Workload and job satisfaction of rural doctors during the new healthcare reform in China: a cross-sectional mixed methods study

With the integration of urban and rural health insurance, the demand for health services from rural residents increases rapidly, which in turn, bring heavier workload for doctors from county public hospitals (CPHs). Meanwhile, township healthcare centers (THCs) are required to provide more additional public health services under the integration of public health and primary care, which also brings challenges for its doctors’ workload. As a result, Chinese rural doctors from both CPHs and THCs have to cope with heavier workload that may have an adverse effect on their job satisfaction. This study sought to investigate the association between the workload and their job satisfaction during the new healthcare reform in China. The workload, dominated by non-medical activities, working hour and clinical visit number, are non-negligible factors that negatively associated with Chinese rural doctors’ job satisfaction in the healthcare reform setting. The growing number of patients towards CPHs and the additional public health service imposed on the THC doctors were the main source for the workload that worsened rural doctors’ job satisfaction. Policy makers should consider some feasible measures to reduce non-medical activities for rural doctors so that they could dedicated their limited energy and time to the medical service.


Introduction
In recent decades, a large number of countries have been exploring healthcare reform to seek for a reasonable trade-off between fairness and e ciency (1,2) . China is not an exception. In the new round of medical reforms which began in 2009, the Chinese government puts forward the goal as "maintaining the basics, strengthening the grassroots, and building mechanisms". With the purpose of equalizing public services domestically, the "new rural cooperative healthcare system", one of the most important elements of new healthcare reform, has been proved to effectively activated the demand for medical services of rural residents in China. Recently, with the aim of universal health coverage, new healthcare reform requires to increase the rural residents' visit rate in their own county and a large proportion of the patients from rural areas urge to county public hospitals (CPHs) which is on the top of the three-tier rural health system. As a result, the workload for doctors from CPHs is heavier than ever accordingly.
Meanwhile, in the orientation of returning to public welfare for primary healthcare organization in Chinese healthcare reform setting, township healthcare centers (THCs) were required to provide more additional public health services. Because of the shortage of public health workers in THCs, quite a few doctors in the institution have to undertake the additional public health services (3) . Doctors often put more values on medical services rather than public health services (3) . Only when employees consider their work meaningful can they recover quickly from work fatigues (4) . Therefore, under this setting, non-medical activities contribute to THC doctors' workload as well.
Previous studies on job satisfaction have found that one of the factors affecting job satisfaction is whether self-worth is realized (5,6) . When health workers feel the recognition of their work and see the improvement of patients' health, they will have a feeling of self-realization, which will improve their job satisfaction. Heavy workload can lead to high work pressure for doctors, increase the risk of health problem, and therefore negatively in uence their job satisfaction (7)(8)(9)(10) . Long-term burdens even increase the chance of medical accidents (11)(12)(13)(14) . Chinese new healthcare reform brings overburden to doctors including both doctors from CPHs and THCs (hereafter, rural doctor). However, less research is targeted on rural doctors' workload and compare what constitute in their workload in CPH doctors and THC doctors differently.
As a stakeholder and essential part of health systems, doctors are central to the achievement of new healthcare reform goal (15, 16) . Health services are highly heterogeneous and personalized (17) , which need doctors put both their professional knowledge and service attitude. Therefore, they are engaged in work with both mental and emotional labors. Their incentive plays an important role in the delivery of medical services (18) and their job satisfaction has a positive relationship with the quality of health services and e ciency (19)(20)(21)(22) and has a negative relationship with their turnover intention (23)(24)(25) .In the context of continuous development and changes of health industry, healthcare professionals face many stresses, and their job satisfaction has attracted increasing attention (26)(27)(28) .Specifying what constitutes the main part of the workload of rural doctors and its distribution in CPHs and THCs and discussing the relationship between workload and job satisfaction would be bene cial to attract and retain rural doctors and improve the sustainability of health services in the background of new healthcare reform in China. At the same time, it also provides empirical evidence for the reform of medical and health system in other countries.
Therefore, this study aimed to portray the workload of rural doctors during the new healthcare reform in China and evaluate the relationship between workload and job satisfaction of doctors, focusing on the differences between CPHs and TCHs in rural areas, by utilizing the data from a cross-sectional survey, including focus group interviews with doctors and hospital managers.

Design and procedure
Both quantitative and qualitative analyses were adopted to evaluate the workload of rural doctors and their association with job satisfaction. The data of a cross-sectional study was collected from doctors from county public hospitals (CPHs) and township healthcare centers (THCs) in Gansu, Shanxi, and Shandong provinces during June-July 2019. County public hospitals include both county people hospital and Chinese traditional medicine hospital. The three provinces were selected, because they are located in the west, middle, and east of China, respectively. Two pilot tests within a small group of Tashang Township Healthcare Center which lies in Lingshou County, Hebei Province, had been conducted to assess the validity of the questionnaire. All respondents were voluntary and anonymous. Thirty-minute focus group interviews with hospital managers and some of the doctors were conducted to acquire additional information of the workload of doctors and other aspects related to their job satisfaction.

Measurements
Demographic and economic variables, such as gender, age, marital status, education level, professional title, and years of work experiences were collected in this study. Moreover, the relationship between workload and job satisfaction was evaluated.
Workload Workload is often measured by work intensity and working hours (12) .As stated above, nonmedical activities also contributes to doctors' workload. This study examined the three different but Interrelated indicators. (i) Clinical visit number (CV) that re ects the number of patients a doctor sees one day on average is directly related to the intensity of doctors' work. Given the potential differences between different departments in the amount of medical services and the number of patients seen, self-assessment of the number of patients was used as a measurement of workload. CV was assessed by their answers of "moderate", or "too many" to a question as "What do you think is the average number of patients you see each day?". (ii) Working hours one day on average (WH). As a more objective measurement of workload, the WH of respondents were evaluated by their answers of working "less than 8 hours", "8-10 hours", "10-12 hours", or "more than 12 hours" per day. (iii) The proportion of time spent on non-medical activities one day on average (PT). A question as "what do you think of the proportion of non-medical activities in a routine working day" with answers of "moderate", or "too much" was used to assess the PT of the respondents. Non-medical activities mainly point to activities that is irrelevant to doctors' clinical job and include sorting administrative matters, attending meetings, and dealing with the inspections of supervising authorities. In addition, for respondents in THCs, non-medical activities also include public health services providing.
Job satisfaction Job satisfaction is an important indicator of employee's occupational well-beings (29) .
Respondents were asked a question as "On the whole, are you satis ed with your job?" to measure their job satisfaction. Likert 5-point method was adopted to assess the corresponding answers as "very dissatis ed" (1 point) to "very satis ed" (5 points), of which "Very dissatis ed/ dissatis ed", "reasonable", and "Satis ed/very satis ed" were converted to 1-3 points, respectively.

Statistical analysis
The statistical package STATA 15.0 was used for data analysis. Socio-demographic data was analyzed, focusing on its subgroup of CPHs and THCs. Joint distribution analyses of workload and job satisfaction were carried out through Pearson's chi-squared test. Three ordered probit regression models were performed to investigate the association between job satisfaction and three dimensions of workload: CV, WH and PT (see Table 3). In model 1, model 2 and model 3, job satisfaction was regressed on CV, WH and PT respectively. In model 4 the three dimensions of workload were added in the regressors.
Furthermore, open-ended questions and focus group interview were presented as a narrative summary.

Characteristics of respondents
A total of 1,043 rural doctors participated in the study. Finally, 849 questionnaires were included for crosssectional data analysis and 194 questionnaires were excluded because of too much missing data, leading to a response rate as 81.3%. Table 1 presents characteristics of the respondents. Among the 849 respondents, 65.49% were from township healthcare centers (THCs). 58.07% were female, and more than a half (51.83%) were at least 35 years old. The average age of THC respondents was older than that of County public hospitals (CPHs) respondents (p < 0.001).
Workload was indicated by three dimensions. 20.73% of rural doctors thought number of patients seen per day was too many, and there was signi cant difference between CPH doctors (40.61%) and THC doctors (10.25%, with p < 0.001). 52% of rural doctors reported their working hours per day was distributed in the range of 8 to 10 hours. The proportion of CPH doctors who reported work more than 8 hours was higher than that of THC doctors (p < 0.001), though the latter had a more fat tailed distribution. The proportion of time spent on non-medical activities was high in rural doctors (52.18%), with 49.83% in CPH doctors and 53.42% in THC doctors but with no signi cant difference. 40.16% of rural doctors reported they were satis ed with their job. The satisfaction rate was slightly higher in CPH doctors (40.61%) than in THC doctors (39.93%). Bivariate analysis: workload and job satisfaction Pearson correlation analysis is used to explore the correlation between job satisfaction and workload of rural doctors, see Table 2. Result illuminate negative correlations between Chinese rural doctors' job satisfaction and their workload and among which the proportion of time spent on non-medical activities had the biggest correlation coe cient of -0.1979 (P < 0.001). As a further investigation, the relation between clinical visit number and job satisfaction (with the coe cient of -0.2253, P < 0.001) was greater than those of the other two dimensions of workload for CPH doctors, while the proportion of time spent on non-medical activities had the strongest negative association with THC doctors' job satisfaction (with the coe cient of -0.2180, P < 0.001). Working hours was negatively associated with the job satisfaction of CPH doctors (p < 0.01), but not that of THC doctors (p > 0.05, Table 2).

Multivariate analysis
Multivariate analysis of Ordered Probit regression was further conducted to analyze the association between workload and job satisfaction of rural doctors in total sample and two subgroups (CPH doctors and THC doctors) (see Table 3)

Discussion
Although Chinese governments have increased their expenditures on domestic rural healthcare system and allocated more resources and try to recruit and retain more human resources for health to the system, there is still a large gap between supply and demand for health services in rural areas in the context of the new healthcare reform. In this cross-sectional empirical study, workload of rural doctors were found to be heavy and associated with low degree of job satisfaction. Furthermore, there was some notable difference of workload between CPH doctors and their counterparts from THC.

Working hours and job satisfaction
The empirical results show that longer working hours lead to a lower degree of job satisfaction. In previous studies, there is a similar conclusion, that is, working too long will reduce job satisfaction (30,31) .
The operating hours of hospitals are 24 hours a day because of the unique nature of health services.
Many health professionals need to take turns on night shifts. Longer working hours may cause burnout syndrome and harm psychological well-being (3) . In the context of emphasizing the function of public welfare of rural health system, there is still an egalitarian mentality in the compensation system. Longer working hours of doctors cannot therefore be reasonably compensated. Furthermore, long working hours inevitably lead to less time with family, which lead to"Work-Family"con icts. This in turn can lower the job satisfaction of doctors. For professional employees (including doctors),"Work-Family"con ict has a more signi cant negative impact on their job incentives (32) . Hausler (2018) pointed out that "Work-Family"con ict has a greater impact on doctors' job burnout than"Effort-Reward"imbalance (33) .
Compared with western countries, China has a stronger family concept (34,35) ;Compared with cities, the concept of rural families is stronger (35) . Previous study indicated that "Work-Family" con ict has a greater impact on reducing job satisfaction among employees with strong tendency of family collectivism (36) .

Proportion of time spent on non-medical activities and job satisfaction
Proportion of time spent on non-medical activities is another factor in uencing job satisfaction. Research has found out that highly skilled professionals tend to show little interest in work irrelevant to their profession (4) .Healthcare sector is a knowledge intensive industry. Engaging in non-medical activities such as managing administrative affairs, attending meetings and preparing for inspections not only prolongs working hours, but also cause a decrease in work enthusiasm. Previous study shows that doctors in public hospitals must nish their paperwork and administrative work outside normal working hours, which further increases their workload (37) . Many doctors regarded those non-medical activities less meaningful (3) . According to our survey over 50% of the respondents were not satis ed with high proportion of time spent on non-medical activities (see Table 1). In the focus group interviews, quite a few doctors reported that they had to make sure the proportion of drug cost for each outpatient and the medical cost for each inpatient not to exceed the limit. Meanwhile, this is included in the performance assessment system, which is closely associated with their salary. The new healthcare reform positioned THCs as part of the public welfare system; consequently, THCs need to provide a comprehensive set of public health services such as health education, family planning and preventive care. Other forms of nonmedical activities include attending meetings and preparing for inspections. There is insu cient or even no nancial compensation for those non-medical activities, leading to an even lower degree of job satisfaction for rural doctors. Let the professionals focus on their professional work, and spend less time on lling in forms or preparing for inspections. Only in this way, doctors can have more time managing patients.

Clinical visit number and job satisfaction
Clinical visit number had a complex impact on job satisfaction for rural doctors, with a signi cant negative impact on CPH doctors rather than THC doctors. Huy et al. (2018) found that the large number of patients received and the insu cient resources of doctors would lead to heavy workload, which led to the imbalance of doctors' "Effort-Reward" (38) . Some studies have found that the perceived "Effort-Reward" imbalance of medical staff will lead to a decline in their job satisfaction and a strong turnover intention (39) ,and reduce the quality of medical services (40) .The new medical reform removed many barriers to seeking health services for rural patients, leading to a rapid increase in demand for health services. One of the original intentions of the new medical reform is for patients with small illnesses to seek health services from THCs and those with severe illnesses to seek health services from CPHs.
However, the capacity of THCs has been impaired by the reform, hindering THCs to provide more medical services. The capacity of THC to provide medical services has been hampered by many factors such as essential drug list, restrictions on surgical quali cations and the egalitarian mentality in the compensation system (3) . These problems have led to the loss of the backbone of THC, the shrinking of the scope of services and the decline in quality of health services. Doctors in THCs reported that they have seen less patients compared to those of several years ago, though they are equipped with more advanced machinery. "We are better equipped, but cannot retain and recruit talented employees and see less patients." A vice president of a THC in Hebei province said in a focus group interview. Medical service provided by healthcare professionals are regarded as credence goods (17) . Low trust from patients lowers the CV further (41) . Patients from rural areas rushed into CPHs, leading to the increase of workload of CPH doctors. Seeing more patients and also a high proportion of time spent on non-medical activities cause a less job satisfaction for CPH doctors.
Basic medical services have the characteristics of "quasi public goods" (17) , the provision of which cannot solely depend on market mechanism. Therefore, governmental efforts are needed to ensure the public welfare nature of rural health system whilst avoiding the egalitarian mentality in the compensation system. The current compensation and incentive policy is still not in favor of the rural health system, so that rural doctors can't be reasonably compensated for their work (42,43) . A health system is weak and expensive if it is not primary care-led (44) .
Appendix I shows the OLS regression results, which are consistent with the Order Probit regression results, ensuring the robustness of the results.

Limitation
There are several limitations in this study: a) the three dimensions of workload may not capture all details of workload of rural doctors. b) the sample from three provinces could be insu cient to represent the national rural doctors. c) there is lack of comparison of the effect of workload on job satisfaction with that of urban counterparts. These problems need to be noted and corrected in future research.

Conclusions
In summary, our ndings are in line with previous studies, which all highlight the importance of workload on job satisfaction (31,45,46) . In addition, we further analyze the impact of three dimensions of workload on job satisfaction, and discussed the difference of this impact between urban and rural areas. Based on our ndings, we found that clinical visit number had a greater impact on the higher-level hospitals (CPHs), while the proportion of non-medical activities had a greater impact on the low-level hospitals (TPHs).
In order to improve the job satisfaction of rural doctors, we should establish a reasonable compensation mechanism, improve the trust of township hospitals, optimize the proportion of non-medical activities, and let doctors do their own professional work. Enhancing the job satisfaction of rural doctors and reducing their workload should be an end rather than a means to achieve better health service. The top priority of hospital administrators shall be increasing employees' enthusiasm for their work, recognizing employees as the linchpin to deliver better health services.