Nurse turnover intention in university hospitals in Iran

Abstract

Background

Nurse turnover poses a significant challenge to healthcare systems worldwide, with direct implications for patient care quality. This study aimed to identify the factors influencing nurses’ intention to leave their positions in hospitals affiliated with the Tehran University of Medical Sciences (TUMS), Tehran, Iran.

Methods

A large-scale, descriptive-analytical cross-sectional study was conducted in 2024 across all 14 TUMS hospitals using stratified proportional quota sampling with systematic random selection (strata: hospital and nursing role) (n = 763; >92% response rate from a target of 5,411 eligible nurses). A rigorously validated 40-item questionnaire (Cronbach’s α = 0.93; CVR 0.6–1.0) assessed four domains: individual, job-related, organizational, and external factors. Data were analyzed using SPSS v26 with descriptive statistics, Pearson correlations, partial correlations (controlling for demographics), and ANOVA/Kruskal-Wallis tests (p < 0.05).

Results

Nurses reported a very high intention to leave (mean = 4.1, 95% CI: 4.09–4.19), with 97% expressing moderate-to-very high intent. Among the four dimensions, job-related factors had the highest overall mean (mean = 4.1), with heavy workload identified as the most influential component (mean = 4.5). Organizational factors followed (mean = 3.9), driven primarily by low salary and compensation (mean = 4.9). External factors ranked next (mean = 3.8), with economic pressures exerting the strongest impact (mean = 4.7). Individual factors showed the lowest mean (mean = 2.9), where anxiety and depression emerged as the most significant component (mean = 4.0). Organizational factors demonstrated the strongest correlation with turnover intention (r = 0.92), whereas individual factors exhibited the weakest correlation (r = 0.62).

Conclusion

High turnover intention was observed among nurses at TUMS hospitals. To address this problem, nursing administrators should implement comprehensive retention strategies that extend beyond mere recognition. Fostering a supportive culture, offering professional growth opportunities, and ensuring fair working conditions are critical for improving job satisfaction and retaining a skilled nursing workforce.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12912-026-05014-y.

Introduction

Nurses are key members of the healthcare team and play a critical role in delivering diagnostic, preventive, therapeutic, and palliative services [1]. Despite their importance, challenges such as staffing shortages and increasing patient demands continue to affect the ability of nurses to provide high-quality, responsive care [1, 2].

Nursing shortages play a critical role in reducing the quality of care and causing disruptions in patient services [2]. Addressing this problem is a critical priority for healthcare systems worldwide, including Iran [3]. According to the WHO projections, the global healthcare sector will face a critical shortage of approximately 50.8 million nurses and midwives by 2030. This nursing deficit is expected to be most acute in low- and middle-income countries, particularly in the WHO Eastern Mediterranean, African, and Southeast Asian regions [4]. One of the main factors contributing to this shortage is the increasing turnover rate of nurses [5, 6].

Nurse turnover refers to the actual departure of nurses from an organization. In contrast, turnover intention describes the preliminary stage in which nurses develop thoughts and plans to leave. This study specifically focuses on turnover intention among nurses, despite being currently employed, experience dissatisfaction and a desire to leave their current jobs due to various factors. Although not all intentions result in actual departure, this intention is a strong correlate of eventual turnover [7].

Recent study reported a global nurse turnover rate of approximately 16%-18%, with notable regional differences (e.g., ~ 20% in Asia, ~ 15% in North America, and ~ 7% in Europe) [8]. The UK Nursing and Midwifery Council reported that 13,945 nurses left the UK register in 2022, highlighting the scale of workforce turnover, even in high-income settings [9]. In Iran, although national-level departure statistics are not publicly centralized, internal reports from major university hospitals indicate hundreds of annual resignations, underscoring a parallel retention crisis.

Turnover poses financial and operational challenges for organizations, including the delays in service delivery, costs of recruiting, training new staff and loss of experienced professionals [10]. In nursing, this can increase the workload of the remaining staff, reduce the quality of care and increase overall healthcare costs [11].

Owing to these consequences, research attention has increasingly been directed toward understanding the reasons for nurses’ intention to leave a job [12]. Several factors affect the intention to leave a job, including individual factors (such as age, gender, education, work experience, and marital status), organizational factors (such as structure, compensation, leadership style, and interpersonal relationships), and job-related factors (such as heavy workload, inappropriate working conditions, and lack of professional independence) [13].

Tehran University of Medical Sciences (TUMS), the oldest and largest academic medical center in Iran, employs a substantial workforce of approximately 20,000 staff members, including approximately 5,000 nurses across its 14 hospitals. Internal HR records from TUMS indicate that approximately 590 nurses resigned from its hospitals between 2021 and 2024, indicating a serious retention crisis that threatens the stability of healthcare services.

Understanding the underlying factors is essential for developing effective nurse retention strategies. Recent global reports emphasize that interventions such as residency and preceptorship programs, supportive leadership, and organizational well-being initiatives play a critical role in reducing nurse turnover [14]. In addition, recent review studies have highlighted a range of evidence-based approaches, including effective workload and staffing management, career development opportunities, adequate compensation, and greater attention to nurses’ well-being and work–life balance. Collectively, this body of evidence indicates that nurse turnover arises from a complex interplay of individual, organizational, and systemic factors, underscoring the need for integrated, multi-level retention strategies across healthcare systems worldwide [15–17].

Although previous studies in Iran have explored factors related to nurse turnover and retention, they have primarily focused on local or organizational-level determinants without fully integrating broader theoretical frameworks or comparing findings with global patterns. Furthermore, most Iranian research has been cross-sectional and limited in scope, with fewer studies adopting comprehensive models that account for personal, organizational, and systemic influences simultaneously. By leveraging an integrated theoretical perspective and comparing determinants identified in Iran with key findings from international research, our study addresses this gap and offers insights relevant to both national policy makers and the global nursing community. Therefore, this study investigated the prevalence of and factors influencing the intention to leave among nurses in 14 TUMS hospitals in 2024.

Methods

Study design and setting

A descriptive analytical cross-sectional study was conducted in 2024 across all 14 hospitals affiliated with the Tehran University of Medical Sciences (TUMS). TUMS, Iran’s oldest and largest medical university (establish in 1851), operates under the Ministry of Health with 11 schools and over above100 research centers. Its 14 hospitals include general and specialized facilities. The inclusion of diverse clinical departments (e.g., medical, surgical, and ICU) enabled a comprehensive assessment of the factors affecting nurse retention.

Study population and sampling

The target population comprised all 5,411 clinical nurses employed in the 14 hospitals affiliated with Tehran University of Medical Sciences (TUMS) in January 2024 (official HR records). The minimum required sample size was 706, calculated using the prevalence proportion formula (expected turnover intention prevalence = 50%, margin of error = ± 4%, 95% confidence level, design effect = 1.3). To account for anticipated non-response, a correction factor of 1.5 was applied, resulting in a target of 1,080 participants. We employed stratified proportional quota sampling with systematic random selection. Strata were defined by hospital and nursing role according to TUMS distribution (staff nurses 70%, nurse assistants 15%, head nurses 8%, clinical supervisors 5%, nursing managers 2%). Quotas were allocated proportionally to each hospital’s nursing workforce size. Within each hospital-role stratum, nurses were further stratified by clinical ward, ordered alphabetically by employee ID, and selected using systematic random sampling (fixed interval k after a random start). Of the 1,080 invited nurses, 827 accessed the online questionnaire. After excluding 64 incomplete responses (> 20% missing data), the final sample was n = 763 (effective response rate 70.7%; completion rate among starters 92.6%). Sensitivity analyses confirmed the robustness of the results.

Inclusion and exclusion criteria

The inclusion criteria required participants to be actively employed in a clinical nursing role, specifically as a nurse assistant, staff nurse, head nurse, supervisor, or manager, in one of the participating hospitals at the time of data collection. The participants provided voluntary informed consent. The exclusion criteria included nurses working exclusively in administrative or non-clinical roles, as well as individuals unwilling or unable to complete the questionnaire in full. These criteria ensured that the study focused on clinical nursing personnel directly involved in patient care, thereby enhancing the relevance of the findings to workforce retention issues in clinical practice.

Data collection instrument

The questionnaire was developed through a sequential mixed-methods design conducted in 2024, integrating a scoping review, qualitative inquiry, and quantitative validation to ensure comprehensive item generation and psychometric rigor. A scoping review was performed in August 2024 using the six-step Arksey and O’Malley framework [18] to extract the factors influencing nurse turnover intention. International databases PubMed, Scopus, Web of Science, ScienceDirect, ProQuest, and Google Scholar were searched using the keywords Intention, Leave the job, Reason, Nursing, and Hospitals (and Persian equivalents). Of the 327 records screened, 23 studies were included (PRISMA-ScR flowchart. Supplementary File 1). In this phase, 78 unique items were identified.

To enrich the item pool, a phenomenological study was conducted with 54 nurses who resigned from TUMS hospitals between 2021 and 2023. Semi-structured interviews used the central question: “What factors led to your decision to leave?” Data collection continued until data saturation was reached. Ritchie and Spencer’s five-stage framework analysis [19] was applied: familiarization, thematic framework identification, indexing, charting, mapping and interpretation. This yielded 60 items.

After removing duplicates and merging conceptually similar items, a preliminary pool of 50 items was formed and classified into four domains: individual (ten items), job-related (eleven items), organizational (twentythree items), and external (six items). The 50-item draft (plus 20 demographic and seven outcome items) underwent rigorous psychometric evaluation. Content validity was assessed by a panel of 12 experts (four PhD faculty in health services management, three matrons, three clinical supervisors, and two head nurses; mean age 50 years, 75% female, five PhD, and seven master’s). Ten items were removed after the expert panel CVR/CVI assessment (CVR < 0.6). Face validity was evaluated through a pilot test involving 40 nurses. Impact scores were calculated as follows: Impact Score=Frequency * Importance. All items exceeded the threshold of 1.5.

The resulting 40-item version was pilot-tested with 40 nurses. Minor linguistic revisions have enhanced clarity. Internal consistency reliability was computed for the pilot sample (n = 40; 75% female, 62% married, 87% bachelor’s degree) using SPSS v26. The domain-specific Cronbach’s α values were as follows: individual (seven items) = 0.8, job-related (nine items) = 0.9, organizational (eighteen items) = 0.9, and external (six items) = 0.8; overall α = 0.9. Only minor linguistic revisions were made post-pilot (no further item removal or addition).

The final questionnaire comprised three sections: Sect.  1 included 20 demographic and occupational items (age, gender, marital status, number of children, education level, position, hospital name, employment type, local/non-local status, years of service and management experience, weekly working hours, salary level, economic status, second job, reason for choosing nursing, spousal support for nursing career, commute distance from residence to workplace, and presence of chronic illness). Section  2 contains 40 core items across four domains, all negatively worded and rated on a 5-point Likert scale from 1 (very low influence) to 5 (very high influence). Section  3 consisted of seven items measuring turnover intention and its consequences. The seven items included one direct item on turnover intention (“To what extent do you intend to leave your current job?”) Six outcomes (job stress, quality of work life, job satisfaction, organizational commitment, willingness to recommend the hospital, and performance impact) were assessed with single items using the same 5-point scale. No composite scores were formed, and all outcomes were reported individually. all rated on a 5-point Likert scale 1.0–1.8 (very low), 1.8–2.6 (low), 2.6–3.4 (moderate), 3.4–4.2 (high), 4.2–5.0 (very high). These cut points are standard in Likert-scale reporting when no empirical clustering is performed and are used here for descriptive consistency and ease of interpretation. The full questionnaire is provided in Supplementary File 2. The instrument was administered electronically through encrypted SMS links. To reduce common-method bias, we took several steps. Participation was anonymous, and we did not collect any personal information. The survey was sent through encrypted SMS links that could not be traced. Questions were clear, and items were randomly ordered within each section. Participation was voluntary, and respondents were told there were no right or wrong answers to encourage honesty.

Data analysis

Data analysis was performed using IBM SPSS Statistics (version 26). Normality was comprehensively assessed using Kolmogorov–Smirnov and Shapiro–Wilk tests, Q-Q plots, histograms, and skewness/kurtosis values (± 2 acceptable). Homoscedasticity was evaluated using Levene’s test and residual plots. Although Shapiro–Wilk tests indicated non-normality (p < 0.01 for all domains), the central limit theorem justified parametric tests given the large sample size (n = 763; >30 per subgroup). Sensitivity analyses using Spearman’s rho yielded nearly identical results.

Descriptive statistics (frequencies, percentages, means, and standard deviations) were used to summarize the sociodemographic and main variables. Pearson’s correlation was used to examine linear relationships between intention to leave and the four factor domains, interpreted as |r| < 0.30 (weak), 0.30–0.40 (moderate), 0.50–0.60 (strong), and ≥ 0.70 (very strong) [20].

One-way ANOVA was used when normality (Shapiro-Wilk P > 0.05) and homogeneity of variances (Levene’s P > 0.05) held; otherwise, Kruskal–Wallis H test was applied. Post-hoc tests (Tukey or Dunn’s) were performed after significant results. Subgroup differences were tested across demographic and occupational variables. Partial correlations controlled for age, gender, education, and work experience confirmed negligible confounding (e.g., organizational r = 0.82 → 0.81). The 95% confidence intervals were reported for key estimates. All tests were performed using a two-tailed approach.

Ethical considerations

The study protocol (internal institutional research proposal, not externally registered due to its non-interventional, survey-based design) was approved by the TUMS Research Ethics Committee (IR.TUMS.SPH.REC.1402.214). All participants were fully informed of the study objectives, procedures, potential risks, and benefits. Informed consent was obtained electronically before participation. The exact wording of the electronic informed consent screen was as follows: “Participation is voluntary and anonymous. The responses will be encrypted and cannot be linked to your identity. You may withdraw at any time, without consequence. By proceeding, you provide informed consent.” This consent form was developed based on the World Medical Association’s Declaration of Helsinki (2013). Participation was fully anonymous (no personal identifiers were collected; IP logging was disabled). Confidentiality was strictly maintained through anonymized data collection and secure and encrypted storage. Participation was voluntary, with no penalties for non-participation or withdrawal. Data will be retained for five years in encrypted form and permanently deleted thereafter. Data analysis and reporting were conducted to ensure scientific integrity and respect for the rights of the participants.

Results

Sociodemographic and work-related characteristics of the nurses

In total, 763 nurses participated in the study. The majority were staff nurses (83.0%), female (78.8%), married (68.2%), and held bachelor’s degrees (78.6%). The mean work and managerial experience were 12 and 1.5 years, respectively. The average weekly working hours were approximately 54. The detailed demographic characteristics are shown in Table 1.

Table 1.

Distribution of nurses by demographic characteristics (n = 763)

Characteristic	Variable	Frequency	Percent
Gender	Female	601	78.8
	Male	162	21.2
Marital Status	Married	520	68.2
	Single	243	31.8
Education Level	Bachelor’s Degree	600	78.6
	Master’s Degree	100	13.1
	PhD	2	0.3
Age	Under 25 years	14	1.8
	25–30 years	154	20.2
	31–35 years	188	24.6
	36–40 years	164	21.5
	41–45 years	134	17.6
	46–50 years	69	9.0
	Over 50 years	40	5.2
Having children	Yes	409	53.6
	No	354	46.4
Job position	Nurse Assistant	60	7.9
	Staff Nurse	633	83.0
	Head Nurse	48	6.3
	Supervisor	22	2.9
History of Temporary Leave from job	Yes	141	18.5
	No	622	81.5
Disease History	Yes	170	22.3
	No	593	77.7
Economic Status	Poor	320	41.9
	Moderate	367	48.0
	Good	76	9.9

Factors influencing nurses’ intention to leave

Nurses exhibited a very high level of intention to leave their current job, with a mean score of 4.1 (95% CI: 4.09–4.19) out of 5. The influencing factors were categorized into four main dimensions: individual (seven items), job-related (nine items), organizational (eighteen items), and external (six items). As shown in Table 2, job-related factors showed the highest mean with turnover intention (mean = 4.1, SD = 0.67, 95% CI: 4.10–4.20), whereas individual factors showed the lowest mean (mean = 2.9, SD = 0.81, 95% CI: 2.91–3.03) (see Table 2). Among the individual factors, the components of anxiety and depression had the greatest impact (mean = 4.0), followed by the intention to migrate. In contrast, the pursuit of higher education had the least effect (mean = 2.1). Regarding job-related factors, heavy workload had the strongest influence (mean = 4.5), whereas the need for specialized skills had the lowest score (mean = 2.7). From an organizational perspective, low salary and compensation (mean = 4.9), lack of managerial attention to nurses’ needs (mean = 4.7), and insufficient appreciation and support from managers (mean = 4.6) were the most significant factors All. external factors had mean scores above 3, indicating moderate to high influence, with economic factors (mean = 4.7) and legal/regulatory factors (mean = 4.2) having the greatest impact. Detailed item-level means and standard deviations for all 40 items are provided in Supplementary File 3.

Table 2.

Factors affecting nurses’ intention to leave their current job (n = 763)

Factors	Very Low	Low	Moderate	High	Very High	Mean ± SD	95% CI
Individual	7.7	25.3	35.9	25.2	5.9	2.9 ± 0.81	2.91–3.03
Job-Related	0.3	2.8	10.6	28.6	57.7	4.1 ± 0.67	4.10–4.20
Organizational	0.3	3.0	17.7	41.9	37.1	3.9 ± 0.68	3.88–3.98
External	0.4	6.9	23.1	36.8	32.8	3.8 ± 0.78	3.70–3.82

SD = Standard Deviation; CI = Confidence Interval

Rate of nurses’ intent to leave and performance outcomes

Turnover intention, organizational commitment, and job stress exhibited above-average mean levels. Quality of work life, job satisfaction, and willingness to recommend the hospital as a workplace were rated as below moderate. Approximately 90% of the nurses reported moderate-to-high turnover intention. The mean scores with 95% confidence intervals are presented in Table 3.

Table 3.

Levels of nurses’ intent to leave and performance outcomes (n = 763)

Factors	Very Low	Low	Moderate	High	Very High	Mean ± SD	95% CI
Quality of Work Life	47.1	13.3	30.3	4.5	4.8	2.0 ± 1.17	1.98–2.16
Job Satisfaction	54.8	13.4	26.6	3.1	2.1	1.8 ± 1.05	1.76–1.92
Organizational commitment	19.4	7.2	33.9	23.9	15.6	3.0 ± 1.30	2.99–3.19
Job stress	2.2	2.8	14.4	23.5	57.1	4.3 ± 0.97	4.24–4.38
Recommendation to work in hospital	58.3	19.9	16.9	2.6	2.3	1.7 ± 0.98	1.64–1.78
Intent to Leave	3.3	6.0	16.8	21.2	52.7	4.1 ± 1.17	4.05–4.23

SD = Standard Deviation; CI = Confidence Interval

Correlation analysis

The Pearson correlation coefficients between the demographic variables and the four factor domains (individual, job-related, organizational, and external) as well as turnover intention are presented in Table 4. Owing to the non-normal distribution of the data (Kolmogorov–Smirnov test, P < 0.05), the interpretation was based on r values and significance levels. Working hours and spouse dissatisfaction showed significant positive correlations with turnover intention, whereas age, financial status, salary, and chronic illness exhibited significant negative correlations (n = 763).

Table 4.

Pearson correlations between demographic variables and study constructs (n = 763)

Variable	Individual	Job-related	Organizational	External	Turnover Intention
Gender	0.07	−0.06	−0.07	−0.02	−0.04
Age	−0.12**	−0.08*	0.02	0.03	−0.11
Education	−0.02	0.03	0.06	0.04	−0.03
Marital Status	−0.02	−0.03	−0.02	−0.04	−0.01
Having Children	0.05	0.06	0.05	0.02	0.04
Number of Children	−0.11**	−0.06	0.00	−0.04	−0.04
Job Position	−0.07*	−0.01	−0.01	0.03	−0.01
Hospital	0.03	0.05	0.01	0.02	0.03
Employment Type	−0.07*	0.02	0.05	0.05	0.03
Work Experience	−0.12**	−0.07*	0.03	0.03	−0.02
Managerial Experience	−0.02	−0.07*	−0.01	0.02	−0.02
Working Hours	0.10**	0.04	0.04	0.07*	0.07
Housing Status	0.09**	0.01	−0.00	0.07*	0.04
Distance to Workplace	0.21**	0.00	0.01	0.03	0.06
Second Job	−0.07*	0.00	0.01	0.02	0.00
Financial Status	−0.18**	−0.10**	−0.06	−0.11**	−0.13
Salary	−0.11**	−0.13**	−0.10**	−0.08*	−0.13
Chronic Illness	−0.03	−0.06	−0.08*	−0.06	−0.08*
Prior Turnover	−0.03	−0.03	−0.03	0.03	−0.04
Spouse Satisfaction	0.35**	0.23**	0.19**	0.15*	0.28

P < 0.01 (**); P < 0.05 (*)

The mean and standard deviation of turnover intention across the demographic subgroups are reported in Table 5. Owing to non-normality, the Kruskal–Wallis H test was applied. Significant differences at the 99% confidence level were found for financial status (P = 0.00), salary (P = 0.01), chronic illness (P = 0.00), and spouse dissatisfaction (P = 0.00). No significant differences were observed in other variables (n = 763).

Table 5.

Mean turnover intention by demographic subgroups (n = 763)

Demographic Variable	Subgroup	Mean ± SD	H (df)	P
Gender	Female	3.8 ± 0.57	1.26 (1)	0.26
	Male	3.7 ± 0.63
Financial Status	Very Poor	4.0 ± 0.56	17.76 (4)	0.00
	Poor	3.8 ± 0.58
	Average	3.7 ± 0.58
	Good	3.7 ± 0.58
	Very Good	4.4 ± 0.31
Salary (Million Rials)	< 100	4.0 ± 0.51	13.93 (5)	0.01
	100–150	3.8 ± 0.57
	150–200	3.7 ± 0.59
	200–250	3.6 ± 0.60
	250–300	3.5 ± 0.65
	> 300	3.8 ± 0.71
Chronic Illness	Yes	3.8 ± 0.54	8.74 (1)	0.00
	No	3.7 ± 0.60
Spouse Satisfaction	No	3.9 ± 0.53	35.64 (1)	0.00
	Yes	3.6 ± 0.61

SD = Standard Deviation; df = degrees of freedom

Table 6, Fig. 1 presents a network visualization of the bivariate Pearson correlations between turnover intention and the four factor domains: individual, job-related, organizational, and external. The edge thickness is proportional to the absolute correlation coefficient (|r|), and all correlations were statistically significant at p < 0.01. The analysis was based on a sample size of n = 763. The correlation results indicated significant positive associations between turnover intention and all four domains. Among these, organizational factors exhibited the strongest positive correlation with turnover intention (r = 0.92, p < 0.01), followed by job-related (r = 0.82, p < 0.01), external (r = 0.77, p < 0.01), and individual factors (r = 0.62, p < 0.01). These findings suggest that contextual and work environment–related conditions are more strongly associated with employees’ turnover intention than individual-level attributes. The particularly strong association between organizational factors and turnover intention may be explained by the high influence of organizational components identified in this study, especially low salary and compensation, lack of managerial attention to nurses’ needs, and insufficient managerial support and recognition. These factors received the highest mean scores among organizational items and appear to be closely linked to nurses’ decisions to leave their current positions.

Table 6.

Correlations between factors and nurses leaving their current job (n = 763)

Dimensions	Individual	Job-related	Organizational	External
Individual	1	0.47**	0.37**	0.33**
Job-related	0.47**	1	0.67**	0.51**
Organizational	0.37**	0.67**	1	0.69**
External	0.33**	0.51**	0.69**	1
Turnover Intention	0.62**	0.82**	0.92**	0.77**

The correlation coefficients used in Fig. 1 were extracted directly from this matrix

p < 0.01 (**); p < 0.05 (*)

Fig. 1.

Correlations between factors and nurses leaving their current job. Pearson correlation coefficient (|r|); all correlations significant at p < 0.01

The correlation analysis presented in Table 7 reveals a coherent and significant pattern of results. The factors that cause nurses to want to leave their current jobs (individual, job-related, organizational, and external) are significantly negatively correlated with positive states such as job satisfaction and organizational commitment but are significantly positively correlated with job stress. Consequently, the conditions that push nurses to leave are the same ones that undermine their job satisfaction and commitment and heighten their stress levels. The results solidly connect the influencing factors not only directly to turnover intention but also to its underlying attitudinal and psychological antecedents.

Table 7.

Correlation between factors affecting turnover intention and outcome dimensions (n = 763)

Dimensions	Individual	Job-related	Organizational	External	Turnover Intention
Quality of Work Life	-0.05	-0.18**	-0.04	-0.03	-0. 92*
Job Satisfaction	-0.19**	-0.32**	-0.21**	-0.15**	-0.27**
Organizational commitment	-0.18**	-0.22**	-0.17**	-0.15**	-0.22**
Job stress	0.26**	0.25**	0.21**	0.19**	0.27**
recommend hospital employment	-0.19**	-0.26**	-0.20**	-0.15**	-0.25**

p < 0.01 (**); p < 0.05 (*)

Discussion

Nurse turnover rate

The present study revealed a very high level of intention to leave (mean score: 4.1 out of 5) among nurses in hospitals at Tehran University of Medical Sciences, with approximately 97% of the respondents reporting moderate to very high intention. This finding aligns with recent global trends, where RN turnover has declined to 15.7% in U.S. hospitals (2025 NSI Report), yet remains elevated in LMICs like Iran (~ 45% intention to leave) due to economic and systemic factors [21]. Systematic reviews have confirmed that residency programs can reduce early turnover by 20%–30% among new graduates [22, 23].

For example, a study in 11 European countries reported that 22% of nurses intended to leave their jobs [24]. In South Korea, the annual turnover rate for all nurses is approximately 12.4%, but for newly licensed registered nurses (NLRNs), it can reach 42.7% in the first year and 46.3% within three years [25]. These figures confirm that the intention to leave is a widespread problem, even in developed healthcare systems in high-income countries.

The Job Demands-Resources (JD-R) model [26] explains turnover intention through an imbalance between job demands (energy depleting) and job resources (motivational). Our findings strongly support this hypothesis. High demands were evident in excessive workload (“high patient-to-nurse ratio per shift,” mean = 4.5), long working hours (54 h/week), emotional load (“dealing with critical patients,” mean = 4.4), and role ambiguity (“unclear duties due to protocol changes,” mean = 4.1), with job-related factors showing a very strong correlation with turnover intention. Insufficient resources included inadequate salary (mean = 4.9), lack of managerial support (mean = 4.7), poor recognition (mean = 4.6), and limited autonomy (mean = 3.8), which reflected the strongest correlation among organizational factors. External pressures, such as inflation (mean = 4.7), acted as contextual demands, extending the JD-R model beyond the workplace—a key adaptation for LMIC settings. The inverse link with organizational commitment confirmed the motivational pathway. This demand-resource imbalance is strongly associated with the observed high turnover intention. and highlights actionable targets: reducing workload and enhancing support and compensation for nurses.

Individual factors influencing turnover

Among the individual factors, work-related anxiety and depression were the most significant correlates of turnover intention, whereas the pursuit of continuing education had a negligible effect. In support of this, another study categorized individual factors into three subgroups: “physical factors,” “mental and psychological factors,” and “social factors.” Within the physical factors section, four components were investigated, among which three, namely, “proximity of residence to the hospital,” “workplace location” and “ease of commute and access to public transportation,” were identified by nurses as influential factors in their decision to continue working [27]. These results indicate that the distance between home and workplace, as well as commuting convenience, significantly affects nurses’ decisions to stay or leave their jobs; difficulties in these areas increase the likelihood of turnover intention. Migration, especially to developed countries, is a growing trend among Iranian nurses, often motivated by the pursuit of better professional and welfare conditions.

According to the World Health Organization, one of the main global reasons for nurse turnover is the desire to migrate to countries that offer better job opportunities, higher salaries, and more favorable working environments [28]. Family related issues create conflicts between work and family roles; such conflicts arise when family needs and expectations are at odds with job duties and responsibilities [29]. Numerous studies have demonstrated a significant positive relationship between work and family conflict and nurses’ intention to leave [30]. In a 2018 study entitled “Family Responsibilities and Nurse Turnover,” nurses who had left their jobs were asked about their reasons for doing so. One prominent and recurring factor reported by participants was family responsibilities; some nurses chose to leave their nursing careers because of pressures related to the need to care for and attend to their spouses and children [24].

In the present study, several demographic and work-related variables showed significant associations with nurses’ intention to leave. Spousal dissatisfaction demonstrated a positive and meaningful correlation with turnover intention, consistent with evidence indicating that emotional and social support from a spouse can buffer work stress and reduce the likelihood of leaving [31]. Among occupational factors, longer working hours were associated with higher intention to leave, aligning with previous studies that link extensive work hours to burnout and withdrawal behaviors [32, 33].

Economic factors also played an important role. Lower monthly salary showed a significant negative correlation with intention to leave, Schug et al. reported that higher perceived rewards reduce absenteeism and turnover intention among nurses. Finally, the presence of a chronic underlying illness was negatively associated with turnover intention, a relationship supported by studies showing that nurses with physical or psychological conditions tend to experience more work strain and absenteeism [34]. The comparatively weaker association of individual factors with turnover intention suggests that personal characteristics alone are less influential than workplace and organizational conditions in shaping nurses’ decisions to leave. While factors such as anxiety, depression, and migration intentions remain important, they may exert their effects indirectly through interactions with job-related and organizational stressors rather than acting as primary drivers of turnover intention.

Job-related factors influencing turnover

Job-related factors were the strongest correlates of turnover intention, predominantly strongly associated with heavy workload and long working hours, whereas the need for specialized skills had a negligible influence. Recent meta-analyses reinforce that heavy workloads are strongly correlated with burnout and turnover intention (r = 0.72), but preceptorship-integrated residency programs mitigate this by enhancing competence and reducing intent to leave by up to 25% in the first year [22].

Bae identified job characteristics, workload, physical and psychological work environment conditions, opportunities for training and advancement, and the degree of control and self-efficacy as key occupational factors [35]. Studies have demonstrated that excessive and inappropriate workloads correlational to turnover among nurses.

Excessive workloads correlational turnover (r = 0.72), but recent industry reports highlight that continuous monitoring and flexible scheduling—such as gig shift platforms—can reduce the intent to leave by 15%–20% through improved work-life balance. In resource-limited settings such as Iran, adopting these strategies could mitigate heavy workloads without large-scale hiring [36]. In the study by Chegini et al., job identity, workload, responsibility, role ambiguity, and role conflict were identified as occupational stressors and correlates of turnover intention among staff. Their results revealed a direct relationship between occupational stress and job identity, workload considering time pressure, the need for increased knowledge to perform tasks (qualitative workload), role ambiguity, and conflicting demands at work, and a significant negative correlation with poor quality of work life and turnover intention [37].

Nurses with lower professional autonomy are more prone to turnover because professional independence plays a crucial role in job satisfaction and their sense of competence [38]. When nurses are unable to make independent decisions or actively participate in treatment processes, feelings of inefficacy and worthlessness gradually develop, which can diminish their motivation to continue in the profession.

One major challenge for nurses, particularly during the COVID-19 pandemic, was the inconsistency and instability of the care guidelines and protocols designed for patient management. Frequent changes in these guidelines and administrative and organizational structures to adapt to new conditions have caused confusion and disruption in nurses’ work flows. This situation directly affects the clarity of their duties and is linked to ambiguity in job descriptions and professional roles [21]. Role ambiguity is a strong factor that increases occupational stress, which not only reduces the quality of patient care but also threatens nurses’ mental health and their motivation. Moreover, the lack of positive feedback and support from managers, leaders, and even the community, despite enduring very high pressure and long working hours, fosters feelings of frustration and burnout among nurses [38]. This lack of appreciation and disregard for nurses’ extensive efforts reduces their motivation and job satisfaction and gradually increases their intention to leave the profession. The prominence of workload as a determinant of turnover intention may reflect the persistent mismatch between patient care demands and available nursing resources. When nurses experience sustained work overload, physical exhaustion, emotional fatigue, and reduced work-life balance accumulate over time, increasing their motivation to seek alternative employment opportunities. Similarly, low compensation may be perceived as inadequate recognition of the intensity and complexity of nursing work, thereby amplifying dissatisfaction and withdrawal intentions.

Organizational factors influencing turnover

Low salaries, lack of managerial attention, and insufficient appreciation and support from managers play the greatest role in nurses’ intention to leave their jobs. Harvey et al. identified organizational factors, including organizational policies and guidelines, structure, leadership, culture, and support. These factors can influence nurses’ attitudes and behaviors, thereby affecting their turnover [39]. The type of leadership and management within an organization can significantly impact turnover rates. In the conceptual model proposed by Untarini et al., elements such as organizational structure, policies and procedures, compensation, promotion and job security, leadership style, training programs, and work shifts are identified as organizational components and sources of occupational stress [40].

The exceptionally strong correlation observed between organizational factors and turnover intention (r = 0.92) further underscores the central role of organizational conditions in shaping nurses’ employment decisions. This finding is consistent with previous studies included in our review, which identified organizational policies, leadership practices, managerial support, compensation systems, and workplace culture as major determinants of nurse retention and turnover intention [39, 40]. The strength of the association observed in the present study may reflect the cumulative impact of multiple organizational challenges reported by participants, particularly dissatisfaction with salary, limited managerial support, and inadequate recognition of nurses’ contributions.

The organizational structure defines the manner in which roles, power, authority, and responsibilities are assigned and managed, as well as the flow of information between different levels of the organizational hierarchy. It also reflects the number of levels within the administrative hierarchy and delineates managers’ or supervisors’ span of control [41]. lack of support and low salaries are strongly associated with higher nurse turnover intention. Conversely, investing in nurse manager well-being—through leadership training, recognition, and supportive practices—significantly lowers turnover rates and early-tenure attrition [42]. The quality of intraorganizational communication among colleagues, managers, and nurses plays an important role in nurse turnover. Poor organizational communication and inefficiency may lead to nurses leaving their jobs [43].

An inadequate salary is a significant factor in nurse turnover. Nursing salaries in Iran are not comparable to those in other countries, which can contribute to turnover [44]. However, a study by Varesteh et al. showed that although financial rewards and incentives are considered motivating factors for continuing work, especially among younger nurses, the nurses in their study preferred job stability over increased financial income, believing that improving employment conditions and establishing job security would eventually lead to better financial situations [45].

External factors influencing turnover

Evaluating external organizational factors can assist managers in identifying environmental opportunities and threats, enabling the development of effective strategies to improve organizational performance. In healthcare organizations, a precise understanding of external environmental forces is especially critical, as changes in these factors can substantially affect organizational functioning and sustainability. External threats, in particular, which often have negative effects on the performance and quality of services, require careful attention and management [46].

Economic issues and inflation were identified as the most significant external factors influencing nurse turnover. This finding aligns with Teti̇k et al. research, which shows that macroeconomic conditions, especially unemployment rates and economic recessions, can increase nurse turnover. During periods of high unemployment, nurses may decide to leave their positions in pursuit of better job opportunities and higher wages [47]. Furthermore, nurses often perceive their financial compensation as disproportionate to their workload and occupational hazards, and this economic dissatisfaction is a primary reason for turnover [45]. Therefore, nurse turnover can be influenced by multiple labor market factors, particularly when alternative job opportunities offering higher income and better conditions are available, intensifying the issue. Indeed, turnover rates are directly related to the frequency and attractiveness of job options with higher economic returns relative to the required effort [48]. Consequently, attention to economic conditions and the creation of an appealing work environment that addresses nurses’ financial and professional needs are essential to reduce turnover rates and maintain workforce stability in the health sector. External economic pressures may also interact with organizational factors. For example, inflation and rising living costs can intensify dissatisfaction with existing salary levels and organizational reward systems, making deficiencies in compensation and managerial support more salient. Consequently, broader economic challenges may strengthen the impact of organizational shortcomings on nurses’ turnover intention.

Theoretical contributions

This study offers several theoretical contributions to the literature on nurse turnover and human resource management in healthcare. First, by examining turnover intention within a resource-constrained and politically complex healthcare system, the study extends the applicability of existing turnover frameworks such as the job demands–resources model and Herzberg’s two-factor theory into a new contextual setting. Our findings reveal that external factors, particularly macroeconomic instability, weak legal protection, and governmental policy shortcomings, play a more prominent role than has been emphasized in prior models developed in high-income countries. Second, the study refines the categorization of turnover antecedents by empirically validating a four-dimensional framework encompassing individual, job-related, organizational, and external factors. This structure provides greater conceptual clarity and highlights the multilevel nature of turnover intention. Third, the data suggest synergistic interactions between these domains, particularly how inadequate managerial support can intensify the effects of heavy workload or low compensation, indicating the need for integrative, systems-level models. Finally, the study challenges HRM theories that overly emphasize individual-level solutions, such as professional development or engagement, by demonstrating that structural and policy-level barriers substantially correlational intent to leave. These insights contribute to the development of more context-sensitive, politically aware models of healthcare workforce retention.

Strengths and limitations

Although correlation coefficients cannot establish causality, the validity of our findings is supported by: (1) a large, representative sample (n = 763, > 92% response rate) with narrow 95% CIs, enhancing precision; (2) robust psychometric properties of the instrument (α = 0.83–0.94 per domain); (3) adjustment for key demographics via partial correlations, which showed negligible change in r values (e.g., organizational:0.828→0.811); and (5) theoretical grounding in the Job Demands-Resources (JD-R) model, which posits that high demands and low resources correlate strongly with burnout and turnover intention—even in cross-sectional data—as a necessary precursor to longitudinal validation.

This study has several limitations. The cross-sectional design precludes causal inferences, as observed high correlations may reflect bidirectional relationships or unmeasured third variables. Although partial correlations controlled for key demographics, residual confounding remains possible, and longitudinal or mixed-methods studies are recommended to establish temporality and causality. Data were self-reported, potentially introducing social desirability bias, recall bias, and common method variance, despite mitigation efforts including ensured anonymity, encrypted distribution, voluntary participation, and absence of reversed items. Additionally, unmeasured confounders—such as personality traits, work-unit characteristics, unit-level staffing ratios, shift patterns, and hospital case-mix—may have influenced the results. The study was restricted to hospitals affiliated with Tehran University of Medical Sciences, limiting generalizability to other regions or healthcare settings in Iran. Finally, although the questionnaire showed strong content validity and internal consistency reliability, exploratory or confirmatory factor analysis was not performed; future research should incorporate such validation along with objective metrics and longitudinal designs. Despite these procedural mitigations, the potential for common-method bias cannot be entirely ruled out in self-report studies. However, the complex multivariate relationships observed (e.g., differential correlations across domains) suggest that common-method variance alone does not account for the pattern of findings. These procedural remedies likely reduced respondents’ evaluation apprehension and response consistency biases, thereby decreasing the likelihood of artificially inflated correlations; however, they cannot completely eliminate the possibility of common method bias in cross-sectional self-report studies.

Recommendations for future research

Future studies should employ longitudinal designs to establish causality between identified factors and actual turnover. Mixed-methods approaches could further explore contextual nuances, particularly the role of external macroeconomic policies. Comparative research across Iranian provinces and with high-income settings would enhance generalizability. Additionally, intervention studies evaluating retention programs (e.g., preceptorship, flexible scheduling, leadership training) in TUMS hospitals are needed to assess effectiveness in reducing turnover intention. Future research should also integrate objective metrics, such as actual turnover data from human resources, to complement self-reported intention measures and provide a more robust assessment of turnover behavior.

Conclusion

Given the very high intent to leave among nurses, it is essential for healthcare managers to take proactive steps to improve working conditions and organizational support. Structured nurse residency and preceptorship programs, along with supportive and transformational leadership training for nurse managers, should be prioritized as evidence-based interventions proven effective in reducing turnover intention. Additional key solutions include increasing financial and psychological support, reducing workload by hiring additional staff and adopting flexible scheduling, establishing clear career progression pathways, and introducing structured recognition programs. Enhancing workplace conditions, clarifying job roles, strengthening managerial communication, and providing organizational benefits are crucial measures for reducing nurse turnover. Implementing these strategies not only helps retain qualified nurses but also contributes to better healthcare service quality and improved patient satisfaction.

Implications for practice and policy

Practical recommendations emerging from this study include implementing structured leadership training programs to increase managerial support and communication, optimizing nurse-to-patient ratios and shifting scheduling to reduce workload, and introducing salary benchmarking and financial incentives to address economic dissatisfaction. For example, hospitals may adopt workload redistribution models based on patient acuity and nurse-to-patient ratios to reduce excessive workload in high-demand units. Structured mentorship and residency programs for newly employed nurses can facilitate professional adaptation, strengthen organizational commitment, and reduce early-career turnover. In addition, compensation reform initiatives, such as performance-based incentives, retention bonuses, and periodic salary adjustments aligned with inflation rates, may help address financial dissatisfaction and improve workforce retention. Additionally, hospitals should invest in psychosocial support systems, such as counseling and debriefing sessions, to mitigate job stress and burnout. Promoting career advancement opportunities and recognition programs can also increase motivation and commitment. Finally, broader policy reforms such as adjusting labor laws and addressing inflation impacts are essential for creating a supportive environment that reduces turnover intention among nurses.

Supplementary Information

Below is the link to the electronic supplementary material.

Abbreviations

TUMS

Tehran university of medical sciences

JD-R

Job demands–resources model

IRB

Institutional review board

Author contributions

Conception and design: Ali Mohammad Mossadegh Rad, Hoda Ghobishipour, Nilofar Amiri Qala Rashidi.Administrative support: Hossein Dargahi, Nilofar Amiri Rashidi.Provision of study materials or patients: Nilofar Amiri Qala Rashidi.Collection and assembly of data: Hoda Ghobishipour, Mohammad Hussein Nozarian.Data analysis and interpretation: Ali Mohammad Mossadegh Rad, Hoda Ghobishipour.Manuscript writing: Hossein Dargahi, Hoda Ghobishipour, Muhammad Hossein Nozarian.Final approval of manuscript: Ali Mohammad Mossadegh Rad, Hoda Ghobishipour, Mohammad Hussein Nozarian, Hossein Dargahi, Nilofar Amiri Qala Rashidi, Maryam Zahmatkesh.

Funding

The authors received no specific funding for this research.

Data availability

“The anonymized dataset (SPSS .sav file with variable/value labels) supporting the findings is available from the corresponding author (Hoda Ghobishipour) upon reasonable request and approval by the TUMS Ethics Committee, in line with institutional data protection policies.”

Declarations

Ethics approval and consent to participate

The study protocol was reviewed and approved by the Research Ethics Committee of the School of Public Health, Tehran University of Medical Sciences, Tehran, Iran (Ethics approval code: IR.TUMS.SPH.REC.1402.214). All procedures performed in this study involving human participants were conducted in accordance with the ethical principles of the Declaration of Helsinki. Before participation, all eligible nurses were fully informed about the purpose, procedures, potential risks, and benefits of the study. Written informed consent to participate was obtained electronically from all participants prior to data collection. Participation was voluntary, and confidentiality and anonymity were strictly maintained throughout the research process.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.