Analysis of Factors Affecting Job Satisfaction and Impact to Burnout on Hopsitalization Nurses in General Hospital

Burnout is a pathological syndrome in which a maladaptive emotion develops in response to prolonged work stress. One of the causes of burnout is low job satisfaction, so it can affect various aspects of work. Job related factors are usually considered acceptable for interventions or strategies to increase satisfaction. Factors from job related itself are working conditions, work autonomy, and work family conflict. Purpose: to determine the factors that influence job satisfaction on nurses at a government hospital in Pemalang Regency. Design: From the total of 160 respondents, there were 30 male respondents and 130 female respondents. The majority group of respondents was in the range of 31 - 40 years old. In terms of education, 48.8% with D3 Nursing education. 94.4% of respondents are married. On average, respondents have worked for > 5 years (85%) and 91.9% of respondents have working hours of 41 – 59 hours/week. The results show that job satisfaction is proven to have negative effects on burnout. Result: The result model (empirical model) of this study is known to have substantial small predictive accuracy (R2) and medium predictive relevance (Q2) on burnout as the predicted dependent variable. Conclusion: Therefore, it can be concluded that the antecedents of job satisfaction in this research model can adequately predict job satisfaction variables. Thus, this research model can be suggested to be replicated and tested further in other nurse populations at other hospitals that are potential for future research.


INTRODUCTION
Burnout is a pathological syndrome in which a maladaptive emotion develops in response to prolonged work stress. Burnout in healthcare can drain energy, reduce productivity, and increase the risk of medical errors (Marzouki et al., 2019;Wang et al., 2020). One of the causes of burnout is low job satisfaction, which can affect various aspects of work (Wang et al., 2020). Job satisfaction for health workers is an important quality indicator for the performance of any health care system. High job satisfaction for health workers is a priority for hospital organizations. Hospitals must create favorable and decent working conditions for health workers, and provide opportunities for them to develop professionally (Afulani et al., 2021) Plenary health services are health services that include promotive, preventive, curative, and rehabilitative health care, so that the main role in hospitals is health workers, one of which is a nurse (Undang-Undang Republik Indonesia Tentang Rumah Sakit, 2009). A nurse is someone who has graduated from a higher nursing education, both at home and abroad, which is recognized by the Government in accordance with the provisions of the legislation (Undang-Undang Republik Indonesia Tentang Keperawatan, 2014). Based on the Decree of the Coordinating Minister for People's Welfare No. 54 of 2013 concerning the Health Personnel Development Plan for 2011 -2025, the target ratio of health workers to the population for nurses is 200 nurses per 100,000 population (Dinas Kesehatan Kabupaten Pemalang, 2020). Pemalang is one of the regencies in Indonesia that has a total nurse availability ratio of 78.42 per 100,000 population (Dinas Kesehatan Kabupaten Pemalang, 2020). RSUD dr. M. Ashari is the only local government-owned hospital located in Pemalang Regency, Central Java. This hospital was accredited with plenary status in 2019 and became one of the referral sites for the entire Pemalang Regency. RSUD Dr. M. Ashari, Pemalang Regency, has a vision to be the first choice hospital for the people of Pemalang and its surroundings (RSUD dr. M. Ashari Kabupaten Pemalang, 2021). This shows that there is still a lack of nurses in improving health services and reducing the medical burden experienced by nurses. To close the gap between the large volume of patients and limited health resources, health facilities require health workers to increase their workload (Zhou et al., 2018) Working conditions include the work environment and all conditions that affect the workforce in the workplace (Manyisa & van Aswegen, 2017)Working conditions affect job satisfaction. The working condition itself will also affects the health of workers.
Unhealthy working conditions will contribute to the development of occupational diseases and emotional disorders such as burnout (Molina Zavala et al., 2022).
Work autonomy has a positive relationship with job satisfaction. Job description, nature of work, and work autonomy are significant variables on job satisfaction in hospitals. Nurses are one of the health professionals. Health professionals who have autonomy in decision-making have 3.0 times higher job satisfaction than their collagues (Geta et al., 2021).
Work-family conflict is a type of conflict that is created between the pressures or demands of work and family simultaneously. The demands of work usually prevent a person from fulfilling the needs, roles, and responsibilities of their family (Ekici et al., 2020). The shortage of nurses can also lead to excessive workloads, conflicts related to scheduling of work hours (hours, shift rotation, and working hours that are too long) which will indirectly affect social and family life (work-family conflict) (Molina Zavala et al., 2022).
Pemalang Regency has one government hospital that is accredited with plenary status in 2019 and is one of the referral hospital in Pemalang Regency. The lack of nurses in Pemalang Regency and the hospital causes an increase in the workload experienced by nurses to close the gap between patient volumes (Zhou et al., 2018). Meanwhile, job satisfaction for nurses is an important factor for the effective functioning of medical facilities (Zhou et al., 2018). Low job satisfaction can cause work stress and burnout in nurses so this needs to be taken into consideration for the welfare of the nurses (Marzouki et al., 2019). One of the factors that influence job satisfaction on nurses is the job related factors. Job related factors are usually considered acceptable for interventions or strategies to increase satisfaction. Factors from job related itself are working conditions, work autonomy, and work family conflict (Zhou et al., 2018).
The purpose of this study was to determine the factors that influence job satisfaction on nurses at a government hospital in Pemalang Regency. The variables included in this study were working conditions, work autonomy, and work-family conflict to determine the variables that affect nurse job satisfaction and their impact on burnout in nurses.

RESEARCH METHODOLOGY
This research used a crosssectional data approach with quantitative survey methods. The independent variables in this study were working conditions, work autonomy, and work-family conflict. The dependent variables of this study were burnout and job satisfaction as mediating variables. This research was conducted at one of the government hospitals in Pemalang Regency from September -October 2022. We chose this timeframe because it indicates that the nurse is currently working at RSUD Dr. M. Ashari Pemalang when we took the sample. The sample of this study was hospitalization nurses who had met the inclusion criteria, which was hospitalization nurses who worked in August -October 2022 and the exclusion criteria, which was not working in shift work. The number of samples in this study was 160 respondents who were taken from each of the 9 hospitalization rooms as many as 18 respondents.
It measured by a structured questionnaire instrument. The measurement of variables in this study uses a five-point Likert scale: (1) Strongly disagree, (2) disagree, (3) neutral, (4) agree, and (5) strongly agree (Hair et al., 2019;Sarstedt et al., JENIUS 574 2021). The questionnaire used in this study was modified from the previous research which was already valid and reliable.
The data were analyzed using a multivariate statistical method using PLS-SEM (Partial Least Square -Structural Equation Model). Partial least squares path modeling (PLS-SEM) aims to test the predictive relationships between constructs by looking at the relationships or influences between those constructs. There were outer model analysis and inner model analysis. Outer model analysis saw the reliability and validity of the indicators. There were 4 stages in the outer model that must be monitored which were indicator reliability (outer loading), construct reliability (Cronbach's Alpha and Composite Reliability), construct validity (average variance expected), and discriminant validity (Heterotrait-Monotrait Ratio). Inner model analysis tests the significance of the variables used as well as the predictive and explanatory capabilities of the model with the bootstrapping function. The quality parameters of the research model analyzed in the inner model were variance inflation factor (VIF), Rsquare, f-square, and Q-square (Hair et al., 2019;Sarstedt et al., 2021).

RESULT
There were 160 eligible respondents whose characteristics are described in Table 1 Analyzing the outer model was the step of PLS-SEM analysis. The outer loading proposed was done to assess the indicator of reliability. In PLS-SEM, the indicator could be reliable when the outer loading value was greater than 0.708 (Sarstedt et al., 2021). The results in this study show that 25 indicators have been reliable to measure their respective constructs ( Figure 1 and Table 2). The next step of the analysis was to test internal consistency. Constructs became reliable of the respective model if the constructs showed that Cronbach's alpha greater than 0.7 (Sarstedt et al., 2021). In this study, the Cronbach's alpha value was greater than the value of 0.7 (Table  3). The third step was measuring Average Variance Extracted (AVE) to assess the convergent validity. This result showed that all of the constructs had an AVE 0.50 as required by the literature (Table 3) (Sarstedt et al., 2021). All of the constructs could explain at least 50 percent of item variance in the model, thus establishing convergent validity (Sarstedt et al., 2021). The last step in the outer model analysis was to check the discriminant validity by Heterotrait-Monotrait Ratio (HTMT). The recommended threshold value for HTMT ratio was below 0.90 referring to the literature, which established that each construct indicator was conceptually different (Sarstedt et al., 2021). All of the HTMT results showed that all of the values were below the 0.90, thus, it concluded that all indicators have adequate discrimination to measure their own respective constructs (Table 4). The outer model analysis had passed reliability and validity testing sequentially. Therefore, it can be concluded that all indicators in this research model were reliable and valid to measure their respective constructs specifically. The inner model analysis was done to evaluate the quality of the proposed model. The variance inflation factor (VIF) was conducted to check multicollinearity issues. The findings showed all the constructs had an inner VIF value below 3 as suggested thus, it could be said that there was no multicollinearity issue in this model (Table 5) (Sarstedt et al., 2021).
The study performed coefficient determinant (R2) to measure prediction accuracy and construct cross-redundancy validation (Q2) to measure the predictive relevance of the model. The R-Square ranges from 0 to 1, with higher values indicating greater explanatory power. R-Square values as low as 0.10 are considered satisfactory. R-squared values between 0.2 -0.5 are categorized as weak, 0.5 -0.75 are categorized as moderate and above 0.75 are categorized as substantial or strong. However, if the R-squared value is above 0.9, it can be considered as overfitting data (Sarstedt et al., 2021). The result found in Table 6 above, found that the job satisfaction had R2 = 0.341, greater than 0.2, and was categorized as small predictive accuracy thus, it could be said the respective model has small capability to predict job satisfaction.
The burnout had R2 = 0.224, indicating that burnout had small predictive accuracy. The next step of the analysis of this structural model is the Q-Squared (Q2) test. The purpose of this test is to determine the predictive relevance of latent variables in the research model. Values of 0.02, 0.15, and 0.35 indicate that the exogenous constructs each have small, medium, and large predictive relevance for certain endogenous constructs (Sarstedt et al., 2021). A blindfolding procedure was conducted to assess the prediction of the proposed model, as seen in Table  6. All the Q2 were found > 0.15 whereas the job satisfaction showed Q2 = 0.192 and burnout had Q2 = 0.137 were categorized as medium (>0.15). Therefore, it could be said that the model approach had sufficient predictive relevance.
This study also tested the f-Square which was used to see the effect size or to see the magnitude of the effect of removing a predictor construct on the R-Square value of the target construct. The f-Square value above 0.02 has a small effect on a latent variable. The f-Square value above 0.15 has a medium effect, while the f-Square value above 0.35 has a large effect (large effect) from a latent variable (Sarstedt et al., 2021). The table 7 above shows the f-Square value from exogenous variables to endogenous variables. The variables working condition, work autonomy, and work-family conflict have a small effect on job satisfaction because they have an fsquare value above 0.02. The job satisfaction variable on burnout has a medium effect because it has an fsquare value of more than 0.15. Hypothesis testing by bootstrapping procedure was conducted to determine the effect of the variables in the model and confirm the hypothesis in this study was supported. The cut-off value criteria to determine hypothesis were significant or not significant by using T-statistic > 1.645 (one-tailed with alpha 0.05) (Hair et al., 2018(Hair et al., , 2019Sarstedt et al., 2021). The results were shown in Table 8 below. In Table 8, there were four hypothesis that were supported with a value of the T-statistics > 1.645 (one-tailed test with 0.05).
First, the working condition of the job satisfaction variable with a path coefficient value of 0.325 which was positive and the Tstatistic value of 4.980, so it can be concluded that if the perception of hospitalization nurses on working conditions was increased, the job satisfaction will be increase. Research conducted on nurses in Oman in 2019 also showed that working conditions had a positive correlation with nurse job satisfaction with a p-value of <0.001. Working conditions had an influence on nurse job satisfaction. Increasing nurse participation in hospital affairs and providing adequate staff were the main means to building healthy working conditions, which was a profitable and cost-effective strategy to increase nurse job satisfaction (Albashayreh et al., 2019). This is in line with previous research, an increase in job satisfaction has been associated with an increase in working conditions during participatory interventions carried out on professional health workers, especially a nurse. Participation in improving working conditions was in the form of motivation of health workers for increasing work control, awareness of work organization, involvement in workplace problems, and understanding of work processes (Tecco et al., 2020). The results of the research hypothesis test on the work autonomy variable on the job satisfaction variable with a path coefficient value of 0.297 which is positive and the T-statistic value is 4,322. This implies that if the perception of hospitalization nurses on work autonomy was increased, the job satisfaction they will be increase. Research conducted in Korea in 2018 showed that work autonomy was positively correlated with job satisfaction (r = 0.457, p < 0.001). Nurses' higher work autonomy has been shown to increase job satisfaction because it promotes organizational coherence by triggering motivation to provide the best possible care. This can be supported by self-determination theory (SDT), which identifies three basic psychological needs, namely autonomy, competence, and relatedness as determinants of satisfaction. According to SDT's theory of human resource development, autonomous workers highly value doing a good job when it is the job they want to do; they can be highly motivated when pursuing goals and assessing what is personally meaningful in their judgment, rather than extrinsic factors such as salary. Recently, SDT theory has been applied to understand work autonomy and job satisfaction among health professionals (Kim et al., 2022) The results of the research hypothesis test on the work-family conflict variable on the job satisfaction variable with a path coefficient value of 0.260 which is negative and the T-statistic value of 3,566. The results of the hypothesis mean that if the perception of hospitalization nurses on workfamily conflict was low, the job satisfaction will be increase. Previous studies have assessed the relationship between work-family conflict and job satisfaction. Two studies identified a statistically significant effect that female health workers reported significantly higher levels of work-family conflict than male health workers. In addition, more female health workers experienced frequent or very frequent work-family conflicts (56% and 41%). Significantly, the women who reported this had a low level of job satisfaction. Female health workers (13%) experience higher levels of job dissatisfaction compared to men (6%). The results of linear regression analysis also showed that work-family conflict predicts job dissatisfaction among women, as well as all health workers. It can be concluded that work-family conflict had a negative effect on job satisfaction (Domagała et al., 2018). Work-family conflict is defined as employees who do not have a balance between work, family, and free time (Tran, 2022). Work-family conflict occurs when an individual was unable to balance work with family demands and will create work that interferes with family life. Nurses were vulnerable to work-life imbalances for many reasons, such as stressful job demands, long and irregular working hours. Work-family conflict was associated with many negative outcomes between professionals and personnel, therefore work-family conflict was significantly related to job satisfaction (Rhéaume, 2022).

JENIUS
The results of the research hypothesis test on the job satisfaction variable on the burnout variable with a path coefficient value of 0.474 which is negative and the T-statistic value of 7.111. The results of the hypothesis mean that if the perception of hospitalization nurses on job satisfaction was increased, the burnout will be decrease. Clark and Lake (2020) studied the relationship between burnout and job dissatisfaction among nurses working in a midwifery clinic. It became clear that 25% of nurses showed symptoms of fatigue and 20% of nurses reported experiencing burnout.
In addition, job dissatisfaction and burnout have a positive relationship (Clark & Lake, 2020). In his research, Sarabi et al. (2020) evaluated the relationship between burnout syndrome, job satisfaction, and other related matters. The data show that burnout and job satisfaction levels are moderate in this rural area of Iran. In addition, the results of the study show that burnout has a statistically significant effect on job satisfaction, and therefore, the higher the level of job satisfaction, the lower the level of burnout (Sarabi et al., 2020). In addition, mediation analysis was also carried out to determine the mediation significance through the specific indirect effects. Based on the result of mediation analysis, all of the mediators tested had a respective Tstatistic above 1,645 thresholds.
The strongest influence path from the independent variable to the dependent variable through the mediating variable is working condition through job satisfaction with a coefficient value of -0.154 and T-statistics 3.7743 (T-statistic was found > 1.645). This shows the important role of job satisfaction as a mediating variable, so that in future research models it can still be used.
Important Performance Map Analysis showed that the target construct of the research model, namely job satisfaction, in the upper right quadrant there are variables working conditions and work autonomy. Thus, it can be said that the variables of working condition and work autonomy have been considered important by respondents and have performed well in the eyes of potential hospitalization nurses. Therefore, it can be suggested for the human resources team to always pay attention to and maintain these two things. The graph above also shows that the work-family conflict variable in the target construct of the research model, namely job satisfaction, is in the lower left quadrant. Thus, it can be said that the work-family conflict variable is considered less important and has poor performance in the eyes of potential hospitalization nurses. Therefore, it can be recommended for the human resources team to always pay attention and reduce problems that may occur related to work-family conflicts.

CONCLUSSION
In conclusion, this study demonstrated that three independent variables, which were working conditions and work autonomy as antecedents of job satisfaction had proven to have a significant and positive influence, and work family conflict had proven to have a significant and negative influence on job satisfaction. Figure 1. demonstrated that three independent variables Furthermore, the results showed that job satisfaction had negative effects on burnout. The result model (empirical model) of this study is known to have substantial small predictive accuracy (R2) and medium predictive relevance (Q2) on burnout as the predicted dependent variable. In this study there is only one path of one variable that predicts burnout, and in this path the f2 value of job satisfaction is classified as a medium size effect (f2 > 0.15). Thus, job satisfaction is considered adequate to predict burnout in hospitalization nurses. The new contribution of this study is to propose a previously modified model, where the dependent variable is burnout, the mediating variable is job satisfaction, the independent variable is working conditions, work autonomy, and work-family conflict. In this model, for the job satisfaction construct, small predictive accuracy (R2) and medium predictive relevance (Q2) are obtained. Therefore, it can be concluded that the antecedents of job satisfaction in this research model can adequately predict job satisfaction variables. Thus, this research model can be suggested to be replicated and tested further in other nurse populations at other hospitals that are potential for future research. This research has other limitations. the research model was only tested empirically in one hospital with a limited number of samples. Therefore, there are limitations in generalizing the findings of this study. It is recommended in future studies to include more hospitals, as well as a larger sample size so that the results of the analysis can be generalized more broadly.