)0.529Gender0.635 Male101 (47.0%)49 (45.4%)52 (48.6%) Female114 (53.0%)59 (54.6%)55 (51.4%)Marital Status0.451 Single4 (1.9%)2 (1.9%)2 (1.9%) Married147 (68.4%)72 (66.7%)75 (70.1%) Widowed62 (28.8%)34 (31.5%)28 (26.2%) Divorced2 (0.9%)0 (0%)2 (1.9%)Education Level0.950 Illiterate or semi-illiterate88 (40.9%)45 (41.7%)43 (40.2%) Primary school75 (34.9%)35 (32.4%)40 (37.4%) Middle school33 (15.3%)18 (16.7%)15 (14.0%) High school15 (7.0%)8 (7.4%)7 (6.5%) Undergraduate4 (1.9%)2 (1.9%)2 (1.9%)Insurance0.321New Rural Cooperative Medical Scheme191 (88.8%)96 (88.9%)95 (88.8%)Basic Medical Insurance for Urban and Rural Residents19 (8.8%)8 (7.4%)11 (10.3%)Urban Employee Basic Medical Insurance5 (2.3%)4 (3.7%)1 (0.9%)Hypertension1.000With125 (58.1%)63 (58.3%)62 (57.9%)Without90 (41.9%)45 (41.7%)45 (42.1%)Oral antidiabetic medication use164 (76.3%)91 (84.3%)73 (68.2%)0.009Insulin use18 (8.4%)12 (11.1%)6 (5.6%)0.226Data are mean ± SD, median (IQR), or n (%) Table 2Baseline characteristics of study participants by groupAll(n = 215)Usual care(n = 108)Integrated care(n = 107)P valueAge (y)72.60 ± 5.6372.95 ± 5.7672.25 ± 5.490.362Duration of Diabetes (m)88 (17–144)77 (19–135)94 (16–154)0.529Gender0.635 Male101 (47.0%)49 (45.4%)52 (48.6%) Female114 (53.0%)59 (54.6%)55 (51.4%)Marital Status0.451 Single4 (1.9%)2 (1.9%)2 (1.9%) Married147 (68.4%)72 (66.7%)75 (70.1%) Widowed62 (28.8%)34 (31.5%)28 (26.2%) Divorced2 (0.9%)0 (0%)2 (1.9%)Education Level0.950 Illiterate or semi-illiterate88 (40.9%)45 (41.7%)43 (40.2%) Primary school75 (34.9%)35 (32.4%)40 (37.4%) Middle school33 (15.3%)18 (16.7%)15 (14.0%) High school15 (7.0%)8 (7.4%)7 (6.5%) Undergraduate4 (1.9%)2 (1.9%)2 (1.9%)Insurance0.321New Rural Cooperative Medical Scheme191 (88.8%)96 (88.9%)95 (88.8%)Basic Medical Insurance for Urban and Rural Residents19 (8.8%)8 (7.4%)11 (10.3%)Urban Employee Basic Medical Insurance5 (2.3%)4 (3.7%)1 (0.9%)Hypertension1.000With125 (58.1%)63 (58.3%)62 (57.9%)Without90 (41.9%)45 (41.7%)45 (42.1%)Oral antidiabetic medication use164 (76.3%)91 (84.3%)73 (68.2%)0.009Insulin use18 (8.4%)12 (11.1%)6 (5.6%)0.226Data are mean ± SD, median (IQR), or n (%) Baseline characteristics of study participants by group Data are mean ± SD, median (IQR), or n (%) Change in HbA1c levels over the 3-month follow-up At the 3-month follow-up, there was no significant difference in mean HbA1c levels between the usual care group (7.57 ± 1.13%) and the integrated care group (7.38 ± 1.25%, P = 0.264; Table 3). From baseline to follow-up, the integrated care group showed a greater reduction in HbA1c (− 0.42 ± 1.04%) compared with the usual care group (− 0.23 ± 1.12%); however, the difference between groups was not statistically significant (P = 0.199). Table 3Comparison of blood glucose monitoring adherence and HbA1c outcomes at 3-month follow-up between integrated care and usual care groupsUsual care(n = 108)Integrated care(n = 107)Difference(95% CI)P valueAdjusted Difference* (95% CI)P valueAdherence76.47 ± 18.35%87.60 ± 6.93%11.17%(5.10%, 17.23%)< 0.00112.29%(6.42%, 18.17%)< 0.001HbA1c (%)3-month follow-up7.57 ± 1.137.38 ± 1.25−0.19 (− 0.51, 0.13)0.264−0.17 (− 0.42, 0.09)0.133Change from baseline−0.23 ± 1.12−0.42 ± 1.04−0.19 (− 0.48, 0.10)0.199−0.16 (− 0.47, 0.14)0.135Data are presented as mean ± SD*The analysis was adjusted for age, duration of diabetes, gender, educational level, marital status, presence of hypertension, oral antidiabetic medication use, and insulin use Table 3Comparison of blood glucose monitoring adherence and HbA1c outcomes at 3-month follow-up between integrated care and usual care groupsUsual care(n = 108)Integrated care(n = 107)Difference(95% CI)P valueAdjusted Difference* (95% CI)P valueAdherence76.47 ± 18.35%87.60 ± 6.93%11.17%(5.10%, 17.23%)< 0.00112.29%(6.42%, 18.17%)< 0.001HbA1c (%)3-month follow-up7.57 ± 1.137.38 ± 1.25−0.19 (− 0.51, 0.13)0.264−0.17 (− 0.42, 0.09)0.133Change from baseline−0.23 ± 1.12−0.42 ± 1.04−0.19 (− 0.48, 0.10)0.199−0.16 (− 0.47, 0.14)0.135Data are presented as mean ± SD*The analysis was adjusted for age, duration of diabetes, gender, educational level, marital status, presence of hypertension, oral antidiabetic medication use, and insulin use Comparison of blood glucose monitoring adherence and HbA1c outcomes at 3-month follow-up between integrated care and usual care groups 11.17% (5.10%, 17.23%) 12.29% (6.42%, 18.17%) Data are presented as mean ± SD *The analysis was adjusted for age, duration of diabetes, gender, educational level, marital status, presence of hypertension, oral antidiabetic medication use, and insulin use Changes in weekly blood glucose levels Figure 2A–B presents the weekly trends of fasting and postprandial blood glucose over the 13-week intervention period. Weighted regression analysis indicated a significant decreasing trend in fasting glucose over time (β = −0.08, 95% CI − 0.11 to − 0.05, P < 0.001; Table 4), with a greater decline observed in the integrated care group compared with the usual care group (group × week interaction: β = −0.05, 95% CI − 0.09 to − 0.01, P = 0.023). The integrated care group generally exhibited lower mean postprandial glucose levels than the usual care group; however, the between-group difference was not statistically significant (β = −1.04, 95% CI − 2.35 to 0.28, P = 0.116). Postprandial glucose showed a modest downward trend over time (β = −0.11, 95% CI − 0.23 to 0.01, P = 0.065), but the group × week interaction was not significant (P = 0.300). Fig. 2Weekly blood glucose levels and monthly adherence in the usual care and integrated care groups. A Weekly trends of fasting plasma glucose (FPG) with 95% confidence intervals (CIs) in the usual care and integrated care groups. B Weekly trends of postprandial plasma glucose (PPG) with 95% CIs in the two groups. C Monthly adherence rates in the usual care and integrated care groups. D Differences in adherence rates between the two groups by month, with 95% CIs. Weekly blood glucose levels and monthly adherence in the usual care and integrated care groups. A Weekly trends of fasting plasma glucose (FPG) with 95% confidence intervals (CIs) in the usual care and integrated care groups. B Weekly trends of postprandial plasma glucose (PPG) with 95% CIs in the two groups. C Monthly adherence rates in the usual care and integrated care groups. D Differences in adherence rates between the two groups by month, with 95% CIs. Table 4Weighted regression analysis of weekly fasting and postprandial blood glucoseVariableβ (95% CI)P-valueFPGGroup (Intervention vs. Control)−0.14 (− 0.44, 0.15)0.330Week−0.08 (− 0.11, − 0.05)< 0.001Group × Week−0.05 (− 0.09, − 0.01)0.023PPGGroup (Intervention vs. Control)−1.04 (− 2.35, 0.28)0.116Week−0.11(− 0.23, 0.01)0.065Group × Week−0.09 (− 0.25, 0.08)0.300FPG Fasting plasma glucose, PPG Postprandial plasma glucose Table 4Weighted regression analysis of weekly fasting and postprandial blood glucoseVariableβ (95% CI)P-valueFPGGroup (Intervention vs. Control)−0.14 (− 0.44, 0.15)0.330Week−0.08 (− 0.11, − 0.05)< 0.001Group × Week−0.05 (− 0.09, − 0.01)0.023PPGGroup (Intervention vs. Control)−1.04 (− 2.35, 0.28)0.116Week−0.11(− 0.23, 0.01)0.065Group × Week−0.09 (− 0.25, 0.08)0.300FPG Fasting plasma glucose, PPG Postprandial plasma glucose Weighted regression analysis of weekly fasting and postprandial blood glucose FPG Fasting plasma glucose, PPG Postprandial plasma glucose Blood glucose monitoring adherence Over the 3-month period, the mean monthly adherence was 76.47% in the usual care group and 87.60% in the integrated care group (P < 0.001; Table 3), corresponding to a group difference of 11.17% (95% CI: 5.10% to 17.23%; P < 0.001). In the adjusted analyses, this difference remained significant at 12.29% (95% CI: 6.42% to 18.17%, P < 0.001). Figure 2C–D shows the monthly adherence to blood glucose monitoring. In the second month, mean adherence was 79.20% in the usual care group and 91.60% in the integrated care group, with an adjusted group difference of 13.05% (95% CI: 4.90% to 21.21%, P < 0.001). In the third month, adherence declined to 56.90% and 79.60%, respectively (P < 0.001), with an adjusted group difference of 24.75% (95% CI: 13.53% to 35.98%, P < 0.001). Teleconsultation frequency and HbA1c change The mean weekly teleconsultation rate in the intervention group was 7.92%. To further explore the relationship between consultation frequency and glycemic control, patients were categorized into three groups according to the number of teleconsultations received during the 3-month follow-up: 0 sessions (n = 34), 1 session (n = 51), and 2 sessions (n = 22). The mean changes in HbA1c were − 0.22% for patients without teleconsultations, − 0.41% for those with one session, and − 0.73% for those with two sessions. Although the between-group difference was not statistically significant (P = 0.210), a dose–response trend was observed, with greater HbA1c reductions among patients who received more teleconsultations. Discussion In this pilot cluster-randomized trial, we found that a village doctor–led integrated care model, combining low-value financial incentives with telemedicine support, significantly improved adherence to weekly blood glucose monitoring among older adults with T2D in rural China. Consistent with this behavioral improvement, fasting glucose showed a significant downward trend over the 13-week intervention, with greater reductions observed in the integrated care group compared with usual care. However, despite these favorable short-term glycemic patterns, the intervention did not result in a statistically significant reduction in HbA1c at 3 months. As this study was designed as a pilot trial, the relatively small sample size and short follow-up period may have limited the ability to detect statistically significant differences in HbA1c. The observed improvement in patient behavior may be attributed to two core components of the intervention: an immediate incentive mechanism and an integrated telemedicine support system. Unlike traditional incentive models that reward task completion or offer delayed reinforcement, our intervention provided small, real-time non-cash rewards (e.g., toothpaste or soap) immediately after each episode of blood glucose monitoring that met predefined criteria. These low-cost items were both practical and culturally acceptable in rural settings, offering sufficient motivation while remaining cost-effective. Prior evidence supports the effectiveness of such low-cost incentives [22–25]. Behavioral economics theory further emphasizes that immediate and certain small rewards are often more effective in sustaining behavior change than delayed or high-value rewards [26]. From a behavioral economics perspective, individuals tend to prioritize immediate rewards over long-term health benefits, a phenomenon known as present bias [27]. Diabetes self-management behaviors, such as regular blood glucose monitoring and lifestyle modification, require sustained effort, whereas the health benefits often become apparent only over the long term. Financial incentives that provide immediate rewards may help offset this temporal imbalance and thereby increase patients’ willingness to engage in recommended health behaviors and improve adherence [28]. In addition, according to behavioral reinforcement theory, behaviors followed by rewards are more likely to be repeated [29]. Therefore, financial incentives can serve as positive reinforcement for self-management behaviors such as regular blood glucose monitoring, facilitating the formation of sustained health behavior patterns. Furthermore, financial incentives may enhance patients’ extrinsic motivation [30], thereby increasing their engagement in diabetes self-management activities. Furthermore, teleconsultation may influence patient adherence and self-management through several potential mechanisms. First, it provides village doctors with specialist support from upper-level physicians, which may improve clinical decision-making and promote more standardized management of chronic diseases [31]. For patients with persistently poor glycemic control, consultation with specialists allows more individualized treatment recommendations and timely therapeutic adjustments. Second, the involvement of higher-level physicians may enhance the credibility of treatment decisions and increase patient trust in the care provided at the primary care level. Previous studies have shown that patient trust in primary care physicians is associated with greater acceptance of medical advice and improved healthcare behaviors [32]. Finally, teleconsultation may increase patient engagement by strengthening interactions between patients and healthcare providers and improving patients’ understanding of disease management, thereby enhancing their capacity for self-management [33]. Although financial incentives have shown potential in promoting self-management behaviors among patients with chronic diseases, their impact on clinical outcomes such as glycemic control remains inconclusive. Previous studies have reported mixed findings, with some observing modest reductions in HbA1c [14, 15, 34–39], while others found no significant effects—particularly when the intervention duration was short or the strategy lacked multiple components [40–42]. In recent years, telemedicine has been increasingly recognized as an effective approach to improving glycemic outcomes [43]. A systematic meta-review by Eberle et al. [44], which included 99 telemetric interventions, reported that real-time video-based interventions achieved an average HbA1c reduction of 1.15% over a 6-month period, making it the most effective modality among those reviewed. In our study, no statistically significant difference in HbA1c levels was observed between the intervention and usual care groups. However, we noted a significant downward trend in fasting glucose across the 13-week follow-up, with greater reductions observed in the intervention group. These findings suggest that the intervention had already produced favorable short-term effects on glycemic patterns. Given that HbA1c reflects average glycemia over the preceding 2–3 months, and that the between-group difference in blood glucose monitoring adherence became evident mainly after the second month, it is plausible that the full effect of the intervention had not yet been captured within the limited follow-up period. Regular self-monitoring of blood glucose (SMBG) has been shown to effectively reduce HbA1c levels and improve glycemic control in patients with type 2 diabetes [45, 46]. However, in China, adherence to SMBG remains suboptimal, with studies reporting that fewer than half of patients perform monitoring on a regular basis [11, 47]. According to the Chinese Guidelines for the Prevention and Treatment of Diabetes (2024 Edition), patients with T2D treated with oral hypoglycemic agents are recommended to monitor their blood glucose 2–4 times per week, while those receiving insulin therapy are advised to perform self-monitoring more frequently. Meanwhile, the national Basic Public Health Service (BPHS) program currently provides only quarterly blood glucose testing, which is insufficient to meet the needs of long-term glycemic management. Against this background, our study was conducted in village clinics in Baima Township, where prior collaboration ensured strong cooperation from village doctors and the clinics were representative of typical rural primary healthcare settings. Patients were encouraged to monitor their blood glucose weekly at village clinics. Adherence to blood glucose monitoring was nearly 100% at the beginning of the study in both groups following intensive supervision and education at enrollment, but gradually declined over time during follow-up. This decline may reflect the attenuation of the initial supervision effect, as well as the sustained effort required for regular self-monitoring of blood glucose, which may reduce patients’ motivation over time. Nevertheless, the integrated care group, which combined low-value financial incentives with teleconsultations for patients with poor glycemic control, demonstrated a slower decline in adherence compared with the usual care group, suggesting that this strategy may help sustain long-term engagement in SMBG and improve glycemic control. This study has several limitations. First, as a pilot study, the 3-month intervention period may be insufficient to capture longer-term effects on clinical outcomes such as HbA1c or sustained adherence. Second, the single-center design limits the generalizability of our findings to other rural settings. Third, because village doctors were instructed to provide lifestyle counseling for patients with poor glycemic control, individuals with severe comorbidities (e.g., coronary heart disease, stroke, or psychiatric disorders) were excluded from enrollment to ensure participant safety. This may limit the generalizability of the findings to the broader rural diabetic population with more complex health conditions. In addition, adherence to blood glucose monitoring was nearly 100% at the beginning of the study due to intensive supervision and education at enrollment, but gradually declined over time. This pattern resulted in limited variability in the early phase and potential reverse causality in the later phase (i.e., poor adherence leading to more teleconsultations); therefore, we did not formally analyze the association between adherence trajectories and the frequency of teleconsultations. Furthermore, although this study adopted a cluster-randomized design, the primary analyses were conducted at the individual level due to the small number of clusters (six village clinics), which limited the feasibility and statistical stability of cluster-level analyses and should be interpreted with caution. In addition, the detailed content of teleconsultations was not systematically recorded, which limited our ability to further examine how specific consultation components may have influenced treatment decisions or clinical outcomes. Building on the insights from this pilot study, the present findings can also provide preliminary parameters for sample size estimation in future larger multi-site randomized trials with longer follow-up to fully evaluate the clinical effectiveness of this intervention model. From a health system perspective, the intervention used in this study involved relatively low-cost components, including small non-cash incentives (e.g., daily necessities) and a modest stipend for village doctors. Such a model may represent a potentially cost-efficient strategy for improving diabetes self-management in rural primary care settings. However, the present pilot trial did not formally evaluate economic outcomes. Future large-scale studies should incorporate comprehensive health economic evaluations, including incremental cost-effectiveness ratios (ICERs), to assess the long-term economic value of combining financial incentives with telemedicine support. Conclusion In summary, this pilot trial demonstrated that a village doctor-led intervention integrating small non-cash incentives and telemedicine support improved adherence to blood glucose monitoring among older adults with type 2 diabetes in rural China. While no significant change in HbA1c was observed during the short follow-up, favorable trends in glycemic outcomes suggest the potential of this model. Larger multi-site studies with extended follow-up are needed to confirm its clinical effectiveness and sustainability. Supplementary Information Supplementary Material 1. Supplementary Material 1.