Chen et al.

BMC Public Health (2019) 19:1526

https://doi.org/10.1186/s12889-019-7894-7

RESEARCH ARTICLE Open Access

Moderating role of job satisfaction on

turnover intention and burnout among
workers in primary care institutions: a
cross-sectional study
Xuyu Chen1†, Li Ran1†, Yuting Zhang1, Jinru Yang2, Hui Yao1, Sirong Zhu1 and Xiaodong Tan1*

Abstract
Background: Global countries are suffering from a shortage of health professionals. Turnover intention is closely
related to job satisfaction and burnout, making good use of these relationships could alleviate the crisis. Our
research aims to examine the mediating role of job satisfaction in the relationship between burnout and turnover
intention.
Methods: This research was conducted in Huangpi, China. The convenience sampling method and self-
administereded questionnaires were used. 1370 of valid samples were collected with 97.72% effective rate.
Descriptive analyses were conducted to describe social demographic factors. The structural equation model (SEM)
was performed to adjust model fitting, and the mediation effect test was carried out by using the bootstrap
method. Sobel-Z test was used to verify the significance of mediation effect.
Results: The mean age was 36.98 (SD = 9.84). The fitting indices of hypothetical model are not good. After the
adjustments, χ2/df = 5.590, GFI = 0.932, AGFI = 0.901, CFI = 0.977, NFI = 0.973, IFI = 0.977, TLI = 0.970, RESEA = 0.058.
The revised model fitted well, and the SEM was put up by using the bootstrap method. The mediating effect is
partial, and Soble-Z test indicates that the mediation effect is significant. Burnout is negatively correlated with job
satisfaction (p < 0.01) and the standardized path coefficient is − 0.41. Job satisfaction is also negatively correlated
with turnover intention (p < 0.01) and the standardized path coefficient is − 0.18. Burnout is positively correlated
with turnover intention (p < 0.01) and the standardized path coefficient is 0.83.
Conclusions: Job satisfaction is a mediating variable that affects the relationship between burnout and turnover
intention. The mediating effect was a partial mediating effect and has a low impact of 7.4%. Improving treatment
and giving more promotion opportunities for workers to improve job satisfaction, conducting career planning
course and paying attention to employee psychological health to reduce job burnout. The above measures may be
helpful to reduce employee turnover rate and alleviating the current situation of a shortage of health personnel in
China.

* Correspondence: 00300469@whu.edu.cn
†
Xuyu Chen and Li Ran are co-first author.
1
School of Health Sciences, Wuhan University, Donghu Road 115, Wuhcang
District, Wuhan 430000, Hubei, China
Full list of author information is available at the end of the article

© The Author(s). 2019 Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0
International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and
reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to
the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver
(http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated.
Chen et al. BMC Public Health (2019) 19:1526 Page 2 of 10

Background Most of researches supported that burnout had a posi-

Turnover is generally viewed as the movement of staff tively strong predictive effect on turnover intention [17–
out of an organization. It was regarded as a two- 22]. For example, emotional exhaustion, negative stagna-
dimensional concept, distinguishing between the act of tion and other factors related to burnout have a signifi-
leaving as voluntary or involuntary, and between the cant positive correlation with turnover intention [23].
leaving and joining of an individual to an organization Beyond that, a study revealed that job satisfaction was
[1]. A previous study [2] defined turnover intention as negative correlated with turnover intention [24]. In the
the next withdrawal behavior when employees encounter study of Guo et al. [25], turnover intention decreased
dissatisfaction. Mobley et al. [3] pointed out that turn- with the increase of job satisfaction, this opinion was
over intention was the intention of a worker to leave an verified from a study of medical workers in Macao. And
organization deliberately after a period of time working Wu et al. [26] also confirmed this result from a research
in a particular organization, after careful consideration, of clinical nurses in Changsha. Therefore, the impact of
which belonged to voluntary turnover. It is considered job satisfaction on turnover intention is important, dir-
as an outcome of affective variables (such as burnout ect and negative. Satisfaction has a significant predictive
and job satisfaction) rather than actual turnover [4]. effect on turnover intention.
That is, turnover intention can predict actual turnover At present, the medical working environment in China
behavior. is tense, more and more medical personnel have left
An American clinical psychologist first proposed the their jobs and changed their posts. As a result, there is a
term “job burnout [5]”. At present, job burnout was de- shortage of medical staff, which destroys the medical en-
fined as the symptoms of practitioners in the service in- vironment and disrupts the medical order and affects
dustry who were unable to cope effectively with the the health of the people. A study based on the National
continuing pressure at work, included emotional exhaus- Health Service Survey in 2013 on China [27] found that
tion, depersonalization and reduced personal accom- the percentage of employees with low, medium and high
plishment [6]. Initially, job satisfaction was described as turnover intentions was 44.5, 43.7 and 11.8% respectively
the physical and psychological satisfaction of staff in in primary care institutions. And a foreign study [28] has
their work [7]. In later studies, it was defined as the investigated the turnover intention of nurses, and also
realization of a person’s work values in the work situ- found that 54% of them have turnover intention and
ation, resulting in a pleasurable emotional state [8, 9]. 35% have turnover behavior. In view of the current
The definition of “the attitude towards one’s work and shortage of medical personnel in various countries [29],
related emotions, beliefs and behaviors” was used in our job satisfaction and turnover intention may therefore be
study, which not only depended on the nature of the a crucial and topic concern.
work, but also the personality, attitude and expectation Most previous research focused on the relationship be-
of medical staff [10, 11]. Low job satisfaction is the most tween job satisfaction, turnover intention and burnout,
common cause of staffs’ turnover and low quality of and exploring the factors affecting the three. However,
health care service. few studies focus on the mediating effect of a certain
There is a plethora of researches on job satisfaction, factor. Such research in China is still in its infancy. The
burnout and turnover intention. A study found that purpose of the research is to examine the mediating role
higher work pressure and lower job satisfaction could of job satisfaction in the relationship between burnout
easily lead to burnout, that is, job satisfaction was one of and turnover intention.
the important predictors of burnout [12]. However, an-
other study of American surgeons had the opposite re- Methods
sult: burnout is the biggest predictor of job satisfaction Participants and setting
[13]. In fact, satisfaction and burnout are in no particular A cross-sectional survey design was used. This quantita-
order and they need to be determined on a case-by-case tive research was conducted in Huangpi District, Wu-
basis. Yin et al. [14] found that satisfaction with work it- han, China. The investigation involved a total of 20
self, occupational risks and off-duty arrangements were public primary care institutions in Huangpi, including
the main factors affecting doctors’ job burnout through 18 township health centers and 2 community health
a survey on doctors in public hospitals. A study showed centers. It conducted from March 2019 to June 2019.
that job satisfaction had a direct predictive effect on The Convenience sampling method was used. Taking
burnout of middle school teacher, and it also existed as into account the scale and level of institutions in rural
a mediating variable between job stress and job burnout area, 25 copies of each institutions were determined.
[15]. Arie R and Yoram N had nosed out that negative Sample size was amplified according to the inefficiency
correlation was existed in job satisfaction and job stress, of 10%, and the final expected sample size was 550. Spe-
burnout [16]. cialized investigators were trained to reduce information
Chen et al. BMC Public Health (2019) 19:1526 Page 3 of 10

bias, inclusion and exclusion criteria was made to reduce shown in Fig. 1 in supplementary materials. There were
selection bias. The inclusion criteria of participants were 1402 questionnaires distributed. Questionnaires with un-
as follows, (1) Eligible participants had at least 6 months’ completed answers or suspected unreal answers were ex-
work experiences in their own workplace. (2) An em- cluded. Finally, a total of 1370 of valid samples was
ployee who had not suffered from mental illness and collected with 97.72% effective rate. All responses were
had not been stimulated by major adverse life events in anonymous to protect the privacy of participants.
the near future. (3) Participants were voluntary. And we
excluded staffs with a working time of less than 6 Measures
months, and employees who were not in the post during We used self-administered questionnaires, which were
the investigation. A flow diagram of participants was classified into three parts, job satisfaction, burnout and

Fig. 1 The hypothetical model of the relationship between job satisfaction, burnout and turnover intention
Chen et al. BMC Public Health (2019) 19:1526 Page 4 of 10

turnover intention. Then we adjusted the questionnaire performed to adjust model fitting, the mediation effect
according to the pilot survey, actual situation and local test was carried out by using the bootstrap method.
culture. All the measures were followed the translation Sobel-Z test was used to verify the significance of medi-
and back-translation process from English to Chinese ation effect. Using SPSS 20.0 (IBM Corp, Armonk, NY,
[30]. The three scales all use the 5-point Likert scale. USA) and AMOS 24.0 (IBM Corp, Armonk, NY, USA)
Content of each scale was shown in Table 1 in supple- to analyze data, and p < 0.05 was determined to signifi-
mentary materials. The KMO measure and Bartlett’s cant in statistics.
spherical test were used to test construct validity, it was
acceptable if values of KMO measure were greater than Results
0.50 and p value of Bartlett’s spherical test less than Descriptive statistics
0.05. Cronbach’s alpha coefficient was calculated to Table 1 showed the sociodemographic characteristics of
examine internal consistency reliability, values higher the respondents. The mean age was 36.98 ± 9.84 years
than 0.70 were considered satisfactory. (minimum: 18 years, maximum: 73 years). 426 (31.09%)
medical staffs were male and 944 (68.91%) were female.
Job satisfaction scale 42.77% of participants were physicians and 41.02% were
Job satisfaction was measured with 18 items selected nurses. The largest number of participants in the 40–49
from the Minnesota Satisfaction Questionnaire (MSQ) age group, accounting for 34.45%, while the group over
[31] and the Job Satisfaction Survey (JSS) [32]. The con- the age of 49 accounts for 11.31%. Most participants
tent of job satisfaction included satisfaction with envir- (77.30%) were married, 724 (52.58%) had a junior title
onment, remuneration, management, the work itself [22, and 63 % had no night shift in their work.
33]. Sample item includes “The comfort level of the
working environment (office environment, greening, Structural equation model constructing and fitting
lighting) will satisfy you.” (KMO measure =0.957, p < The Hypothetical model was established, as shown in
0.01, Cronbach’s α = 0.970). Fig. 1. We have constructed four paths: (1) Path a: Path
from independent variable to potential mediator vari-
Burnout scale able, the path coefficient of path a represents the indir-
We used 5 items from the Maslach Burnout Inventory ect effect of burnout to job satisfaction (Job satisfaction
(MBI) [34] to measure individual burnout, and aggre- ← Burnout). (2) Path b: The path from potential medi-
gated it to measure a positive effect on the burnout. Par- ator variable to dependent variable, the path coefficient
ticipants respond to the following items: “I feel that my of path b represents the indirect effect of job satisfaction
daily [35] work is meaningless”, “I can’t find a sense of to turnover intention (Turnover intention ← Job satis-
accomplishment at work”, “I feel exhausted when I get faction). (3) Path c: the path from independent variable
off work every day”, “this job has made me indifferent” to dependent variable, the path coefficient of path c rep-
and “this job makes me feel restless”. (KMO measure = resents the total effect of burnout to turnover intention
0.857, p < 0.01, Cronbach’s α = 0.925). (Turnover intention ← Burnout). (4) Path c’: Under the
influence of potential mediator variables, the path from
Turnover intention scale the independent variable to the dependent variable, the
The turnover intention questionnaire was designed with path coefficient of path c’ represents the direct effect of
reference to turnover intention scale explored by Grif- burnout to turnover intention (Turnover intention’ ←
feth [36]. We measured turnover intention using 5 Burnout’).
items. Participants respond to the following items: “I had As shown in Table 2. From the results of the hypo-
the idea of leaving this organization”, “within a year, I thetical model operation, we found that all fitting indices
will go to find a new job”, “If there is an opportunity, I did not meet the fitting criteria, indicated that the hypo-
will definitely accept a better job”, “I think the employ- thetical model was not ideal, so we revised the model.
ment situation in this organization is very good” and The model path was modified according to the amend-
“Currently, I agree to find a good job in the market”. ment advice given by AMOS. We removed some items
(KMO measure =0.800, p < 0.01, Cronbach’s α = 0.721). (A18, B1 and C4) and added lots of bidirectional arrows
to make the model fitting better. The final fitting indices
Statistical analysis results were also shown in Table 2, and the revised stan-
Data entry and conversion was completed with Epi- dardized path coefficient map was displayed on Fig. 2.
Data 3.0. Double machine imputing method was used to After the adjustments, validity and reliability of the three
enter the collected data into the computer. Descriptive scales remained acceptable. KMO measure, p value for
analyses were conducted to describe social demographic batrtlett’s spherical test and Cronbach’s α for job satis-
factors. The structural equation model (SEM) was faction is 0.957, < 0.01 and 0.976, respectively; For
Chen et al. BMC Public Health (2019) 19:1526 Page 5 of 10

Table 1 Demographic characteristics of the respondents (n = 1370)

Variables Group n % Means (SD a)
Gender Male 426 31.09 –
Female 944 68.91
b
Age(year) < 30 438 31.97 36.98 (9.84)
30–39 305 22.26
40–49 472 34.45
> 49 155 11.31
Occupation Physician 586 42.77 –
Nurse 562 41.02
Medical technician 92 6.72
Public health personnel 66 4.82
Pharmacist 34 2.48
Others 30 2.19
Educational background University and above 423 30.88 –
College 636 46.42
High school/Technical school 286 20.88
Junior high school and below 25 1.82
Marital status Married 1059 77.30 –
Unmarried 270 19.71
Divorced/widowed 41 2.99
Professional title Senior title 44 3.21 –
Middle title 207 15.11
Junior title 724 52.85
No title 395 28.83
Monthly income (RMB) > 5000 55 4.01 –
4001~5000 274 20.00
3001~4000 519 37.88
2001~3000 413 30.15
< 2000 109 7.96
Hire form Personnel agent staff 205 14.96 –
permanent staff 221 16.13
Contract staff 185 13.50
Temporary staff 759 55.40
b
Working time (hours a week) < 30 18 1.31 42.92 (8.48)
31–40 973 71.02
41–50 235 17.15
> 50 144 10.51
b
Working years 1–5 404 29.49 14.65 (10.87)
6–10 230 16.79
11–15 134 9.78
15–20 169 12.34
21–25 149 10.88
> 25 287 20.95
Frequency of night shift (times a week) 0 868 63.36 –
1–3 450 32.85
>3 52 3.80
a b
Note: SD, standard deviation, this indicator was calculated only for quantitative data. represented quantitative data
Chen et al. BMC Public Health (2019) 19:1526 Page 6 of 10

Table 2 Comparison of different fitting indices on hypothetical model and adjusted model
Fit index χ2 a χ2/df b
GFI c
AGF d CFI e
NFI f
IFI g TLI h RMSEA i AIC j
Optimum model – 2–5 > 0.90 > 0.90 > 0.90 > 0.90 > 0.90 > 0.90 < 0.05 –
Hypothetical model 9695.143 27.940 0.600 0.533 0.801 0.796 0.802 0.784 0.140 9813.143
Results – unfit unfit unfit unfit unfit unfit unfit unfit –
Adjusted model 1157.059 5.590 0.932 0.901 0.977 0.973 0.977 0.970 0.058 1343.059
Results – acceptable fit fit fit fit fit fit acceptable –
Note: a χ2, Chi-square. b χ2/df, Chi-square divided by degree of freedom. c GFI Goodness of fit index, d AGFI Adjusted goodness of fit index, e CFI Comparative fit
index, f NFI Normed fit index, g IFI Incremental fit index. h TLI Tucker-Lewis index. I RMSEA Root mean square error for approximation. j AIC Akaike
information criterion

turnover intention is 0.857, < 0.01 and 0.910, respectively; As shown in Fig. 2, the standardized path coefficient
For burnout is 0.798, < 0.01 and 0.879, respectively. of burnout to job satisfaction is − 0.41, indicates that
As shown in Table 3, we estimated the significance of burnout is negatively correlated with job satisfaction
total effect, direct effect and indirect effect by bias- (p < 0.01). It shows that when the other conditions are
corrected approach. The results showed that the total ef- unchanged, the turnover intention decreases by 0.41
fect was significant of independent variable (burnout) to units for each additional unit of burnout. The standard-
dependent variable (turnover intention) (p < 0.01), that ized path coefficient of job satisfaction to turnover
was, the total effect of path c was statistically significant. intention is − 0.18, demonstrates that job satisfaction is
The direct effects of path a (Job satisfaction Burnout) also negatively correlated with turnover intention (p <
(p < 0.01), path b (Turnover intention ← Job satisfac- 0.01). Under the same other conditions, the turnover
tion) (p < 0.01) and path c’ (Turnover intention’ ← Burn- intention decreases by 0.18 units for each additional unit
out’) (p < 0.01) were also significant. The results of of job satisfaction. The standardized path coefficient of
indirect effect test again proved that path c’ (Turnover burnout to turnover intention is 0.83, reveals that burn-
intention’ ← Burnout’) was statistically significant. We out is positively correlated with turnover intention (p <
concluded that this mediating effect was a partial medi- 0.01). That is, under the influence of job satisfaction, the
ating effect. turnover intention increases by 0.83 units for each add-
As shown in Table 4, standardized estimates and its itional unit of burnout.
standard errors were calculated by using the bootstrap The mediation effect is statistically significant (p <
method. Standardized path coefficient of path a (Job Sat- 0.01), and the impact of burnout on turnover intention
isfaction ← Burnout) is − 0.410 and its standard error through the intermediary effect of job satisfaction is
(Sa) is 0.038. Standardized path coefficient of path b 0.074 (a*b = (− 0.180)*-(0.410)). It manifests that when
(Turnover intention ← Job satisfaction) is − 0.180 and other conditions remain unchanged, the turnover
its standard error (Sb) is 0.028. Standardized path coeffi- intention will be indirectly increased by 0.074 units for
cient of path c’ (Turnover intention’ ← Burnout’) is each unit of burnout.
0.824 and its standard error (Sc′) is 0.029. Looking up
the related tables, the standardized path coefficient of Discussion
path c is 0.899 and its standard error (Sc) is 0.020. Our results demonstrated that for medical workers in
Soble-Z test is carried out according to the formula z primary care institutions, a mediator variable was existed
in burnout and turnover intention: job satisfaction. Job
¼ pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
ab ffi . Finally, z = 4.506. According to MacKin-
satisfaction was usually regarded as a dependent variable
sa b þsb a
2 2 2 2

non’s critical value table, the result is p < 0.05, indicating [40–42] or an independent variable [43, 44] in most of
that the mediation effect is significant. the current studies. And work to family conflict [42],
work engagement [40], burnout and workload [45] were
viewed as mediator variables. However, an American
Interpretation of the revised model study suggested that burnout was the biggest predictor
Model fit is acceptable if χ2/df ≤ 4.0 [37], GFI > 0.90, of job satisfaction [13]. There were few researches focus
AGFI > 0.90, CFI > 0.90, NFI > 0.90, IFI > 0.90 [38], TLI on job satisfaction as a mediator variable, which provides
> 0.90 and RMSEA < 0.05. As shown in Table 2, all fit new ideas for future research. That’s why we try to study
indexes are up to standard, except for χ2/df and RMSEA. how job satisfaction as a mediating variable affects the
Our sample size is larger than 1000, the value of χ2/df is correlation between burnout and turnover intention.
acceptable. In another study, the author points out that The results of the study provided further support for the
0.05 < RMSEA < 0.08 is also acceptable [39]. Overall, the importance of job satisfaction in engaging the workforce
model fits well and the model is established. and retaining staff to settle the demands and challenges
Chen et al. BMC Public Health (2019) 19:1526 Page 7 of 10

Fig. 2 The revised structural equation modeling of the relationship between job satisfaction, burnout and turnover intention

facing health care setting in primary care institutions. intention ← Job satisfaction) is relatively small, only −
Turnover intention was negatively related to job satisfac- 0.18. We speculated that traditional studies which usually
tion and positively related to burnout. Negative correl- used methods such as multivariate linear regression, logis-
ation was found between job satisfaction and burnout. tic regression, and ANOVA may neglect the measurement
Some studies conducted in China have also obtained simi- error, so its negative correlation is stronger. However, the
lar results [14, 15, 22–24, 26, 27, 46], but the correlation error is taken into account in the SEM.
coefficient of their results is greater than ours. In our Our finding showed that the mediating effect was a
study, the correlation coefficient of path b (Turnover partial mediating effect. First of all, we need to realize
Table 3 The p-value of significance test of total effect, direct effect and indirect effect by bias-corrected approach
Burnout Job satisfaction Turnover intention
Total effect Job satisfaction 0.001 –
Turnover intention 0.001 0.001 –
Direct effect Job satisfaction 0.001 – –
Turnover intention 0.001 0.001 –
Indirect effect Job satisfaction – – –
Turnover intention 0.001 – –
Chen et al. BMC Public Health (2019) 19:1526 Page 8 of 10

Table 4 Standardized path coefficient and standard error of three main paths by bootstrap
Path SE SE-SE Mean Bias SE-Bias
Job Satisfaction ← Burnout 0.038 0.001 −0.410 0.000 0.001
Turnover ← Job Satisfaction 0.028 0.000 −0.180 −0.002 0.001
Turnover’ ← Burnout’ 0.029 0.000 0.824 −0.002 0.001
Note: SE standard error. SE-SE standard error caused by using bootstrap to estimate standard errors. SE-Bias the standard error of bias

what the difference between partial mediation and Our study also has some limitations. Firstly, Because
complete mediation is. In the process of partial medi- of the diversity of the participants (including doctors,
ation, any variable in the causal chain, when it controls nurses, technicians.), we don’t use the international scale
the variable before it (including the independent vari- completely. According to the research purpose, the ap-
able), it will affect its subsequent variables significantly. propriate items have been chosen from these inter-
And in the process of complete mediation, after control- national scales, and the reliability and validity of the
ling the mediation variables, the influence of the inde- questionnaire are still guaranteed. Secondly, sample rep-
pendent variables on the dependent variables is not resentativeness needs to be improved. This research can
significant. Job burnout is closely related to turnover not be generalized to all china as our study place limited.
intention, and job burnout affects turnover intention dir- In the future, we will continue to cooperate with other
ectly [17–22], thus causing a partial mediating effect. local governments in central China and to conduct simi-
Although the mediation effect test confirms that the ex- lar surveys to solve the problem.
istence of job satisfaction as a mediating variable of job
burnout affects turnover intention, the mediation effect has Conclusions
a low impact of 7.4%. That is to say, the mediation effect Our study provides a clear understanding of how job
accounts for 7.4% of the variation of dependent variables. satisfaction can mediate the relationship between burn-
Despite the low mediation effect, we still believe that the re- out and turnover intention. In the process of burnout af-
search is valuable and meaningful. We speculated that the fecting turnover intention, job satisfaction can be
low mediation effect was caused by the following reasons. regarded as a mediating variable to influence its effect,
Firstly, it is related to the choice of independent variables and the mediating effect was a partial mediating effect.
closely. Job burnout includes three dimensions: emotional And the mediation effect has a low impact of 7.4%.
exhaustion, depersonalization and reduced personal Turnover intention was negatively related to job satisfac-
accomplishment [6]. For example, family conflict and tion and positively related to burnout, job satisfaction
doctor-patient relationship are more intuitive than job was negative related to burnout. We can make full use
burnout as important factors that directly affect emotions. of this relationship to adjust the impact of job burnout
If we take them as independent variables directly, the re- on turnover intention by improving job satisfaction.
spondent would understand the meaning of the question Some operable and useful measures were taken to re-
more clearly, therefore, the path coefficients of path a duce employee turnover rate and alleviating the current
(mediator variable ← independent variables) would be lar- situation of shortage of health personnel in China, such
ger than those of the study (job satisfaction ← burnout), as improving treatment and giving more promotion op-
which will lead to a higher mediation effect. Secondly, portunities for workers to improve job satisfaction, con-
turnover intention, burnout and job satisfaction are diffi- ducting career planning courses and paying attention to
cult to measure directly. The items and measuring employee psychological health to reduce job burnout.
methods may be somewhat various in different studies
[47]. Therefore, the difference of instrument selection is Supplementary information
also one of the important reasons for the inconsist- Supplementary information accompanies this paper at https://doi.org/10.
1186/s12889-019-7894-7.
ency of results. Thirdly, the influence of job satisfac-
tion on turnover intention is limited. There are many
Additional file 1: Figure S1. A flow diagram of participants.
factors affecting turnover intention. Turnover
Additional file 2: Table S1. Contents of three scales in the
intention is influenced not only by job satisfaction, questionnaire.
but also by social demographic factors such as educa-
tion [43], years in work [43], family’s relationship Abbreviations
[43], monthly income, social support [48], mentoring AGFI: Adjusted goodness of fit index; AIC: Akaike information criterion;
[49], etc. and other unobserved factors. Therefore, ANOVA: Analysis of Variance; CFI: Comparative fit index; GFI: Goodness of fit
index; IFI: Incremental fit index; JSS: Job Satisfaction Survey; MBI: Maslach
these demographic factors and unobserved factors Burnout Inventory; MSQ: Minnesota Satisfaction Questionnaire; NFI: Normed
should be taken into account in the future study. fit index; RMSEA: Root mean square error for approximation; Sa: The standard
Chen et al. BMC Public Health (2019) 19:1526 Page 9 of 10

error of standardized path coefficient of path a; Sb: The standard error of 10. Peters DH, Chakraborty S, Mahapatra P, Steinhardt L. Job satisfaction and
standardized path coefficient of path b; Sc: The standard error of motivation of health workers in public and private sectors: cross-sectional
standardized path coefficient of path c; Sc′: The standard error of analysis from two Indian states. Hum Resour Health. 2010;8(1):27.
standardized path coefficient of path c’; SD: Standard Deviation; SE: Standard 11. Wang H, Tang C, Zhao S, Meng Q, Liu X. Job satisfaction among health-care
Error; SE-Bias: The standard error of bias; SEM: Structural Equation Model; SE- staff in township health centers in rural China: results from a latent class
SE: The standard error caused by using bootstrap to estimate standard errors; analysis. Int J Environ Res Public Health. 2017;14(10):1101.
TLI: Tucker-Lewis index; χ2: Chi-square; χ2/df: Chi-square divided by degree of 12. Visser MR, Smets EM, Oort FJ, De Haes HC. Stress, satisfaction and burnout
freedom among Dutch medical specialists. Cmaj. 2003;168(3):271–5.
13. Shanafelt TD, Balch CM, Bechamps GJ, Russell T, Dyrbye L, Satele D, Collicott
Acknowledgements P, Novotny PJ, Sloan J, Freischlag JA. Burnout and career satisfaction among
We would like to thank Wuhan University and Huangpi District Center for American surgeons. Ann Surg. 2009;250(3):463–71.
Disease Control and Prevention for their support of this project, as well as 14. Yin W, Wang Z, Fan Y, Tian J, Qin S, Sun K, Li Y. Analysis of current situation
the efforts of team partners in the project. and influencing factors of job burnout of doctors. Chin J Hosp Admin. 2008;
24(3):184–7.
Authors’ contributions 15. Xiang G. A study on middle school teachers' job burnout, job stress, job
XyC participated in the survey, the data analysis and the writing of the satisfaction and their relationship. Wuhan: Central China Normal University; 2005.
article. LR took part in the design of the study and the writing of the article. 16. Reichel A, Neumann Y. Work stress, job burnout, and work outcomes in a
YtZ and JrY contributed to the data collection and screening. HY and SrZ turbulent environment. Int Stud Manag Org. 1993;23(3):75–96.
were involved in the data analysis and participated in the literature research. 17. Estryn-Behar M, Van-Der-Heijden BH, Camerino D, Le-Nezet O, Conway P,
XdT made a second revision of the manuscript. All authors have read and Fry C, Hasselhorn H. The impact of social work environment, teamwork
approved the final version. characteristics, burnout, and personal factors upon intent to leave among
European nurses. Med Care. 2007;45(10):939–50.
Funding 18. Hongryun W, Hyunhee K, Sangmin P. Burnout and turnover intentions
Not applicable. among junior counseling faculty: moderating role of mentoring. J Employ
Couns. 2019;56(2):85–94.
Availability of data and materials 19. Rachel W, Margae K, Beatrice H, Hali H, Coleen K, Kevin G. Burnout and
The data that support the findings of this study are available from the CDC health care workforce turnover. Ann Fam Med. 2019;17(1):36–41.
of Huangpi district but restrictions apply to the availability of these data, 20. Shimizu T, Feng Q, Nagata S. Relationship between turnover and burnout
which were used under license for the current study, and so are not publicly among Japanese hospital nurses. J Occup Health. 2005;47(4):334–6.
available. Data are however available from the authors upon reasonable 21. Lee RT, Ashforth BE. A meta-analytic examination of the correlates of the
request and with permission of the CDC of Huangpi district. three dimensions of job burnout. J Appl Psychol. 1996;81(2):123–33.
22. Zhang Y. The model study on the relationship between job satisfaction,
Ethics approval and consent to participate career burnout and turnover intention among physicians from urban state-
The ethics committee of Wuhan University School of Medicine (WUSM) owned medical institutions. Shanghai: Fudan University; 2011.
reviewed it, and verified it to comply with the Declaration of Helsinki and its 23. Bing Z, Lei Z, Na Z. Research on the relationship between job burnout and
revised version, as well as the relevant regulations of biomedical journals, turnover intention among clinical nurses born in the 80s. J Nurs Sci. 2013;
and approved the research (No.2018YF0080). All participants were provided 28(8):16–18.
with information about the investigation and gave written informed consent 24. Chang X. Study on the relationship among job satisfaction, Career Burnout
to participate. and Intent to Stay in General Practitioners. Jinan: Shangdong University;
2015.
Consent for publication 25. Guo B. Relevant study on job characteristics, Job Satisfaction and Turnover
Not applicable. Intention of Medical Nurses in Macao. Guangzhou: South China Normal
University; 2007.
Competing interests 26. Wu L. Study of work satisfaction and turnover intention among clinic nurses
The authors declare that they have no competing interests. in Changsha. Changsha: Central South University; 2007.
27. WangShuai CM. XuLing, MengQun: Analysis of turnover intention of grass-
Author details roots medical staff in China. Chin J Health Inform Manag. 2016;13(2):206–13.
1 28. Gardulf A, Söderström IL, Orton ML, Eriksson LE, Arnetz B, Nordström G. Why
School of Health Sciences, Wuhan University, Donghu Road 115, Wuhcang
District, Wuhan 430000, Hubei, China. 2School of Clinical Medicine, Wuhan do nurses at a university hospital want to quit their jobs? J Nurs Manage.
University of Science and Technology, 947 Heping Avenue, Qingshan District, 2010;13(4):329–37.
Wuhan 430000, Hubei, China. 29. Guilbert JJ. The world health report 2006: working together for health. Educ
Health (Abingdon). 2006;19(3):385–7.
Received: 21 July 2019 Accepted: 1 November 2019 30. Brislin RW. Back-translation for cross-cultural research. J Cross-Cult Psychol.
1970;1(3):185–216.
31. Weiss DJ, Dawis RV, England GW. Manual for the Minnesota satisfaction
References questionnaire. In: Minnesota Studies in Vocational Rehabilitation, vol. 22; 1967.
1. Bluedorn AC. A taxonomy of turnover. Acad Manag Rev. 1978;3(3):647–51. 32. P ES. Job satisfaction. Thousand Oaks: Sage publication; 1997.
2. Porter LW, Steers RM. Organizational, work, and personal factors in 33. Meng R, Li J, Zhang Y, Yu Y, Luo Y, Liu X, Zhao Y, Hao Y, Hu Y, Yu C. Evaluation
employee turnover and absenteeism. Psychol Bull. 1973;80(2):151–76. of Patient and Medical Staff Satisfaction regarding Healthcare Services in
3. Mobley WH. Intermediate linkages in the relationship between job Wuhan Public Hospitals. Int J Environ Res Public Health. 2018;15(4):769.
satisfaction and employee turnover. J Appl Psychol. 1977;62(2):237–40. 34. Sueoka N, Nisigaki H, Yonezawa M, Tsukui T, Sakamoto C, Tabuchi M. The factorial
4. Hellman CM. Job satisfaction and intent to leave. J Soc Psychol. 1997;137(6): validity of the Maslach burnout inventory-general survey (MBI-GS) across
677–89. occupational groups and nations. J Occup Organ Psychol. 2011;73(1):53–66.
5. Freudenberger HJ. Staff Burn-Out. J Soc Issues. 2010;30(1):159–65. 35. Maissiat GS, Lautert L, Pai DD, Tavares JP. Work context, job satisfaction and
6. Maslach C, Jackson SE. The measurement of experienced burnout. J Organ suffering in primary health care. Revista Gaúcha De Enfermagem. 2015;36(2):42–9.
Behav. 1981;2(2):99–113. 36. Griffeth RW, Hom PW. A comparison of different conceptualizations of
7. Hoppock R. Age and job satisfaction. Psychol Monographs. 1936;47(2):115–8. perceived alternatives in turnover research. J Organ Behav. 1988;9(2):103–11.
8. Bussing A, Bissels T, Fuchs V, K-M P. A dynamic model of work satisfaction: 37. Bollen KA, Long JS. TESTING STRUCTURAL EQUATION MODELS. Bms Bulletin
qualitative approaches. Hum Relat. 1999;52(8):999–1028. Sociol Methodol. 1993;23(39):66–7.
9. Locke E, Latham G. Work motivation and satisfaction: light at the end of the 38. Hair JF, Black WC, Babin B, Anderson RE, Tatham RL. Multivariate data
tunnel. Psychol Sci. 1990;1(4):240–6. analysis. 7th ed. New York: Prentice Hall; 2009.
Chen et al. BMC Public Health (2019) 19:1526 Page 10 of 10

39. Byrne BM. Structural equation modeling with AMOS. Basic concepts,
applications, and programming; 2009.
40. Van Bogaert P, Peremans L, Van Heusden D, Verspuy M, Kureckova V, Van
de Cruys Z, Franck E. Predictors of burnout, work engagement and nurse
reported job outcomes and quality of care: a mixed method study. BMC
Nurs. 2017;16(1):5.
41. Marie-Josée F, Guy G, Jean-Marie B, François C. Associated and mediating
variables related to job satisfaction among professionals from mental health
teams. Psychiat Quart. 2018;2(89):415.
42. D X, Y Y, S D, Z T, L L, L H. Can Job Control Ameliorate Work-family Conflict
and Enhance Job Satisfaction among Chinese Registered Nurses? A
Mediation Model. Int J Occup Environ Med. 2018;2(9):97–105.
43. Chen I, Brown R, Bowers BJ, Chang W. Work-to-family conflict as a mediator
of the relationship between job satisfaction and turnover intention. J Adv
Nurs. 2015;71(10):2350.
44. Billie C, B KL. Impact of job satisfaction components on intent to leave and
turnover for hospital-based nurses: a review of the research literature. Int J
Nurs Stud. 2007;44(2):297–314.
45. Van Bogaert P, Clarke S, Willems R, Mondelaers M. Nurse practice
environment, workload, burnout, job outcomes, and quality of care in
psychiatric hospitals: a structural equation model approach. J Adv Nurs.
2013;69(7):1515–24.
46. Meng J, Wang J. Reasons and countermeasures of clinical nurses'turnover
intention. Fam Nur. 2007;5(5):61–2.
47. Pinder CC. Work motivation in organizational behavior. Upper Saddle River:
Prentice-Hall; 1998.
48. Duan XJ, Ni X, Shi L, Zhang LJ, Ye Y, Mu HT, Li Z, Liu X, Fan LH, Wang YC.
The impact of workplace violence on job satisfaction, job burnout, and
turnover intention: the mediating role of social support. Health Qual Life
Out. 2019;17(93). https://doi.org/10.1186/s12955-019-1164-3.
49. Woo H, Kim H, Park S. Burnout and turnover intentions among junior
counseling faculty: moderating role of mentoring. J Employment Couns.
2019;56(2):85–94.

Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in
published maps and institutional affiliations.