Factors Associated with Turnover Intention Among Hospital Nurses: A Systematic Review and Meta-Analysis
Article information
Abstract
Purpose
The aim of the study was to conduct a systematic review and meta-analysis of nurses’ turnover intention in existing studies.
Methods
A literature search was conducted in two rounds. The first round was performed on December 16, 2018. The search period was from January 1, 2008, to December 16, 2018, and involved seven databases: MEDLINE and EMBASE, which are international academic databases, and KoreaMed, KERIS, KISS, KISTI, and KMbase, which are domestic academic databases. The second search round was performed on July 2, 2022. The search period was from 2019 to 2022.
Results
Fifty-three variables associated with turnover intention were identified in articles in the first round, with significantly associated work- and organization-related parameters, including burnout, emotional labor, job stress, role conflict, job satisfaction, and organizational commitment. Thirty-eight variables were identified in the second round, with mental and psychological parameters, including mental health and psychological capital, having a more significant influence.
Conclusion
The findings of this study can be used to prevent the turnover of hospital nurses and develop strategies for retaining excellent and professional hospital nurses.
INTRODUCTION
Hospital nurse managers have long aimed to retain and manage skilled nurses by lowering turnover [1]. Cutting-edge medical systems are increasingly replacing nursing staff in hospitals; however, the loss of skilled nursing professionals is alarming [2]. Retaining and managing personnel is essential because retaining excellent staff for prolonged periods and assisting them to utilize their competencies developed through experience and expertise contributes to attaining organizational goals [3]. Furthermore, experienced nurses with strong organizational commitment and job satisfaction serve as role models for newly hired personnel, assisting them in overcoming the shock and burden of being in an unfamiliar work environment [4]. In addition, they serve as valuable mentors, assisting new hires in adjusting to the organization and staying [5]. Finally, experienced nurses motivate other members of the organization to commit to and trust the organization, facilitating a virtuous cycle in nursing human resource management [6].
The United States has pursued various systemic efforts to recruit proficient nurses and prevent their loss, one of which is accreditation programs like magnet hospitals, which promote how hospital organizations provide good work environments and benefits for nurses [7]. The magnet hospital system offers several benefits, which can broadly be divided into nursing-related outcomes, patient outcomes, and organizational outcomes [8]. Nursing-related outcomes include reduced job dissatisfaction and burnout [9], lower nursing turnover [7] and consequently reduced costs [10]. In other words, nurses in magnet hospitals enjoy a better work environment, nurse-patient ratio, and quality of staff than nurses in non-magnet hospitals [11]. Finally, magnet hospital cultures prevent bullying and hostile behavior among professionals [12]. In Korea, the level of nursing staff is a critical evaluation criterion for hospital accreditation [13].
Nevertheless, nurse turnover remains high. The increased workload caused by the coronavirus disease (COVID -19) pandemic has significantly increased the global turnover rate. In the United States, the nurse turnover rate was 17.2% in 2018 but rose to 22.7% in December 2022 [14]. Prior to the pandemic, turnover intention was primarily linked to sociodemographic and work-related factors. Van der Heijden, Brown Mahoney, and Xu (2019) reported that high burnout leads to high turnover intention [15], while Sharififard et al. (2019) discovered that overtime is a predictor of turnover intention [16]. Vevoda, Vevodova, Bubenikova, Kisvetrova, and Ivanova (2016) pinpointed wage, access to modern technology and equipment, and social benefits for employees as the factors associated with turnover intention. In contrast, after the COVID-19 outbreak, Khattak et al. (2021) identified fear of COVID-19, secondary trauma, showing that COVID-19 has had adverse impacts on the retention and management of hospital nurses [17].
These findings emphasize the importance of nursing organizations taking proactive measures to minimize the loss of proficient nursing human resources. Therefore, the first step to devising nursing management strategies is identifying the key factors influencing nurses’ turnover. Additionally, determining which of the factors mentioned in previous studies contribute to lower and higher nurse turnover is crucial.
1. Aim
This study aimed to conduct a systematic review and meta-analysis of existing studies on nurses’ turnover to identify factors reinforcing increased turnover, as well as management factors that contribute to a decreased turnover, based on statistical evidence.
The specific goals are:
1) To systematically review studies on the turnover intention of hospital nurses
2) To examine the meta-correlation coefficients of turnover intention and associated factors
METHODS
1. Design
This is a secondary data analysis study that identified factors that significantly correlate with the turnover intention of hospital nurses using a literature review and meta-analysis.
2. Search Methods
The literature search was conducted in two rounds. The first round of search was performed on December 16, 2018. The search period was from January 1, 2008 to December 16, 2018, and involved seven databases: MEDLINE and EMBASE, which are international academic databases, and KoreaMed, KERIS, KISS, KISTI, and KMbase, which are domestic academic databases. In total, 1,345 studies were found. Following the screening test and quality appraisal process for the abstract and full text by two independent researchers, 129 studies from the first round were identified for inclusion in the systematic review and meta-analysis. During the first search literature analysis, COVID-19 occurred, and it was thought that the pandemic situation would have the most sensitive impact on hospital nurses, so a second literature search was planned to compare with the post-COVID-19 situation.
The second round of search was performed on July 2, 2022 using the same search strategy and databases as in the first round. The search period was from 2019 to 2022. After excluding 387 duplicate publications, 835 articles were extracted. Subsequently, the primary abstract and secondary full texts were screened using literature selection and exclusion criteria. Finally, 112 papers were identified for inclusion in the second round (Figure 1).
Preferred reporting items for systematic reviews and meta-analyses of this study.
3. Inclusion and Exclusion Criteria
The literature for this study was extracted based on the PICOTS-SD strategy, which includes seven factors: participant, intervention, comparison, outcome, time, setting, and study design. Selection criteria included literature with
1) general hospital nurses as study participants (P),
2) intervention as the analysis of organizational behavioral factors (I),
3) outcome as turnover intention (O),
4) the aim to explore the research on turnover intention measurement comprehensively, so comparison (C) was unlimited
5) setting (S) as a general hospital,
6) study design (SD) as a correlation study, and
7) time (T) as studies published after 2008.
Inclusion criteria involved studies:
1) involving hospital nurses,
2) that measured turnover intentions and reported correlations between turnover intentions and combinations of personality and job-related characteristics, and
3) published in Korean or English.
Exclusion criteria included studies:
1) where nurses did not participate,
2) published in languages other than Korean and English,
3) that did not report the entire study results,
4) that provided only abstracts, such as conference materials, and
5) that did not present accurate statistical results, such as correlation coefficients.
3. Quality appraisal
Cummings and Estabrooks (2003) evaluated the quality of the literature to be analyzed using the quality assessment and validity tool for correlation study [18]. The quality evaluation tool consists of the following 11 items:
1) whether a theoretical framework was used,
2) whether randomization was used,
3) whether a prospective design was used,
4) whether an appropriate sample size was used,
5) whether data was collected from multiple centers,
6) whether anonymity was ensured,
7) whether a response rate over 60% was achieved,
8) whether a tool with established reliability and validity was used,
9) whether a tool with an internal consistency of 0.7 or more was used,
10) whether the statistical analysis was suitable for the study, and
11) whether statistical management for outliers was presented.
The quality of the literature was evaluated independently by two evaluators who were not members of the research team. Then, consensus was reached through discussion, including the research team, on items with differences of opinion.
4. Synthesis
1) General characteristics of selected studies
The general characteristics of articles were analyzed in four categories: publication year, publication country, sample size, and quality evaluation score.
2) Summary of descriptive statistics for variables related to turnover intention
Using the correlation coefficient r and p-value, frequency, and percentage, this analysis identified variables that significantly correlated with the turnover intention of general hospital nurses.
3) Effect size calculation for meta-analysis and homogeneity testing
Meta-correlation analysis was performed using Comprehensive Meta-Analysis 3.0 software. Standardized Zr was analyzed based on the standardized Fisher's Z equation, and 95% confidence intervals were presented. The fixed-effects model was used in cases of high homogeneity, whereas the random-effects model was used in cases of high heterogeneity. Effect sizes were interpreted based on the following criteria: r≤0.1 (small effect size), 0.3<r≤0.5 (middle effect size), and 0.5<r (high effect size). This is in accordance with Cohen's criteria (1988). Homogeneity was analyzed using Q and I2 values following the χ2 distribution.
4) Publication bias in studies included in the meta-analysis
Publication bias was analyzed using a funnel plot and the trim and fill method.
4. Ethical Considerations
This study was exempted from review by the institutional review board of our university because of the secondary data analysis of literature [details blinded for peer review].
RESULTS
1. General Characteristics of the Studies
Of the 241 articles included, 129 were published between 2008 and 2018, and 112 between 2019 and 2022. Most of the studies were published in Asia (n=226), including South Korea, China, and East Asia (12 nations). The most common sample size was 100~300 (81.82%), and the quality rating was 10 (75.97%) (Table 1).
General Characteristics of the Studies
2. Meta-analysis of Turnover and associated Variables and Homogeneity Test
Prior to 2018, 53 variables were associated with the turnover of hospital nurses. Twenty-one had a moderate or higher correlation of .3 or higher. Six of these variables had a positive association, with burnout (r=.52, z=10.98, p<.001) having the strongest association, followed by emotional labor and job stress. Fifteen variables had a negative association, with organizational commitment (r=-.54, z=-11.59, p<.001) having the strongest association, followed by job involvement and job satisfaction.
Since 2019, 38 variables have been associated with hospital nurse turnover. Twenty had a moderate or higher association of .3 or higher. Eight of these variables had a positive association, with role conflict (r=.52, Z=14.28, p<.001) having the strongest correlation, followed by emotional labor and traumatic shock. Twelve variables had a negative correlation, with organizational commitment (r=-.58, z=-14.55, p<.001) and job embeddedness (r=-.58, z=-14.75, p<.001) having the strongest correlation, followed by work engagement and work-life balance.
The random-effects model was used because the I2 index was .00~.99.61, indicating heterogeneity (Tables 2, 3).
Meta-analysis of Variables Associated with Turnover Intention (2008~2018)
Meta-analysis of Variables Associated with Turnover Intention (2019~2022)
3. Publication Bias of Selected Studies
The funnel plot demonstrated no publication bias because it appeared symmetrically around the integrated estimate. Furthermore, there was no difference before and after correction in the Trim and Fill testing, indicating no publication bias (Figure 2).
Funnel plot of this study.
DISCUSSION
This study systematically reviewed existing studies on hospital nurses' turnover intention and conducted a meta-analysis. As a result of the study, 53 variables were identified in the research literature before the outbreak of COVID-19. In particular, significant correlations were shown between psychological variables such as burnout, emotional labor, job stress, and role conflict, and work and organization-related variables such as job satisfaction and organizational commitment. In research literature published after COVID-19, 38 variables with a more significant mental and psychological impact, such as mental health, were identified.
Turnover of hospital nurses is the most significant phenomenon in retaining and managing proficient nursing professionals. Moreover, one of the most critical policies of nursing organizations is to reduce nurse turnover. One hundred and twenty-nine studies between 2008 and 2018 and 112 between 2019 and 2022 investigating predictors of turnover intention showed that hospital nurses’ turnover is a crucial factor for retaining nurses in nursing organizational management. Moreover, more studies on turnover intention were published after 2019 than during the 10 years between 2008 and 2018, suggesting that nurse turnover has emerged as a critical issue due to the COVID-19 pandemic.
Among studies published between 2008 and 2018, 53 variables were associated with turnover intention. Of the 21 variables with a moderate to high negative correlation with turnover intention, job satisfaction, and organizational commitment had the strongest association. In contrast, burnout, emotional labor, and job stress were positively correlated with turnover intention. These results suggest that hospital nurses’ job and organizational characteristics influence nurses’ decisions to stay or leave their jobs. This is consistent with Kim and Kim’s (2014) structural modeling study, which discovered that job satisfaction, organizational commitment, stress, and burnout had significant direct and indirect effects [19]. These results highlight the need for organizations to implement organizational support measures and measures that enhance nursing work efficiency to lower hospital nurse turnover.
Among studies published between 2019 and 2022, 38 variables were associated with turnover intention, with 20 having a moderate or high correlation. Psychological capital was the most strongly correlated factor. However, other variables, including quality of nursing work life, job embeddedness, organizational commitment, role conflict, job stress, mental health, traumatic shock, and work-life balance, also had strong correlations. Burnout, job stress, role conflict, and organizational commitment were associated with turnover intention in both periods; however, the degree of association varied. This suggests that factors directly linked to work, such as burnout, stress, and role conflict, remain strong predictors of nurses’ turnover intention.
In contrast, the influences of psychological and mental factors, including psychological capital, mental health, and traumatic shock, were more strongly linked to turnover intention in studies published in 2019. Traumatic shock refers to the phobia of COVID-19 and COVID-19-related traumatic events. Furthermore, these results indicate that changes in the healthcare and work environments, increased workload, and psychological burnout caused by COVID-19 intensified nurses’ intent to leave their jobs. Amid a battle against a virus in an unprecedented pandemic, hospital nurses perceived COVID-19 as a threat [2] and suffered from psychological stress, such as fear and secondary trauma, leading to their decisions to leave the clinical setting.
A study on healthcare workers during the COVID-19 pandemic [20] reported that 9.6~51.0% of healthcare workers had post-traumatic stress disorder, and 20–75% experienced psychological and emotional repercussions, such as anxiety and depression. This shows that various forms of psychological burden and burnout have emerged as significant problems. These factors are expected to continue to influence nurses’ turnover intentions with the extension of the COVID-19 pandemic. Studies report that clear communication, managerial support, alleviation of work stress through precise work distribution, periodic work training, adequate professional staffing, and counseling with another professional are some protective factors against nurses’ psychological burden and stress. Therefore, these factors should be considered when devising strategies to lower nurse turnover. The advances in information and communication technology and information technology have ameliorated work overload and efficiency problems. However, role conflicts and emotional and psychological problems persist, and their influence is growing, necessitating strategies to address these problems. Organizations must improve their systems, train employees to facilitate clear communication and implement practical training for infectious diseases and response. Such measures would increase job embeddedness and help nurses stay in the clinical setting.
Extensive research has been conducted on hospital nurses’ turnover, primarily concerning nurses’ sociodemographic and work-related factors; however, researchers have recently begun to focus on nurses’ psychosocial problems. Comprehensive approaches that address this complex relationship, rather than dichotomous strategies that focus on personal or organizational dimensions, are required because nurses’ psychosocial problems are closely linked to personal and organizational factors. Various policies have been implemented at the organizational level, such as presenting evidence for adequate nursing staffing through healthcare institution accreditation and evaluation, various training and education programs, and improving work efficiency using information and communication technology and information technology. However, the findings of our study amid a global emergency due to COVID-19 highlight the urgency of establishing measures to minimize the loss of competent nursing human resources in organizations. Furthermore, the findings accentuate the importance of support measures at the organizational level, work efficiency, securing safe nursing environments during an infectious disease pandemic, ensuring adequate hospital facilities and equipment, and implementing clear standard protocols in relevant governmental agencies to protect nurses from psychological conflict and stress.
CONCLUSION
Nurse turnover is the most significant phenomenon in retaining and managing skilled nursing professionals. Considering the pandemic situation caused by COVID-19, the analysis was divided into two periods based on 2019. Our analysis revealed that organizational commitment, job stress, burnout, and role conflict were strongly correlated with turnover intention in both periods. Psychological factors such as psychological capital and mental health were more strongly correlated with turnover intention in studies published in 2019 and after. The results suggest that psychological stress related to the fear and secondary trauma from COVID-19 has increased hospital nurses’ intent to leave the clinical setting amid the prolonged pandemic. Strategies reflecting these findings must be developed and implemented through relevant policies at the hospital’s organizational and governmental levels to lower hospital nurses’ turnover intention and retain proficient and specialized nursing staff. Thus, this study’s findings would be useful as evidence for developing such policies.
Notes
The authors declared no conflicts of interest.
AUTHOR CONTRIBUTIONS
Conceptualization and Methodology: Lim, JY & Kim, GM
Data curation: Lim, JY
Formal Analysis: Lim, JY & Kim, GM
Funding acquisition: Lim, JY
Project administration: Lim, JY
Supervision: Kim, EJ
Validation: Lim, JY, Kim, EJ & Kim, GM
Original draft: Lim, JY, Kim, EJ & Kim, GM
Writing - review & editing: Lim, JY, Kim, EJ & Kim, GM