Consistent with previous research findings, the turnover intention (p) among village doctors in China is approximately 50%, spanning a range from 35.8 to 84.44% (20–24). Employing a significance level of α = 0.05 for a two-tailed test, and considering a permissible margin of error (d) of 0.1 times the proportion (p), the required sample size is determined to be 400 based on the applicable statistical sampling formula: n = Z n / 2 2 × p q d 2 .Considering a 10% non-response rate, the final sample size should be 440.

In 2023, according to the method of un-random sampling, a designed cross-sectional study was carried out and an electronic self-filled questionnaire survey was conducted among a sample of NGVDs who have worked at the community level in Shanghai. Inclusion criteria for this cross-sectional survey are as follows: Having a minimum of 6 months of clinical experience immediately preceding the survey and be consenting to engage in the survey process. The participants were asked questions related to their experiences, opinions, and perceptions. The study data were anonymous to protect privacy.

Based on previous studies, this study aimed to explore: (1) NGVDs reselection intentions; (2) reasons for attending RTME; (3) job satisfaction; (4) career identity; (5) the effects of job satisfaction and career on reselection intention; (6) other determinants. The questionnaire was generally divided into five parts.

The NGVDs’ intention of reselecting RTME was measured by asking “If you had the chance to reselection, would you choose to be a Village Doctors of Directional Training?,” with responses of “yes,” “no” and “not clear.” The answers “yes” were coded to be yes (coded 1); while “no” and “not clear” were not (coded 0). The information collected will be kept confidential.

Data analysis was performed using SAS version 9.4 and Effective results were plotted using GraphPad Prism 9. Descriptive statistics were used to describe the characteristics of the NGVDs. continuing variable such as age was presented as mean and standard deviation (SD), while the statistical description of categorical variables used frequency and percentage. Differences in “If you had the chance to choose again, would you choose to be a Village Doctors of Directional Training under the order-directed program?” were examined using ANOVA for continuing variables and x²-test for categorical variables. Finally, logistic regression analysis was conducted in order to determine factors associated with the intention of NGVDs reselect RTME. A p-value =0.05 was considered significant.

A total of 419 participants were successfully recruited, although 21 participants did not complete the questionnaire in accordance with the specified requirements. The NGVDs’ average age were 27.31 ± 2.59 years, with 25.06% male and 74.94% female respondents. In terms of the respondents’ work unit levels, 21.96% were from township health centers, while the majority, 76.61% of the NGVDs, were from community health service centers. 98.09% were Still practicing as a doctor, with the rest worked in other types of jobs. Among the respondents, 49.40% of which had less than 3 years of work experience, only 1.7% of them held medium-grade and above professional titles (see Table 1).

When asked about the reasons for applying as NGVD of RTME, the top three were “No job pressure as work unit determined upon enrollment,” “Volunteering to serve the grassroots people” and “No financial pressure during schooling period” (see Table 2).

During the initial application phase, 16.23% of them had a thorough understanding of the government’s policy on RTME, while 47.26% had some understanding. 29.83% had a general understanding, and 6.68% had no understanding of the policy. In the current stage, these proportions have changed to 18.62, 50.12, 26.73, and 4.54%, respectively.

Based on the data in Figure 1, there has been a significant increase in the level of understanding of the government’s policy on RTME compared to the application phase. Initially, 63.49% of individuals had a high level of understanding of the policy. In the current stage, this proportion has increased to 68.74%, representing a 5.25% increase compared to the application phase. Conversely, during the application phase, 36.51% of individuals did not have a deep understanding of the policy. In the current stage, only 31.24% of individuals lack a deep understanding, indicating a 5.17% decrease compared to the application phase.

In terms of how they view the signing of the RTME, 70.64% believe that their initial choice was correct. On the other hand, 27.21% express doubt about their initial choice, and 2.15% admit that their initial choice was completely wrong (see Figure 2).

In the case of reselection, 270 individuals (64.44%) expressed their willingness to select the RTME again, while 149 individuals (35.56%) indicated that they would not (no clear) select it again.

From Table 3, it shows that the reasons for reselecting RTME such as “Volunteering for grassroots service” and “enjoying the medical profession” and “meeting family expectations” were related to the intention to reselection (p < 0.05). In the face of reselection, different choices will be made based on these reasons. The level of understanding of the national policy during two stages, importance of VDs and job satisfaction were all related to the intention to reselection (p < 0.05).

Variables that achieve statistical significance in the t-test or chi-square test are incorporated into the logistic regression analysis. Table 4 shows the results of the logistic regression for factors associated with reselection. NGVDs who chose the reason for “Volunteering to serve the grassroots people” (p < 0.05, OR = 2.29, 95%CI: 1.22–4.30), “Enjoying the medical profession” (p < 0.05, OR = 2.64, 95%CI: 1.33–5.27) and not fulfilling family wishes (p < 0.05, OR = 0.47, 95%CI: 0.24–0.88) were more likely to reselect this career. NGVDs who think village doctors can leverage their professional ability with “okay” (p < 0.05, OR = 32.87, 95%CI: 3.69–293.64) and “yes” (p < 0.05, OR = 120.03, 95%CI: 12.78- > 999.99) and who think NGVDs were “important” (p < 0.05, OR = 3.74, 95%CI: 1.21–11.55) were more inclinatively to reselect to be NGVD. Satisfied with the current salary also increased their probability of reselecting this profession (p < 0.05, OR = 2.39, 95%CI: 1.27–4.49).

Numerous similar studies have made research on turnover intention which refers the likelihood of an employee to leave the current job he/she are doing (25). In our study, 35.56% of the NGVDs will not reselect RTME, while the rest of them will still reselect this career which is lower than the rate of other provinces in China. 66.5% of rural-oriented tuition-waived medical students (RTMSs) in Shaanxi had no intention to remain after the contract expired (20) while 57.7% RTMSs in Shandong province, 84.44% in Chongqing and 45.5% in Chengdu, had low and medium willingness to fulfill the contract (21). A investigation results showed that 36.8% of the village doctors in Xiangyang City intended to resign (22) and the number was 42.3% in Chongqing (23). A survey about village doctor in Shandong province showed that 46.9% of them had a higher turnover intention (24). According to systematic research, which integrated 20 research among 23,284 VDs in China, the prevalence of turnover intention among VDs in China was as high as 44.1% (3). A study of comparison based on three surveys in a province in eastern China found that the turnover intention of village clinic doctors has increased by time (26).

But the rate is higher than Inner Mongolia autonomous region which approximately 33% of rural order-oriented medical students were not willing to work in rural areas (27). The other counties, Grassroots doctors in British, for example, only 11.8% of primary care family physicians had high turnover intention (28). Of a 1,174 family physicians studied in England, only 6.30% of them stated high intention to leave (29).

This study showed clearly the top three reasons for applying as NGVD of RTME were guaranteed employment, volunteering to serve the grassroots people and no financial pressure, which is in accordance with findings in previous studies (16). Understanding the motivation why they select RTME program is very important because a motivated individual is willing to exert and maintain an effort to provide good quality health services (30). Simultaneously, we reported that students who enrolled in the RTME program due to “Volunteering to serve the grassroots people” and “Enjoying the medical profession” were more likely to reselect this career. But, “RTME for Fulfilling family wishes” were less likely to reselect the RTME program. Therefore, medical schools and employers should focus on researching how to correctly improve the professional identity of village doctors, so that the RTME project can sustainably develop and achieve maximum resource utilization.

The results of our study showed that career identity related to reselection conclude “NGVDs can leverage their professional ability” and “Importance of NGVDs assigned to primary health care development.” Study by Zhang et al. strongly confirmed that career identity are early and powerful predictors of the willingness to fulfill the contract of RTMSs (21). Village doctors had poor working and living conditions, such as low income, lack of social security, inappropriate performance assessment, inadequate training, heavy workload and insufficient cooperation from rural residents (31), which affected the career identity of them (32). Simultaneously, limited professional development prospect, unsound medical equipment and facilities, and inadequate financial remuneration, influencing the professional identity of RTMSs (33).

Job satisfaction is an important predictor and it have a direct and indirect negative effect on turnover intention of health workers (24, 34). A scientometrics perspective study by Chen highlighted that job satisfaction had a negative predictive effect on VDs’ turnover intention (35). Our study indicated that NGVDs were dissatisfied with their jobs, including salary and performance of professional competence and continuing education policies, especially among those who will not reselect NGVD of RTME. A cross-sectional study showed 48.6% of VDs felt satisfied with their job (36). Renmin Jin’s study found that job satisfaction rate of village doctors with RTME program was 56.50% (16).

Studies showed that income satisfaction contribute to the turnover intention of village doctors (22, 37). Income can be a major contributor to dissatisfaction and low income was directly linked to the retention rates of rural physicians (38). The mismatch between income and workloads is a major contributor toward job dissatisfaction (39). This result showed that among the 419 NGVDs surveyed, more than half of them, 301 (71.84%) dissatisfied with the current salary. Huang’s survey of village doctors’ mobility found that 70% of village doctors were unsatisfied with their income (40). A series of changes, such as cancelling drug mark-ups and integrating management of village doctors’ activities, have dramatically reduced village doctors’ income and employment satisfaction (41).

The research findings suggest that a considerable number of NGVDs are inclined to opt for the RTME program again, indicative of its favorable reception. Nonetheless, delving into the motivations behind the reluctance of others is imperative for refining enrollment and retention approaches. Although the RTME program has played a pivotal role in mitigating the scarcity of healthcare personnel in rural regions, the study underscores key domains that necessitate policy interventions to bolster the program’s impact and long-term viability.

The study suggests that limited professional development prospects and financial remuneration affect the professional identity of NGVDs. Health policies should focus on fostering a strong professional identity among medical students by emphasizing the importance of their role in rural healthcare and providing opportunities for community engagement. Moreover, policies must be crafted to present enduring career pathways, which include prospects for specialization and continuing education within the rural healthcare infrastructure. Given the extensive training requirements of village doctors (42), it is imperative to develop a standardized Continuing Medical Education (CME) framework. This framework should guarantee NGVDs access to ongoing training in both clinical and public health domains, thereby enhancing their skills and knowledge, and ultimately elevating the quality of care they deliver.

Consistent with prior empirical research that corroborated the direct association between job satisfaction and the propensity for turnover (43), the current investigation reveals that job satisfaction, specifically related to salary and career advancement opportunities, is a crucial determinant in the reselection decision-making process of NGVDs. Health policies should aim to improve the compensation and benefits for NGVDs. Tackling the challenges of substandard income and insufficient social protection is imperative for the recruitment and retention of skilled NGVDs. The adoption of policies guaranteeing equitable compensation and benefits, including performance-related pay and extensive insurance schemes, is likely to bolster job satisfaction and enhance staff retention rates.

This investigation underscores the critical role that workplace conditions and support systems play in affecting the reselection of NGVDs. Health policies must prioritize the enhancement of the work environment, the provision of sufficient resources, and the establishment of social security measures for these healthcare professionals. Additionally, health policies could foster community engagement in healthcare provision and establish support networks for NGVDs, thereby reinforcing their sense of belonging and societal contribution. Establishing peer support networks, which appear to substantially impact job satisfaction (44), can offer NGVDs a communal and emotional lifeline. Ensuring access to mental health support services, including counseling and stress management initiatives, is vital for helping them navigate the emotional and psychological demands of rural healthcare work.

This study has several limitations. First, data was collected using convenience sampling, not representative of the experience of all NGVDs in Shanghai. Meanwhile our study provides valuable insights into the reselection intentions of NGVDs in Shanghai, it is important to acknowledge the potential limitations in the generalizability of our findings. The sample was drawn from a specific urban setting in Shanghai, which may not fully represent the experiences and perspectives of NGVDs in rural or less developed areas. Second, the cross-sectional study design made it difficult to identify causal relationships. The cross-sectional nature of our study precludes the examination of changes over time which means that we cannot track the development of phenomena or assess the stability of the observed associations. Further study with a longitudinal design is needed. Third, data was collected by self-report of the respondents, which could cause recall bias, and structured scales were not used to measure turnover intention, job satisfaction, and career identity. These limitations should be considered in future studies.

While acknowledging the limitations inherent in our study’s design and sample, several aspects of our findings suggest potential broader applicability beyond the specific context of Shanghai. The identified factors influencing reselection intentions, namely career identity, job satisfaction, and understanding of government policies, are likely to be relevant in other settings where rural-oriented tuition-waived medical education programs are implemented. Furthermore, the challenges faced by VDs, including low income and limited professional development opportunities, are widely recognized issues impacting healthcare workers in rural areas across China.