ABSTRACT

Objective

Türkiye is experiencing a significant physician attrition with departure of young physicians. This study aimed to assess the rate and associated factors of intention to leave (ITL) among resident physicians working in a university hospital.

Methods

A cross-sectional, questionnaire-based survey was conducted with the actively working resident physicians in a government university hospital. Data were collected using five different questionnaires during face-to-face interviews: a questionnaire for sociodemographic data, the Turkish versions of the work-family conflict (WFC) and family-work conflict (FWC), Job Satisfaction Inventory (JSI), Copenhagen Burnout Inventory, and the Intention To Turnover Scale (ITS). Correlation and regression analysis determined factors associated with the ITL.

Results

45.8% of the residents would quit their jobs if they had any chance, and 41.8% had recently thought more often about quitting their jobs. 12.2% of the residents were actively looking for a new job, and 43.4% planned to quit. Marital status, having children, compulsory service history, working hours, number of shifts at the weekends, income status, institution satisfaction, satisfaction with the area of specialization, JSI scores, WFC-FWC scores and burnout scores were found to be correlated with ITS scores. Among these factors, total job satisfaction, FWC, and work-related burnout scores predicted ITL the current job independently among these physicians.

Conclusion

Resident physicians have a high ITL their current jobs. The main contributing factors were decreased job satisfaction, FWC, and work-related burnout. Improvement of job satisfaction and work conditions, and implementation of family-friendly practices are essential to better retain these physicians.

INTRODUCTION

Türkiye is currently experiencing significant physician attrition that leads to the increasing departure of young Turkish physicians abroad. A recent countrywide multicenter study reported that 70.7% of medical students have emigration intentions (1). The “working conditions in the country” were reported to be the main push factor (1). Türkiye already has a physician shortage, with 1.9 doctors per 1,000 people, less than many Organization for Economic Co-operation and Development countries: the average is 3.5/1,000 (2). Numerous socioeconomic, personal, interpersonal, and organizational factors contribute to this tendency to emigrate.

The practice of medicine has always been challenging. It requires extensive medical school education, additional postgraduate resident training, and many years of practice to gain adequate experience. In modern times, physicians are confronted with lengthy, inflexible working hours, many night shifts, frequent overtime, a heavy workload, stressful working environments, emotional contacts, inadequate hospital resources, conflicts with patients and patients’ families, and economic issues (3). The response of the physicians to these conditions varies greatly. Some of them burn out from this stressful environment, feel unable to continue their work, and want to leave their current job. In Türkiye, almost no empirical study assessed this topic among physicians; however, the rate of physicians who want to leave their jobs ranges from 3.2% to 53.7% (4) in developed countries. A national survey study reported this rate as 20.5% among Chinese physicians (5) and 14.5% among Taiwanese physicians (6).

The departure of physicians from jobs decreases access to healthcare for the population. In addition, recruiting and training new physicians is not easy, as the training period is longer and more expensive than in other occupations. Retaining physicians has become an essential issue for the continuation of health services. Studies show that intention to leave (ITL) is a vital precursor of actual leaving (7); and underlying reasons must be understood for prevention. Therefore, it is crucial to study physicians’ ITL in order to identify the high-risk group for early intervention before leaving occurs.

The time of medical specialization is considered the most challenging period during a physician’s career (8). This residency period directly follows graduation from medical school and usually has a heavy workload with high demands and limited job control. Also, this stage of life is the family-founding and child-having stage, which may also contribute to work-family conflict (WFC) and family-work conflict (FWC) (9).

The explanatory factors of ITL the current job are usually grouped under demographic factors, family or personal domain, working hours and conditions, job-related well-being, and career aspects (4). The present study aims to assess the rate and associated factors of ITL among resident physicians working in a university hospital, including the above-mentioned explanatory factors.

METHODS

The study was designed as a cross-sectional, questionnaire-based survey conducted at Mersin University Hospital between 1 December 2023 and 31 January 2024. The inclusion criteria were being actively working as a resident physician at Mersin University Hospital and agreeing to contribute to the study. There were 496 actively working resident physicians and all were invited to participate. Of these, 371 responded and completed the questionnaires. Informed consent was obtained from all of the responders. Data were collected using five different questionnaires during face-to-face interviews with two researchers. The questionnaires were selected to assess all possible explanatory factors: demographic characteristics, family/personal space, working time, salary and job satisfaction. The study was approved by Mersin University Clinical Research Ethics Committee (approval no: 2023/824, date: 05.12.2023).

The first questionnaire included questions regarding the resident physicians’ sociodemographic data, working conditions, and income status. Sociodemographic data included age, gender, marital status, parenthood, area of speciality, professional status, personnel cadre, weekly working hours, number of night shifts during weekdays and weekends, income status, institution satisfaction, satisfaction with the area of specialization and satisfaction with the salary. Satisfaction was assessed with a three-question scale: not satisfied, neither satisfied nor dissatisfied, and satisfied.

The second questionnaire was the Turkish version of the WFC-FWC inventory. The original WFC-FWC inventory was developed by Netemeyer et al. (10). The questionnaire is composed of two subscales, WFC-FWC, and each includes five items and uses a 5-point Likert response scale (from “strongly agree” to “strongly disagree”). The values of the WFC-FWC Scales are calculated as the sum of the responses to the five items, with five being the lowest and 25 being the highest score. The higher scores represent more conflict. The Turkish version was adapted and validated by Kolbaşı and Bağcı (11) with a Cronbach alpha of 0.88 for WFC and 0.89 for FWC. The Cronbach alpha values for WFC and FWC were calculated as 0.896 and 0.852 in the present study.

The third questionnaire was the Job Satisfaction Inventory developed by Kuzgun et al. (12) in 2005. It consists of 20 items and uses a 5-point Likert scale (from “never” to “always”). High scores on the scale mean that the individual’s job satisfaction is high. The Cronbach alpha value of the original inventory was reported to be 0.91, and in the present study, it was found to be 0.876.

The fourth questionnaire was the Turkish version of the Copenhagen Burnout Inventory (CBI) developed by Kristensen et al. (13) in 2005. This inventory consists of 19 items and assesses burnout across three domains: personal, work-related, and client (patient in the present study)-related burnout (14). The Turkish version was adapted and validated by Bakoğlu et al. (14) with Cronbach alpha values of 0.903, 0.857, and 0.669 for personal burnout, work-related burnout and patient-related burnout domains, respectively. It uses a 5-point Likert scale (from “never: 0 points” to “always: 100 points”), and higher scores indicate a higher level of burnout symptoms. A total domain score is obtained, by taking the average scores for each question within each domain. A total domain score of 50 points or higher is considered a moderate-to-high degree of burnout (13). The Cronbach alpha values for personal, work-related, and patient-related burnout domains were calculated as 0.885, 0.867 and 0.838, respectively, in the present study.

The fifth questionnaire was the Intention to Turnover Scale (ITS), developed by Rosin and Korabik (15) in 1995, translated to Turkish, and validated by Tanrıöver (16) in 2005 with a Cronbach’s alpha value of 0.930. The scale was revised by Torun (17) with a Cronbach alpha value of 0.835, and this revised form was used in the present study. The questionnaire consists of 4 items and uses a 5-point Likert scale (from “strongly disagree” to “strongly agree”). The fourth question, “I do not plan to quit my job”, is scored in reverse. The higher scores indicate an increased ITL the current job. “Strongly agree” and “agree” were considered positive for the first three questions, and “strongly disagree” and “disagree” were considered affirmative for the last question. The Cronbach alpha value was found to be 0.825 in the present study.

Statistical Analysis

Statistical analysis was accomplished with IBM SPSS Statistics version 21 (IBM Corp., Armonk, NY, USA). The normality of the data was analyzed with the Kolmogorov-Smirnov test and histograms. Descriptive data were expressed as percentages, means and standard deviations. T-tests and one-way analysis of variance (ANOVA) tests were used where appropriate to compare normally distributed parametric data. For post hoc comparisons of ANOVA, the Bonferroni test was used. Correlation analysis was used to determine the variables related to the ITS scores. Regression analysis was conducted with the significant variables. A p-value of <0.05 was considered to be significant.

RESULTS

The response rate was 74.8% among all the resident physicians working in the hospital; the sociodemographic characteristics of the 371 responding physicians are shown in Table 1. The mean age was 29.8±3.3 years, and more than half of the residents, 50.7%, were younger than 30. The male-to-female ratio was almost equal (1.07), and 55.5% were married. Most of the residents were living in rented houses (70.9%), and among those who were married, 85.4% of the spouses were working. Seventeen residents (4.6%) had completed residency training in a main specialty and had passed an examination to get training in a subspecialty. 55.5 percent of the residents were working in internal medicine sciences departments, and 40.2 percent were working in surgical departments. Türkiye has two main bodies administering residency programs for Turkish physicians: The Council of Higher Education and the University of Health Sciences. After medical school, physicians are obligated to complete a compulsory service in different parts of the country for a limited time if they do not pass the examination for residency. In the enrolled population, 55.5% of the residents were in the University of Health Sciences program. There were only eight physicians (2.2%) from various foreign countries. 79.5% of the Turkish physicians had completed their compulsory service obligations. 46.1% of the attendants worked more than 60 hours per week. The average number of night shifts during weekdays and weekends in a month were 3.8±1.6 shifts and 1.9±0.8 shifts, respectively. 43.7% of the residents reported that their income was less than their expenses, and only 2.2% were satisfied with their salaries. Satisfaction with the institution and the area of specialization was reported to be 29.6% and 42.96%, respectively (Table 1).

Table 2 shows the residents’ responses to the ITS questions. It was found that 45.8% of the residents would quit their jobs if they had any chance, and 41.8% had recently been thinking more often about quitting their jobs (Table 2). 12.2% of the residents were actively looking for a new job, and 43.4% planned to quit their jobs (Table 2).

Comparisons of the ITSv scores between the groups, based on residents’ characteristics, are shown in Table 3. The total ITS score was significantly higher in married residents than in single residents: 11.71±4.4 vs. 10.61±4.1, p<0.004. The total score was higher among residents who do not have any children (12.09±4.0 vs. 10.87±4.9) and whose spouses were working (11.88±4.6 vs. 10.60±3.9). However, these differences were not significant. ITL was greater among residents who completed their compulsory service tasks (11.57±4.3 vs. 9.95±4.1; p=0.003). Total ITS scores were significantly higher for residents working more than 80 hours than for those working less than 60 hours. There was no difference in total ITS scores between residents working 60 to 79 hours per week and more than 80 weeks concerning total ITS scores. The scores were found to be significantly lower in residents whose income was more than their expenses when compared to those whose income was less than or just covering their expenses (9.47±3.8 vs. 12.01 4.3, p<0.001; 9.47±3.8 vs. 11.18±4.4, p=0.002). The residents who were satisfied with the institution and the area of specialization, had significantly lower total ITS scores than those who were not satisfied and neither satisfied nor dissatisfied. The personal, patient-related, and work-related burnout scores were ≥50 in 75.2%, 85.4%, and 81.4% of the resident physicians, and the mean total ITL scores were significantly higher in these physicians (Table 3). Work-related burnout is “the degree of physical and psychological fatigue and exhaustion that the person perceives as related to his/her work” (13). This domain assesses work-related fatigue with seven questions. 66.5% of the residents responded “often” or “always” to the two questions: “Do you feel burnt out because of your work?” and “Is your work emotionally exhausting?”

Factors that were significantly correlated with the total ITS scores are depicted in Table 4. Being married, having children, having non-completed compulsory service tasks, income status, satisfaction with income, satisfaction with the area of specialization, and total job satisfaction scores were negatively correlated with the total ITS scores (Table 4). Weekly working hours, number of weekend shifts per month, WFC scores, FWC scores, patient-related burnout scores, personal-related burnout scores, and work-related burnout scores were positively correlated with the total ITS scores (Table 4). Among these significantly correlated parameters, only total job satisfaction scores, FWC scores, and work-related burnout scores were independent predictors of the ITS scores (Table 5).

DISCUSSION

This study shows that the assessed resident physicians have a high ITL their current jobs, and about 1 in every 8 of them not only planned to quit, but also had already been looking for another job. Although most of them were not satisfied with their salaries and had more expenses than income, the main contributing factors for the ITL were job dissatisfaction, WFC, and work-related burnout.

Intention to quit is the tendency to deliberately leave the organization in a conscious and planned manner (18). Medical doctors are accomplished individuals who have undergone many selective examinations in every population and have a pivotal role in the healthcare system. Therefore, the departure of doctors from jobs risks decreasing healthcare delivery capacity. In addition, the burden of increased demand for clinical care shifts to the remaining physicians. This would eventually lead more physicians to leave their jobs, causing a vicious cycle. Studies showed that ITL is a vital precursor of actual turnover (4). The reported rate of intent to leave among physicians ranged from 11.8% to 22% in five developed countries (4). However, these studies were conducted before the coronavirus disease 2019 pandemic. In a recent study from the United States of America conducted during the pandemic, 32.6% of 15 890 physicians reported ITL (19). Our study expands the literature by providing novel information about the prevalence of ITL among Turkish resident physicians. The rate of ITL among Turkish physicians has yet to be studied, the rate of ITL among the residents only has not been assessed in the literature before. Residency is a unique, lengthy, and rigorous training period during a medical career, and continuing the medical career is associated with how successfully residents manage the demanding work-life balance of medicine. In the present study, the rate of ITL was found to be 43.4% among the residents. Additionally, 12.2% were actively looking for a new job.

Studies have shown that burnout is associated with increased physician turnover (7). Several organizational, social, and job-related factors result in feelings of burnout, which interrupt coping with personal and interpersonal demands (20) and lead to quitting clinical practice. The prevalence of burnout among physicians has been reported to range from 0% to 86.2% (21). This comprehensive range results from the heterogeneity in the criteria used to measure burnout, and there is still no consensus on the best way. The CBI, developed to overcome criticisms of the widely used Maslach Burnout Inventory (13) and being freely obtainable, was used in the present study. The burnout rates were considerably high in all domains, and ITL scores significantly increased among these residents experiencing burnout. However, among the three domains, only the work-related burnout domain was found to be an independent predictor of residents’ ITL. Extended working hours and night shifts are reported to elevate stress, fatigue, and risk of burnout (22). Lin et al. (23) also showed that burnout rates increase 1.5 times faster after reaching 60, 74, and 84 hours of work per week in healthcare professionals. Similarly, ITS scores were higher in residents working more than 80 hours per week. Long, heavy working hours that usually involve inflexible schedules, the presence of numerous and irregular shiftwork including weekends, a stressful hospital environment, emotional and economic issues, and patients’ and patient-related issues all contribute together to feelings of burnout that increase the risk of ITL among the residents.

Job satisfaction is an important feeling that strengthens commitment and plays a role in physician retention (24). It is reported to mediate between the workplace and workers’ ITL (25). This study confirmed that job satisfaction is negatively correlated with ITL among resident physicians, consistent with previous studies. Satisfaction with the institution and area of specialization was associated with lower total ITS scores. In particular, job satisfaction is one of the independent predictors of turnover intention in these young physicians. These findings show that residents’ welfare and job satisfaction must be improved to prevent resident turnover. Increasing physician satisfaction rates is not difficult as physicians derive high internal satisfaction levels due to intrinsic rewards such as societal respect and the fulfilment derived from positively impacting lives (26).

Family work is a form of inter-role conflict that describes the conflict arising from family obligations that disturb one’s work. It occurs when the pressure from the family and work domains is mutually incompatible (27). According to scarcity theory, a person’s resources are limited, and time devoted to the requirements of one role makes it difficult to fulfil the requirements of another. Therefore, spending more energy, and time on family roles necessitates spending less on work (28). Marital obligations, responsibilities for children and childcare, and economic issues are family obligations. Family time commitment and family-related emotional distress are shown to be positively associated with FWC (29). This conflict between family and job demands is found to be another independent predictor of ITL among the residents. The residents devote more energy to their family obligations and want to quit their jobs when there is a high incompatibility between family and work.

This study is the first that assesses the prevalence and factors associated with ITL among resident physicians in Türkiye. The response rate was high. Reliability analysis of the included inventories indicated they were all reliable with high internal consistency coefficients (Cronbach’s alpha >0.8).

Study Limitations

However, some limitations should be addressed. The study is a single-centre study, which is the main limitation limiting the generalizability of the results. The study’s cross-sectional design is another potential limitation as absolute causal effects cannot be assessed, and findings cannot be generalized. The data were obtained with self-reported questionnaires and could not be cross-validated. Responses might have been influenced by method bias. Therefore, absolute objectivity cannot be claimed.

CONCLUSION

This cross-sectional study describes the prevalence of ITL among resident physicians working in a government university hospital and identifies factors associated with this intention. A large percentage of the residents reported their ITL their current jobs. This intention is related to many factors, including marital status, having children, compulsory service history, working hours, number of shifts at the weekends, income status, institution satisfaction, satisfaction with the area of specialization, and personal burnout scores. None of these factors independently predicts the ITL in the residents. Instead, one by one, the combination of these factors, which are components of total job satisfaction, FWC and work-related burnout, predicts ITL the current job among these physicians.

This research indicates that organizational and work-design interventions should be implemented to improve job satisfaction among resident physicians. Public health policymakers should consider the underlying factors and take measures that will positively affect the satisfaction of this population. Improving work conditions and implementing family-friendly practices to enhance the appeal of the hospital workplace are essential to better retain these physicians.