Research Article

Journal of The Korean Society of Living Environmental System. 31 October 2026. 576-586
https://doi.org/10.21086/ksles.2026.10.33.5.576

ABSTRACT


MAIN

  • 1. Introduction

  • 2. Method

  •   2.1. Study Design

  •   2.2. Participants and Data Collection

  •   2.3. Measures

  •   2.4. Ethical Considerations

  •   2.5. Data Analysis

  • 3. Results

  •   3.1. General Characteristics of the Participants and Differences in Turnover Intention

  •   3.2. Mean Scores and Correlation Coefficients among the Study Variables

  •   3.3. Hierarchical Regression Analysis for Turnover Intention

  • 4. Discussion

  • 5. Conclusion

1. Introduction

The global nursing shortage remains a persistent crisis, undermining health system capacity and patient safety worldwide (World Health Organization [WHO], 2025). A central driver of this shortage is turnover intention, which has been consistently linked to increased workloads, reduced staffing, poorer nurse outcomes, and substantial replacement costs (Aiken et al., 2023; Bae, 2022). The coronavirus disease 2019 (COVID-19) pandemic further intensified these pressures, with organizational support frequently failing to keep pace with rising workloads and psychological stress (Organisation for Economic Co-operation and Development [OECD], 2025; WHO, 2025), suggesting that the factors underlying turnover intention may have shifted in the post-pandemic period.

In South Korea, these challenges are particularly acute. The number of practicing clinical nurses per 1,000 population remains well below the OECD average, despite a more than twofold increase in nursing school admissions since 2008 (Ministry of Health and Welfare [MOHW], 2025a; OECD, 2025). Turnover among newly licensed nurses has recently shown signs of improvement: the proportion of new nurses who resigned within their first year peaked at 57.4% in 2022 and declined to 45.9% in 2024 (Hospital Nurses Association, 2026). Government pilot programs for shift-work improvement and dedicated nurse educators, launched in 2022, have accordingly monitored not only new-nurse turnover but also the retention rate of experienced nurses with three or more years of tenure, reporting a decrease in new-nurse turnover from 15.7% to 10.6% and an increase in experienced-nurse retention from 53.0% to 56.5% (MOHW, 2025b). This shift in policy focus, from replacing departing new graduates toward retaining experienced nurses, underscores the need to understand turnover intention across the clinical workforce rather than among new graduates alone.

The Korean nursing environment has also undergone substantial structural change. The Nursing Act, enacted in 2024 and implemented in June 2025, established an independent legal basis for nursing practice, including advanced practice and the medical support tasks that physician assistant nurses had long performed without clear legal standing (Woo & Woo, 2026). Concurrently, the mass walkout of resident physicians that began in February 2024 reduced service capacity at tertiary hospitals, and part of the resulting care gap was filled by nurses and advanced practice nurses (Ha & Kim, 2026). These developments have directly affected staffing, role boundaries, and nurse–physician working relationships in tertiary hospitals, the setting of the present study, making it timely to re-examine the factors associated with turnover intention in this context.

Research has consistently identified three categories of determinants: the nursing work environment, job satisfaction, and individual psychological resources such as resilience. The Job Demands-Resources (JD-R) Model (Bakker & Demerouti, 2007) frames a supportive work environment and job satisfaction as job resources that buffer nursing’s demands. Conservation of Resources (COR) Theory (Hobfoll, 1989) further conceptualizes personal and contextual resources as interacting rather than independent: resilience, as a personal resource, may help nurses cope with resource loss, but its protective effect is expected to depend on the adequacy of surrounding organizational resources rather than operate in isolation. Together, these frameworks motivate the hypothesis that organizational factors may exert stronger direct effects on turnover intention than individual psychological resources such as resilience.

Empirically, the work environment, particularly staffing adequacy, foundations for quality care, and nurse-physician collaboration, has emerged as a primary, consistent predictor across settings, including in Korean studies using the Practice Environment Scale of the Nursing Work Index (PES-NWI) (Aiken et al., 2023; Lee & Yu, 2025). Korean evidence also indicates that nurse participation in hospital affairs and collegial nurse–physician relations are directly associated with nurses’ intent to stay (Kim & Lee, 2023), highlighting the relevance of nurses’ involvement in hospital governance and interprofessional relationships alongside staffing. Job satisfaction is similarly one of the most proximal determinants of nurses’ retention decisions (Enea et al., 2024; Lim et al., 2024; Pressley & Garside, 2023) and moderates the relationship between the nursing work environment and turnover intention (Lee & Yu, 2025). Resilience, by contrast, shows inconsistent effects on turnover intention: some studies report inverse associations (Gao et al., 2026), while others find no direct effect once job satisfaction and social support are accounted for (Zhao et al., 2021), and Korean meta-analytic evidence suggests organizational and job-related variables remain the stronger correlates overall (Lim et al., 2024).

Despite this breadth of evidence, whether resilience retains independent predictive value after accounting for the nursing work environment and job satisfaction has not been directly tested, and studies examining all three factors together using recent Korean data that reflect these post-pandemic and policy changes remain limited. This study aimed to identify factors associated with turnover intention among tertiary hospital nurses in South Korea. Specifically, it examined whether the nursing work environment and job satisfaction are associated with turnover intention after controlling for general and clinical characteristics, and whether resilience is independently associated with turnover intention after the nursing work environment and job satisfaction are accounted for.

2. Method

2.1. Study Design

This study employed a descriptive cross-sectional design with data collected via an online self-administered questionnaire.

2.2. Participants and Data Collection

The study population comprised registered nurses employed at tertiary hospitals in Seoul and Gyeonggi-do, South Korea, with a minimum of three months of clinical experience. The required sample size was calculated a priori using the G*Power 3.1.9.7 program for hierarchical multiple regression, based on a significance level of .05, a medium effect size of .15, a power of .80, and 14 predictors (7 general and clinical characteristics variables and 7 variables of interest: the 5 nursing work environment subscales, resilience, and job satisfaction), resulting in a minimum of 135 participants. Considering a potential attrition rate of 10%, 148 questionnaires were targeted for distribution. Of 146 responses received, 12 were excluded for incomplete data, yielding a final analytic sample of 134 participants.

Data were collected over a two-week period in December 2024 using convenience sampling. The survey was administered via Google Forms, with the questionnaire link initially distributed to eligible nurses who then forwarded it to colleagues within their professional networks, including ward-based communication channels and peer contacts. This approach was adopted to facilitate access to participants across a range of hospital settings. Prior to participation, prospective respondents received information regarding the study purpose, voluntary nature of participation, and confidentiality assurances. Eligibility criteria were reviewed at the outset, and progression to the questionnaire required completion of the electronic informed consent process.

2.3. Measures

Nursing work environment, defined as the organizational characteristics of the workplace that facilitate or constrain professional nursing practice, was assessed using the Korean version of the PES-NWI, originally developed by Lake (2002) and translated and validated for Korean nurses by Cho et al. (2011). The instrument consists of 29 items rated on a 4-point Likert scale (1 = not at all to 4 = very much so) across five subscales: nurse participation in hospital affairs (9 items; range 9–36), nursing foundations for quality of care (9 items; range 9–36), nurse manager ability, leadership, and support of nurses (4 items; range 4–16), staffing and resource adequacy (4 items; range 4–16), and collegial nurse–physician relations (3 items; range 3–12). Subscale scores were calculated by summing the item scores, yielding a possible total score of 29 to 116, and the five subscale scores were used in the analyses; higher scores indicate a more favorable nursing work environment. In the Korean validation study, Cronbach’s α was .93 for the total scale and ranged from .80 to .93 across the subscales (Cho et al., 2011); in the present study, it was .91 for the total scale and ranged from .70 to .83 across the subscales.

Job satisfaction, defined as the degree to which clinical nurses perceive positive feelings and fulfillment from their work and professional role, was assessed using the Job Satisfaction Scale for Clinical Nurses developed by Lee et al. (2018). The instrument consists of 33 items rated on a 5-point Likert scale (1 = not at all to 5 = very much so) covering six domains: recognition from the organization and professional achievement (9 items), personal maturation through the nursing profession (6 items), interpersonal interaction with respect and recognition (8 items), accomplishment of accountability as a nurse (4 items), display of professional competency (3 items), and stability and job worth (3 items). The total score was calculated by summing all item scores (range 33–165) and was used in the analyses; higher scores indicate greater job satisfaction. Cronbach’s α was .95 in the development study (Lee et al., 2018) and .92 in the present study.

Resilience, defined as a nurse’s capacity to adapt positively and recover from stress and adversity in the clinical setting, was measured using the Resilience Scale for Nurses developed by Park and Park (2016). The instrument consists of 30 items rated on a 5-point Likert scale (1 = not at all to 5 = very much so) covering five domains: dispositional pattern (5 items), relational pattern (4 items), situational pattern (10 items), philosophical pattern (6 items), and professional pattern (5 items). The total score was calculated by summing all item scores (range 30–150) and was used in the analyses; higher scores indicate greater resilience. Cronbach’s α was .95 in the development study (Park & Park, 2016) and .93 in the present study.

Turnover intention, defined as a nurse’s conscious and deliberate willingness to leave their current workplace, was measured using the 6-item scale used by Kim (2007). Items are rated on a 5-point Likert scale (1 = not at all to 5 = very much so), and item 4, which reflects intention to stay, was reverse-coded. The total score was calculated by summing all item scores (range 6–30) and was used in the analyses; higher scores indicate stronger turnover intention. Cronbach’s α was .76 in the original study (Kim, 2007) and .83 in the present study.

2.4. Ethical Considerations

This study was approved by the affiliated university’s Institutional Review Board. Electronic informed consent was obtained, participation was voluntary, and no personally identifiable information was collected, in accordance with the Declaration of Helsinki.

2.5. Data Analysis

Data were analyzed using SPSS 29.0. Group differences in turnover intention were examined using t-tests and ANOVA with Scheffé’s post hoc test. Pearson’s correlations assessed bivariate associations. Hierarchical multiple regression was conducted with four sequential models: general and clinical characteristics (Model 1), nursing work environment (Model 2), job satisfaction (Model 3), and resilience (Model 4). Regression assumptions were evaluated using the Durbin-Watson statistic and variance inflation factors.

3. Results

3.1. General Characteristics of the Participants and Differences in Turnover Intention

Participant ages ranged from 23 to 50 years (M = 29.10, SD = 5.38). The majority were female (92.5%), single (84.3%), without religious affiliation (67.2%), and held a bachelor’s degree (92.5%) (Table 1). Mean clinical experience was 5.39 years (SD = 5.17), with 43.3% having less than three years; most worked in general ward settings (70.9%).

Turnover intention differed significantly by age (F = 4.338, p = .015), sex (t = 2.130, p = .035), and educational level (t = 2.675, p = .008). Scheffé’s post hoc test indicated that nurses aged 23–39 years reported significantly higher turnover intention than those aged 40 or older. In addition, female nurses reported higher turnover intention than male nurses, and bachelor’s-degree nurses reported greater turnover intention than those with a master’s degree. No significant differences were found for the remaining variables.

Table 1

General Characteristics of the Participants and Differences in Turnover Intention
(N=134)

Variables N (%) Turnover intention
M ± SD t/F (p)
Age (Range: 23–50)
  23–29a 94 (70.1) 19.32 ± 6.35 4.338
  30–39b 29 (21.6) 19.48 ± 4.14 (.015)
  40≤c 11 (8.2) 14.73 ± 3.74 c < a, b*
Sex
  F 124 (92.5) 19.24 ± 4.90 2.130
  M 10 (7.5) 15.70 ± 6.88 (.035)
Marital status
  Single 113 (85.0) 19.15 ± 5.26 0.610
  Married 20 (15.0) 18.55 ± 3.80 (.546)
Religious affiliation
  No 90 (67.2) 18.77 ± 5.05 0.680
  Yes 44 (32.8) 19.41 ± 5.32 (.498)
Educational level
  Bachelor’s degree 124 (92.5) 19.31 ± 5.08 2.675
  Master’s degree 10 (7.5) 14.90 ± 3.99 (.008)
Clinical career (yr)
  ~ ≤3 58 (43.3) 19.52 ± 5.01 0.815
  3< ~ ≤5 29 (21.6) 19.21 ± 6.06 (.488)
  5< ~ ≤10 28 (20.9) 18.64 ± 4.58
  10< ~ 19 (14.2) 17.47 ± 4.77
Type of working unit
  General unit 95 (70.9) 19.27 ± 4.99 1.044
Specialized unit† 39 (29.1) 18.26 ± 5.45 (.298)

*Scheffé post hoc test; Percentages for marital status are based on n = 133, as one participant did not identify with either category.

†Intensive care unit, Emergency room, Operating room.

3.2. Mean Scores and Correlation Coefficients among the Study Variables

The mean turnover intention score was 18.98 (SD = 5.13; possible range 6–30). The total scores and mean item scores of all study variables are presented in Table 2. All study variables were significantly and negatively correlated with turnover intention. Among the nursing work environment subscales, staffing and resource adequacy showed the strongest negative correlation (r = –.414, p < .001), followed by nursing foundations for quality of care (r = –.335), collegial nurse–physician relations (r = –.299), nurse manager ability, leadership, and support of nurses (r = –.289), and nurse participation in hospital affairs (r = –.246, p = .004; all others p < .001). Job satisfaction (r = –.356, p < .001) and resilience (r = –.239, p = .005) were also negatively associated with turnover intention.

Table 2

Mean Scores and Correlation Coefficients among the Study Variables
(N=134)

Variables Total score Mean item score Turnover intention
M ± SD M ± SD r (p)
Nursing work environment - - -
  Nurse participation in hospital affairs 23.07 ± 4.01 2.56 ± 0.45 –.246 (.004)
  Nursing foundations for quality of care 25.75 ± 3.41 2.86 ± 0.38 –.335 (< .001)
  Nurse manager ability, leadership, and support of nurses 10.99 ± 2.02 2.75 ± 0.50 –.289 (< .001)
  Staffing and resource adequacy 9.31 ± 2.13 2.33 ± 0.53 –.414 (< .001)
  Collegial nurse-physician relations 8.09 ± 1.64 2.70 ± 0.55 –.299 (< .001)
Job satisfaction 106.00 ± 13.92 3.21 ± 0.42 –.356 (< .001)
Resilience 115.77 ± 12.73 3.86 ± 0.42 –.239 (.005)
Turnover intention 18.98 ± 5.13 3.16 ± 0.85 -

3.3. Hierarchical Regression Analysis for Turnover Intention

Hierarchical multiple regression confirmed no autocorrelation (Durbin–Watson = 2.049) and acceptable multicollinearity (variance inflation factor [VIF] = 1.117–4.580; Table 3). Model 1 (general and clinical characteristics) was not significant (F = 1.212, p = .290, R² = .090, Adjusted R² = .016). Adding nursing work environment variables in Model 2 produced a significant model (F = 3.905, p < .001, R² = .334, ΔR² = .243, Adjusted R² = .248); three subscales emerged as significant predictors: nursing foundations for quality of care (β = –.285, p = .021), staffing and resource adequacy (β = –.286, p = .003), and collegial nurse–physician relations (β = –.209, p = .023). Age 40 or older was also significant (β = –.413, p = .012).

Table 3

Hierarchical Regression Analysis for Turnover Intention

Model 1 Model 2 Model 3 Model 4
B (SE) β p B (SE) β p B (SE) β p B (SE) β p
Constant 19.515 (0.784) < .001 38.004 (3.356) < .001 42.213 (3.848) < .001 41.158 (4.211) < .001
Age: 30–39 (ref. 23–29) 0.253 (1.488) .021 .865 0.168 (1.303) .014 .897 –0.094 (1.289) –.008 .942 –0.106 (1.293) –.009 .935
Age: 40≤ (ref. 23–29) –5.415 (3.452) –.284 .119 –7.887 (3.083) –.413 .012 –7.682 (3.038) –.402 .013 –7.658 (3.047) –.401 .013
Male (ref. Female) –3.000 (1.771) –.150 .093 –2.737 (1.584) –.137 .087 –2.374 (1.570) –.118 .133 –2.458 (1.580) –.123 .122
Married (ref. Single) 1.229 (1.554) .087 .431 1.154 (1.386) .082 .407 0.786 (1.376) .056 .569 0.780 (1.379) .055 .573
Religious affiliation: Yes (ref. No) 0.682 (0.981) .064 .488 0.394 (0.878) .037 .654 0.421 (0.865) .039 .487 0.364 (0.872) .034 .677
Master’s degree (ref. Bachelor’s degree) –1.329 (2.554) –.066 .604 0.867 (2.301) .043 .707 1.325 (2.276) .066 .562 1.354 (2.283) .068 .554
Clinical career: 3< ~ ≤5 (ref. ~ ≤3) –0.266 (1.145) –.022 .817 –0.791 (1.048) –.065 .452 –0.469 (1.044) –.038 .654 –0.528 (1.051) –.043 .616
Clinical career: 5< ~ ≤10 (ref. ~ ≤3) –0.925 (1.391) –.075 .507 –1.826 (1.270) –.148 .153 –1.486 (1.262) –.120 .241 –1.515 (1.266) –.123 .234
Clinical career: 10< ~ (ref. ~ ≤3) 1.045 (2.417) .071 .666 0.462 (2.140) .031 .829 0.764 (2.113) .052 .718 0.788 (2.119) .054 .711
Type of working unit: Specialized unit (ref. General unit) –0.381 (1.034) –.034 .713 –0.444 (0.916) –.040 .628 –0.654 (0.908) –.059 .473 –0.699 (0.913) –.063 .446
Nursing Work Environment
  Nurse participation in hospital affairs 0.105 (0.157) .083 .504 0.138 (0.156) .109 .378 0.145 (0.157) .115 .355
  Nursing foundations for quality of care –0.422 (0.180) –.285 .021 –0.404 (0.178) –.273 .025 –0.423 (0.181) –.286 .021
  Nurse manager ability, leadership, and support of nurses 0.180 (0.277) .071 .518 0.254 (0.275) .101 .358 0.276 (0.278) .109 .324
  Staffing and resource adequacy –0.678 (0.225) –.286 .003 –0.636 (0.223) –.268 .005 –0.646 (0.224) –.272 .005
  Collegial nurse–physician relations –0.643 (0.278) –.209 .023 –0.585 (0.276) –.190 .036 –0.626 (0.284) –.204 .030
Job satisfaction –0.068 (0.032) –.187 .035 –0.079 (0.036) –.217 .032
Resilience 0.024 (0.038) .061 .534
R² .090 .334 .359 .361
ΔR² .243 .025 .002
Adjusted R² .016 .248 .270 .267
F(p) 1.212 (.290) 3.905 (< .001) 4.059 (< .001) 3.823 (< .001)
Durbin-Watson 2.049
Tolerance .218 ~ .895
Variance inflation factor (VIF) 1.117 ~ 4.580

Model 3 added job satisfaction, which was a significant predictor (β = –.187, p = .035), yielding a significant model (F = 4.059, p < .001, R² = .359, ΔR² = .025, Adjusted R² = .270); nursing foundations for quality of care (β = –.273, p = .025), staffing and resource adequacy (β = –.268, p = .005), collegial nurse–physician relations (β = –.190, p = .036), and age 40 or older (β = –.402, p = .013) remained significant. Model 4 added resilience, which was not a significant predictor (β = .061, p = .534), yielding a final model (F = 3.823, p < .001, R² = .361, ΔR² = .002, Adjusted R² = .267) in which job satisfaction (β = –.217, p = .032), nursing foundations for quality of care (β = –.286, p = .021), staffing and resource adequacy (β = –.272, p = .005), collegial nurse–physician relations (β = –.204, p = .030), and age 40 or older (β = –.401, p = .013) remained significant.

4. Discussion

This study examined factors associated with turnover intention among South Korean hospital nurses, focusing on the nursing work environment, job satisfaction, and resilience. After controlling for general and clinical characteristics, three nursing work environment subscales and job satisfaction were independently associated with lower turnover intention, whereas resilience was not independently associated with turnover intention in the final model.

Job satisfaction emerged as a significant independent predictor, consistent with its established role as a proximal determinant of nurses’ retention decisions (Lim et al., 2024; Pressley & Garside, 2023). Lower satisfaction with personal development and professional recognition has been identified as a key predictor of turnover intention among Korean nurses specifically (Moon & Kang, 2025), which may explain why job satisfaction retained independent predictive value even after the nursing work environment was entered in Model 2. Similarly, in a recent study of Korean general hospital nurses, job satisfaction, rather than resilience, was a significant predictor of retention intention (Lee et al., 2024). A Korean study of Magnet hospitals has further reported that job satisfaction moderates the relationship between the nursing work environment and turnover intention (Lee & Yu, 2025); although the present study did not test this interaction, nursing work environment and job satisfaction each remained a significant independent predictor of turnover intention in the final model, suggesting that job satisfaction explains additional variance beyond the work environment. These results are consistent with the JD-R Model, which regards job resources as conditions that support nurses’ retention (Bakker & Demerouti, 2007).

Three nursing work environment subscales were significant predictors: nursing foundations for quality of care, staffing and resource adequacy, and collegial nurse-physician relations. These domains reflect the direct, structural conditions of daily clinical practice. This finding is consistent with evidence that inadequate staffing and unfavorable working conditions are primary drivers of burnout and turnover intention internationally (Aiken et al., 2023; Skela-Savic et al., 2025). Recent Korean policy data point in the same direction: in the government’s shift-work improvement pilot program, which funded additional nurses for participating wards, new-nurse turnover decreased and experienced-nurse retention increased (MOHW, 2025b). Even hospitals with Magnet designation, which formally recognizes hospitals with strong nursing practice environments, have reported nurse dissatisfaction with workload and staffing (Turnbach et al., 2024). This suggests that formal structural recognition alone does not guarantee that nurses experience their day-to-day working conditions as adequate. The significance of collegial nurse–physician relations may also reflect the period in which the data were collected. In December 2024, tertiary hospitals were operating amid the resident physician walkout, and nurses’ roles and working relationships with physicians were being redefined as nurses absorbed part of the care gap and medical support tasks were institutionalized under the Nursing Act (Ha & Kim, 2026; Woo & Woo, 2026). Under these conditions, the quality of collaboration with physicians may have been a particularly tangible aspect of nurses’ daily work.

Nurse participation in hospital affairs and nurse manager ability, leadership, and support of nurses were negatively correlated with turnover intention but were not significant in the regression models; this indicates not that these domains are unimportant; rather, their independent associations were not confirmed after other variables were controlled. This may reflect variance shared among the interrelated PES-NWI subscales (Cho et al., 2011) and differences in outcome and analytic approach, as a Korean path analysis found these domains to be directly associated with intent to stay, partly through burnout (Kim & Lee, 2023), whereas the present study examined turnover intention without burnout. The modest sample size may also have limited the power to detect small unique associations.

Resilience was negatively correlated with turnover intention but was not independently associated with it in the final model; this should not be interpreted as evidence that individual psychological resources are unimportant for nurse retention. Similar patterns have been reported in Korean nurses, in whom resilience was correlated with turnover or retention intention but was not a significant predictor once burnout, stress responses, or job satisfaction were included (Cha & Baek, 2026; Lee et al., 2024), although findings vary by the variables and analytic approach used (Gao et al., 2026; Zhao et al., 2021). Given that nurses rate improvements in staffing and the work environment as at least as important as resilience-focused programs (Aiken et al., 2023, 2024), resilience-building interventions are likely to be most useful when delivered alongside organizational improvements.

Nurses aged 40 or older reported significantly lower turnover intention than younger counterparts, consistent with broader evidence that younger nurses are disproportionately likely to consider leaving (Enea et al., 2024; Pressley & Garside, 2023). Notably, nurses aged 30–39 reported turnover intention as high as those aged 23–29. Considering the recent policy emphasis on retaining experienced nurses (MOHW, 2025b), this suggests that retention efforts should extend beyond the first year of practice to nurses in the middle of their careers. Because turnover intention longitudinally predicts actual departure (Ki & Choi-Kwon, 2022), mentoring and structured preceptorship for early-career nurses, together with career development support for mid-career nurses, are warranted.

This study’s cross-sectional design precludes causal inference, and self-report measures introduce possible response bias. Generalizability is limited by convenience sampling and the geographic restriction to tertiary hospitals in Seoul and Gyeonggi-do; dissemination through nurses’ networks may also have introduced clustering by workplace. In addition, data were collected in December 2024, during the resident physician walkout, which may limit the generalizability of the findings to other periods. Because resilience was entered in the final step, changes in its association before and after the inclusion of the work environment and job satisfaction could not be examined, and mediating or moderating pathways were not tested. The sample may not have provided adequate power to detect the small effect observed for resilience (β = .061). Future studies would benefit from longitudinal designs, larger and more diverse samples, and analyses testing mediating and moderating pathways among resilience, job satisfaction, the work environment, and turnover intention.

Improving the nursing work environment, particularly staffing adequacy, interprofessional collaboration, and conditions supporting quality care, should be an organizational priority. Concrete strategies include adjusting nurse-to-patient staffing ratios (Aiken et al., 2023) and expanding replacement-nurse and shift-work improvement programs, which have been associated with lower new-nurse turnover and higher experienced-nurse retention (MOHW, 2025b). As medical support tasks are institutionalized under the Nursing Act, clarifying role boundaries and collaboration structures between nurses and physicians may also help (Woo & Woo, 2026). Because job satisfaction independently predicted turnover intention beyond the work environment, interventions should also extend to recognition of professional contributions and career development support, such as clinical ladder advancement. Retention strategies co-designed with nurses, in alignment with their values and lived experiences, are likely to be most effective for younger nurses (aged 23–39), who reported higher turnover intention, and to reinforce nurses’ sense of being valued as professionals.

5. Conclusion

This study found that, after controlling for general and clinical characteristics, three nursing work environment subscales (nursing foundations for quality of care, staffing and resource adequacy, and collegial nurse–physician relations) and job satisfaction were independently associated with lower turnover intention among tertiary hospital nurses in South Korea. Resilience was negatively correlated with turnover intention but was not independently associated with it after the work environment and job satisfaction were accounted for. Hospitals should prioritize adequate staffing, interprofessional collaboration, and job satisfaction as retention targets, complement these organizational efforts with support for nurses’ personal resources, and co-design retention initiatives with nurses themselves.

References

1

Aiken, L. H., Lasater, K. B., Sloane, D. M., Pogue, C. A., Fitzpatrick Rosenbaum, K. E., Muir, K. J., et al. (2023). Physician and nurse well-being and preferred interventions to address burnout in hospital practice: Factors associated with turnover, outcomes, and patient safety. JAMA Health Forum, 4(7), e231809. https://doi.org/10.1001/jamahealthforum.2023.1809

10.1001/jamahealthforum.2023.180937418269PMC10329209
2

Aiken, L. H., Sermeus, W., McKee, M., Lasater, K. B., Sloane, D., Pogue, C. A., et al. (2024). Physician and nurse well-being, patient safety and recommendations for interventions: Cross-sectional survey in hospitals in six European countries. BMJ Open, 14(2), e079931. https://doi.org/10.1136/bmjopen-2023-079931

10.1136/bmjopen-2023-07993138346890PMC10862305
3

Bae, S. H. (2022). Noneconomic and economic impacts of nurse turnover in hospitals: A systematic review. International Nursing Review, 69(3), 392-404. https://doi.org/10.1111/inr.12769

10.1111/inr.1276935654041PMC9545246
4

Bakker, A. B., & Demerouti, E. (2007). The Job Demands-Resources model: State of the art. Journal of Managerial Psychology, 22(3), 309-328. https://doi.org/10.1108/02683940710733115

10.1108/02683940710733115
5

Cha, C., & Baek, G. (2026). Differential predictors of turnover intention among registered nurses and nursing assistants: A cross-sectional study in South Korea. International Nursing Review, 73(2), e70190. https://doi.org/10.1111/inr.70190

10.1111/inr.7019042249669PMC13241861
6

Cho, E., Choi, M., Kim, E. Y., Yoo, I. Y., & Lee, N. J. (2011). Construct validity and reliability of the Korean version of the Practice Environment Scale of Nursing Work Index for Korean nurses. Journal of Korean Academy of Nursing, 41(3), 325-332. https://doi.org/10.4040/jkan.2011.41.3.325

10.4040/jkan.2011.41.3.325
7

Enea, M., Maniscalco, L., de Vries, N., Boone, A., Lavreysen, O., Baranski, K., et al. (2024). Exploring the reasons behind nurses’ intentions to leave their hospital or profession: A cross-sectional survey. International Journal of Nursing Studies Advances, 7, 100232. https://doi.org/10.1016/j.ijnsa.2024.100232

10.1016/j.ijnsa.2024.10023239224233PMC11367642
8

Gao, L., Wang, Z., Meng, S., Xue, F., Zhang, B., Long, Z., et al. (2026). Associations of individual and team-level resources with turnover intention in nurses: A multilevel analysis. Journal of Nursing Scholarship, 58(2), e70075. https://doi.org/10.1111/jnu.70075

10.1111/jnu.70075
9

Ha, J., & Kim, M. (2026). Experiences of work performance among physician assistant nurses during the period of healthcare disruption in South Korea: A qualitative content analysis. Journal of Korean Academy of Nursing, 56(1), 98-107. https://doi.org/10.4040/jkan.25133

10.4040/jkan.25133
10

Hobfoll, S. E. (1989). Conservation of resources: A new attempt at conceptualizing stress. American Psychologist, 44(3), 513-524. https://doi.org/10.1037/0003-066X.44.3.513

10.1037/0003-066X.44.3.513
11

Hospital Nurses Association. (2026). Survey on the working conditions of hospital nurses (2025). [Press release, in Korean]. Hospital Nurses Association. Retrieved October 5, 2026, from https://khna.or.kr/home/pds/utilities.php

12

Ki, J., & Choi-Kwon, S. (2022). Health problems, turnover intention, and actual turnover among shift work female nurses: Analyzing data from a prospective longitudinal study. PLoS One, 17(7), e0270958. https://doi.org/10.1371/journal.pone.0270958

10.1371/journal.pone.027095835802575PMC9269367
13

Kim, M. R. (2007). Influential factors on turnover intention of nurses: The affect of nurse’s organizational commitment and career commitment to turnover intention. Journal of Korean Academy of Nursing Administration, 13(3), 335-344.

14

Kim, Y. B., & Lee, S. H. (2023). The nursing work environment, supervisory support, nurse characteristics, and burnout as predictors of intent to stay among hospital nurses in the Republic of Korea: A path analysis. Healthcare, 11(11), 1653. https://doi.org/10.3390/healthcare11111653

10.3390/healthcare1111165337297792PMC10252290
15

Lake, E. T. (2002). Development of the Practice Environment Scale of the Nursing Work Index. Research in Nursing & Health, 25(3), 176-188. https://doi.org/10.1002/nur.10032

10.1002/nur.10032
16

Lee, B. S., Eo, Y. S., & Lee, M. A. (2018). Development of job satisfaction scale for clinical nurses. Journal of Korean Academy of Nursing, 48(1), 12-25. https://doi.org/10.4040/jkan.2018.48.1.12

10.4040/jkan.2018.48.1.12
17

Lee, G. R., Lee, I., Chung, M., & Ha, J. (2024). Effects of grit, calling, and resilience on the retention intention of general hospital nurses. International Nursing Review, 71(4), 766–775. https://doi.org/10.1111/inr.12908

10.1111/inr.12908
18

Lee, W., & Yu, S. (2025). Moderating effect of job satisfaction in the relationship between nursing work environment and turnover intention: A perspective from Magnet hospitals in South Korea. SAGE Open, 15(2), 21582440251344769. https://doi.org/10.1177/21582440251344769

10.1177/21582440251344769
19

Lim, J. Y., Kim, G. M., & Kim, E. J. (2024). Factors associated with turnover intention among hospital nurses: A systematic review and meta-analysis. Journal of Korean Academy of Psychiatric and Mental Health Nursing, 33(1), 70-79. https://doi.org/10.12934/jkpmhn.2024.33.1.70

10.12934/jkpmhn.2024.33.1.70
20

Ministry of Health and Welfare. (2025a, February 27). The enrollment quota for the College of Nursing for the 2026 academic year has been set at 24,883, the same as the previous year. [Press release, in Korean]. Retrieved October 5, 2026, from https://www.mohw.go.kr/board.es?mid=a10503010100&bid=0027&act=view&list_no=1484827&tag=&nPage=1

21

Ministry of Health and Welfare. (2025b, August 31). Expansion of the pilot program for improving nurse shift work for patient safety and nurses’ work-life balance. [Press release, in Korean]. Retrieved October 5, 2026, from https://www.mohw.go.kr/board.es?act=view&bid=0027&list_no=1487311&mid=a10503000000

22

Moon, M., & Kang, H. (2025). Predicting turnover intention among newly graduated nurses in South Korea: A decision tree analysis. BMC Nursing, 24(1), 1411. https://doi.org/10.1186/s12912-025-04041-5

10.1186/s12912-025-04041-541254662PMC12625515
23

Organisation for Economic Co-operation and Development. (2025). Health at a glance 2025: OECD indicators. OECD. Retrieved October 5, 2026, from https://www.oecd.org/en/publications/2025/11/health-at-a-glance-2025_a894f72e.html

24

Park, M. M., & Park, J. W. (2016). Development of resilience scale for nurses. Journal of Korean Academy of Fundamentals of Nursing, 23(1), 32-41. https://doi.org/10.7739/jkafn.2016.23.1.32

10.7739/jkafn.2016.23.1.32
25

Pressley, C., & Garside, J. (2023). Safeguarding the retention of nurses: A systematic review on determinants of nurses’ intentions to stay. Nursing Open, 10(5), 2842-2858. https://doi.org/10.1002/nop2.1588

10.1002/nop2.158836646646PMC10077373
26

Skela-Savic, B., Sermeus, W., Bahun, M., Pivac, S., & Albreht, T. (2025). Reasons for the intention to leave among nurses working in internal medicine and surgery departments of Slovenian hospitals: A cross-sectional study. Zdravstveno Varstvo, 64(3), 160-166. https://doi.org/10.2478/sjph-2025-0021

10.2478/sjph-2025-002140909102PMC12406988
27

Turnbach, E., Coates, L., Vanek, F. D., Cotter, E., Pogue, C. A., Clark, R. R. S., et al. (2024). Emergency nurses’ well-being in Magnet hospitals and recommendations for improvements in work environments: A multicenter cross-sectional observational study. Journal of Emergency Nursing, 50(1), 153-160. https://doi.org/10.1016/j.jen.2023.06.012

10.1016/j.jen.2023.06.01237498276PMC10811286
28

Woo, S. J., & Woo, S. G. (2026). An analysis of changes in the healthcare system following the enforcement of the Nursing Act in 2025: Focusing on comparison with the Medical Service Act and its implications. Journal of the Korea Contents Association, 26(7), 735-745. https://doi.org/10.5392/JKCA.2026.26.07.735

10.5392/JKCA.2026.26.07.735
29

World Health Organization. (2025). State of the world’s nursing report 2025. World Health Organization. Retrieved October 5, 2026, from https://www.who.int/publications/i/item/9789240110236

30

Zhao, Y., Wang, H., Sun, D., & Ma, D. (2021). Job satisfaction, resilience and social support in relation to nurses’ turnover intention based on the theory of planned behaviour: A structural equation modelling approach. International Journal of Nursing Practice, 27(6), e12941. https://doi.org/10.1111/ijn.12941

10.1111/ijn.12941
페이지 상단으로 이동하기