Junior doctors’ workplace well-being and the determinants based on ability–motivation–opportunity (AMO) theory: Educational and managerial implications from a three-year longitudinal observation after graduation

Abstract

Purpose

Junior doctors function as trainees from an educational perspective and as employees from a human resource management perspective. Employing the ability-motivation-opportunity (AMO) theory as a conceptual framework, this study longitudinally investigated the factors affecting the workplace well-being and career progression of junior doctors over a 3-year period following their graduation from medical schools.

Materials and methods

This 3-year prospective cohort study enrolled junior doctors who graduated from 2 medical schools in June 2019 in Taiwan. This study collected data by implementing web-based, self-administered structured questionnaires at 3-month intervals between September 2019 and July 2022. The collected data encompassed ability indicators (i.e. academic performance and perceived preparedness for clinical practice), motivation indicators (i.e. educational and clinical supervision), opportunity indicators (i.e. clinical unit cultures), and workplace well-being indicators (i.e. burnout, compassion satisfaction, and job performance). A total of 107 junior doctors participated, providing 926 total responses. The data were analysed using univariate analyses and structural equation modelling with path analysis.

Results

Over the 3-year period following graduation, the junior doctors’ confidence in their preparedness for clinical practice and the educational and clinical supervision had varying degrees of influence on the junior doctors’ workplace well-being. The influence of clinical unit cultures, which can provide opportunities for junior doctors, became evident starting from the second year postgraduation; notably, unit cultures that emphasised flexibility and discretion played positive and critical roles in enhancing the junior doctors’ workplace well-being lasted to the third year.

Conclusions

Our findings provide insights into the distinct critical factors that affect the socialisation of junior doctors within workplace environments over 3 consecutive years. These findings can provide guidance for medical educators and healthcare managers, helping them understand and support the progressive integration of junior doctors into their work environments.

Keywords:

• Junior doctors
• medical education
• healthcare human resource management
• workplace socialisation
• burnout
• compassion satisfaction
• job performance
• ability-motivation-opportunity (AMO) perspective

Practice points

• The critical factors across the progress of the junior doctors’ medical career on their workplace well-being were identified based on AMO theory.

• The junior doctors’ preparedness for clinical practice upon graduation was associated with their well-being throughout the 3 years following their graduation, especially for their self-reported job performance.

• The motivation factors of educational and clinical supervision during workplace training were associated with junior doctors’ well-being to varying extents throughout the 3 years following their graduation.

• The workplace cultures of flexibility and discretion play critical roles on junior doctors’ well-being at their later medical career (after 2–3 years of graduation).

Introduction

The transition from the learning environment of medical school to the realm of clinical practice constitutes a crucial stage in the professional socialisation of junior doctors. This involves a shift from the role of medical student to that of doctor. Junior doctors must engage with their clinical practice communities, accept their roles, and strive to become members of these communities (Cope et al. 2000). However, junior doctors typically experienced poorer mental health than senior doctors do (Tong et al. 2012), resulting in absenteeism from duty, inefficiency in daily activities, and general health problems (Abdulghani et al. 2015; Monrouxe et al. 2017) and a higher likelihood of being charged with suboptimal patient care practices (Toral-Villanueva et al. 2009). Despite the presence of these challenges for junior doctors, one survey for doctors and medical students by the British Medical Association revealed that junior doctors were the least likely to be aware of how to access help or support (Bhugra et al. 2019).

The effectiveness of medical education programmes is associated with the quality of their results, that is, the association between education and practice. However, studies on medical education have primarily evaluated the results of medical school education. Investigating the development of medical students and doctors solely from the perspective of school education does not fully account for the work and life conditions of junior doctors after they graduate and begin clinical practice. From an organisational management perspective, junior doctors in medical institutions play the roles of trainees, practitioners delivering medical care, and members (i.e. employees) of an organisation. Because junior doctors’ experiences are affected by the human resource management (HRM) practices within medical institutions (Wilson and Myers 1999), their performance and well-being must be considered at the workplaces (Ogbonnaya and Messersmith 2019). The ability-motivation-opportunity (AMO) theory delineates 3 essential elements or designs of HRM systems—The ability aspect (A) encompasses the abilities and skills required for employees to align with the objectives of organisations and reflects the effectiveness of medical school education. The motivation aspect (M) involves job support systems designed to enable employees to strive for success and effectively apply their skills. The opportunity aspect (O) of HRM practices involves enhancing employee engagement by guiding employees towards developing values that are in line with those of the workplace and fostering a greater sense of responsibility for their work and organisational identification (Jiang et al. 2012; Ogbonnaya and Messersmith 2019).

By employing the AMO theory as a conceptual framework, therefore, the present study evaluated the workplace well-being of junior doctors and explored the factors influencing the workplace well-being of junior doctors, tracing their career progression over the 3 years following their graduation from medical school. Our study findings can be used to differentiate the critical factors influencing the workplace socialisation of junior doctors across consecutive career years. These findings can serve as a guide for medical educators and healthcare managers, offering insights into the progression of junior doctors’ workplace socialisation and enabling educators and managers to offer step-by-step assistance.

Theoretical framework and hypothesis development

Workplace well-being, a comprehensive concept that involves the overall quality of work experiences, has substantial implications across the individual, organisational, and societal levels (Schulte and Vainio 2010). Given the broad scope of workplace well-being, studies have identified numerous indictors for junior doctors, including psychological well-being factors, such as mental health (Bartholomew et al. 2022), stress and burnout (Zhou et al. 2020), and anxiety and depression (Mason et al. 2016). Studies have also employed the indicators of self-reported confidence (Bryant et al. 2021), competence (Mason et al. 2016), and performance (Sladek et al. 2019) in clinical practice as well as job satisfaction and motivation (Mason et al. 2016) and workforce retention (Lock and Carrieri 2022).

It was argued that the structural and systemic changes required to develop a workforce fosters well-being in junior doctors and mitigates institutional obstacles that impede the implementation of well-being strategies (Hobi et al. 2022). The AMO theory outlines 3 essential elements or designs within HRM systems (Jiang et al. 2012; Ogbonnaya and Messersmith 2019). The ability aspect (A) of HRM practices involves recruitment, selection, and employee training. Considering the ability aspect ensures that employees possess the abilities required to meet organisational performance requirements. Recruiting and selecting new employees on the basis of their skills can lead to the hiring of employees with superior working abilities (Gong et al. 2009; Jiang et al. 2012). For example, grade point average (GPA), which serves as a composite measure of ability and performance during medical school, was positively associated with improved workplace performance among junior doctors (Carr et al. 2018). Moreover, it was found that the self-reported preparedness of junior doctors as they transition from medical school to clinical settings has implications for both patient safety and the well-being and development of junior doctors’ professional identity (Chan et al. 2023). A systematic review and meta-analysis of trainee physician studies revealed that low self-efficacy was associated with an increased likelihood of burnout and stress (Zhou et al. 2020). Therefore, we proposed our first hypothesis:

Hypothesis 1: Junior doctors’ ability, as the ability aspect of HRM practices in the healthcare organisations (i.e. academic performance and perceived preparedness for clinical practice upon graduation), was associated with their well-being during workplace training.

The motivation aspect (M) of HRM practices refers to job support systems designed to enable employees to strive for success and effectively apply their skills. Practices such as establishing a system for promotion and career development can encourage employees to work towards organisational goals (Jiang et al. 2012). The transition from senior medical student to working safely and effectively as a new junior doctor was described as one of the primary challenges encountered by new graduates (Dalgaty et al. 2017). High-quality supervision is a key factor in effective clinical training (Dolmans et al. 2002) and a method of providing monitoring, guidance, and feedback to junior doctors to ensure their successful personal and professional development (Kilminster and Jolly 2000). Effective support structures in the workplace can also counteract the effects of negative role stressors and demanding, socially isolating work schedules (Brown et al. 2010). Educational supervision and clinical supervision are formal components of postgraduate medical education and provide support for trainees throughout their training (Patel 2016). Educational supervision refers to ‘the provision of guidance and feedback on matters of personal, professional, and educational development in the context of a trainee’s experience of providing safe and appropriate patient care’ (Kilminster and Jolly 2000, p. 828), whereas clinical supervision refers to ‘an exchange between practising professionals to enable the development of professional skills’ (Butterworth and Faugier 1992, p. 12). An integrative review revealed that the level and type of supervision in the clinical environment were critical factors that influenced the capacity of junior doctors (Callaghan et al. 2017). Therefore, we proposed our second hypothesis:

Hypothesis 2: The motivation aspect of HRM practices in healthcare organisations (i.e. educational and clinical supervision) was associated with junior doctors’ well-being during workplace training.

The opportunity aspect (O) of HRM practices involves to enhance employee engagement (Ogbonnaya and Messersmith 2019) through strategies such as empowering employees with job resources, offering flexible job designs, organising work teams, and sharing information to achieve organisational goals (Jiang et al. 2012; Perreira et al. 2018; Subramony 2009). Organisational culture refers to a set of values and beliefs adopted by an organisation that serve as guidelines for its members (Schein 1985). Organisational culture can be considered a job resource (Lopez-Martin and Topa 2019), and studies have revealed that it enhances workforce well-being with respect to health, job involvement, effort, satisfaction, productivity, and ultimately, quality of care (Brown and Leigh 1996; Hsiung et al. 2021; Katz et al. 2019). Moreover, organisational culture can be explored at multiple levels (Hofstede et al. 1990; Keeton and Mengistu 1992; Martin 1992) and across professional and occupational subgroups (Davies et al. 2000; Schein 1985) within a healthcare organisation as subcultures. Junior doctors undergo a process of socialisation into different professional cultures as they rotate through individual clinical specialty units. Therefore, we proposed our third hypothesis:

Hypothesis 3: The opportunity aspect of HRM practices in the healthcare organisations (i.e. clinical unit cultures) was associated with junior doctors’ well-being during workplace training.

Methods

We conducted a 3-year survey study. This study was approved by the Research Ethics Committee (NTU-REC No.201811ES026).

Study participants and data collection

The term ‘junior doctors’ in our study refers to postgraduate year 1, 2, and 3 medical doctors (Sekhar et al. 2021). In Taiwan, medical graduates in their first 3 years of practice as junior doctors are referred to as residents training in general medicine, following specialty medicine. We implemented a 3-year prospective cohort study of junior doctors who graduated from 1 of 2 undergraduate medical schools in June 2019. In August 2019, we sent study recruitment letters to junior doctors who graduated in June 2019 by using e-mails provided by the medical schools. Of the 478 junior doctors, 145 (30%) agreed to participate in the study and provided written informed consent.

The participants first completed a background survey with a validated, structured, self-administered questionnaire that collected information on their personal background and ability indicators (A), including their academic performance and perceived preparedness for clinical practice within the AMO framework. Subsequently, a series of web-based, self-administered, structured questionnaire surveys were distributed at 3-month intervals over a 3-year period, with the survey concluding in July 2022. These surveys were used to gather data related to motivation indicators (M), namely, educational and clinical supervision, and opportunity indicators (O), namely, clinical unit cultures, within the AMO framework. The surveys were also used to collect data on the junior doctors’ well-being within their clinical workplaces, including data regarding their burnout, compassion satisfaction, and job performance. The participants were permitted to autonomously complete individual routine surveys. Those who provided one background survey and at least 5 responses to our routine surveys over a span of more than 1.5 years were considered eligible for inclusion in our analysis. Finally, we included 926 responses from 107 junior doctors in our study. Table 1 summarises the personal and undergraduate academic backgrounds of the participants. The 107 junior doctors comprised 67 men (63%), which was comparably representative to the sex percentage (male: female ≒ 2:1) of the 1,597 graduates from undergraduate medical schools in June 2019 in Taiwan (Taiwan Medical Association 2019).

Table 1. Personal and academic backgrounds of junior doctors (N = 107).

Variables	Mean	SD	Freq	%
Personal Background
Sex
Male			67	63
Female			40	37
Age^a (years)	25.552	1.387
Marital status^a
Single			105	98
Married			2	2
Financial well-being^a (self-reported 5-point scale with endpoints ranging from 1 to 5)	3.333	0.583
Religion^a
None			71	66
One or more			34	32
Academic background
Medical school
A school			62	58
B school			45	42
Medical programme (years of study)
7 years			51	48
6 years			56	52
Academic performance (score 0 to 100; 60 = minimum passing score)
Score 70 to 79			10	9
Score 80 to 89			76	71
Score 90 and higher			19	18
Preparedness for clinical practice upon graduation^b (self-reported 5-point scale from 1 to 5)	3.715	0.506

Note. ^aPersonal background information was subject to change over the course of the longitudinal surveys and was documented during each survey time. ^bForty-five items were measured using a 5-point scale to measure junior doctors’ clinical preparedness upon graduation and one composite score was generated by the mean of the 45 items showed in Table A1.

Measures

Ability (A) from AMO framework

This study measured the junior doctors’ academic performance and perceived preparedness for clinical practice as ability indictors within the AMO framework. Academic performance refers to junior doctors’ official scores of the formal curricula at medical schools ranging from 0 to 100. Medical students in Taiwan must achieve a minimum score of 60 to graduate and qualify for completing the national licensing examination. We requested that the junior doctors extract data from their medical school transcripts. The provided scores were categorised as follows: 90 and higher (coded as 4), 80 to 90 (coded as 3), 70 to 79 (coded as 2), and 60 to 69 (coded as 1).

The junior doctors’ perceived preparedness for clinical practice was measured using an adapted version of a questionnaire by Morrow et al. (2012)’s 53 items designed to capture the perceptions of graduates from medical schools covering the dimensions of complex communication, practical procedures, self-direction, professionalism, multi-professional working, paperwork management, examination skills, clinical judgement, professional development, leadership skills, and respiratory care. The items are rated on a 5-point scale, with 1 indicating not at all prepared and 5 indicating fully prepared (Morrow et al. 2012). Because the instrument was originally developed for graduates of medical schools in the United Kingdom, we recruited interns at a medical centre in Taiwan and subjected the 53 items to factor analysis in our previous project (Lin et al. 2020). We retained the 45 items that had factor loadings of ≥0.4 that were contextually suitable for the Taiwanese environment. A questionnaire comprising the 45 items was then administered to the junior doctors in the present study. To evaluate construct validity, this study implemented a two-level confirmatory factor analysis using an asymptotically distribution-free bootstrap model (Al-Nasir and Robertson 2001; Yung and Bentler 1996), and the results indicated a moderate model fit. The Cronbach’s α was .967. Detailed information is presented in Supplementary Table A1. The average scores for the 45 items were calculated to generate a composite score representing the collective perceptions of the junior doctors, which was used in further analysis.

Motivation (M) from AMO framework

This study measured educational and clinical supervision to junior doctors during their workplace training as motivation indicators within the AMO framework.

Educational supervision in the context of junior doctors can be understood to be a structured process involving mentoring dialogues (Patel 2016). The present study focused on clinical mentors assigned to junior doctors. In Taiwan, clinical mentors are hospital and community mentors who are assigned across clinical training and practice facilities during junior doctors’ postgraduate years (Taiwan Ministry of Health and Welfare 2017). Overall perceived training support from clinical mentors was measured using 15 items developed by Scandura and Ragins (1993), which assessed psychosocial support, career development, and role modelling, rated on a 5-point scale with 1 indicating strongly disagree, 2 indicating disagree, 3 indicating neutral, 4 indicating agree, and 5 indicating strongly agree. It was the first time we applied this instrument exploring junior doctors’ perceived supports from their mentors in Taiwan. Exploratory factor analysis was implemented, and 2 common factors were identified, labelled as ‘relation-oriented,’ comprising 8 items, and as ‘profession-oriented,’ comprising 7 items. Construct validity was then evaluated by implementing confirmatory factor analyses with the maximum likelihood method, and the results revealed a good model fit. The subconstructs of ‘relation-oriented’ and ‘profession-oriented’ mentoring both had a Cronbach’s α value of .939. The factor scores (i.e. 2 composite scores) for two subconstructs of ‘relation-oriented’ and ‘profession-oriented’ mentoring during the junior doctors’ clinical training were calculated using regression methods, and the results were used in subsequent analysis.

Clinical supervision is completed by clinical supervisors who facilitate collaboration between the medical workforce and the medical education framework to enhance the development of junior medical staff (Dwyer et al. 2011). The present study focused on clinical teachers assigned to junior doctors. For junior doctors, clinical teachers may be hospital clinical teachers, community and health centre teachers, geriatric medicine teachers, and any educator involved in clinical practice training across clinical training and practice facilities in Taiwan (Taiwan Ministry of Health and Welfare 2017). Overall perceived training support from clinical teachers was measured using 16 positive characteristics of role model clinicians related to clinical attributes, teaching skills, and personal qualities from Burgess et al. (2015)’s qualitative study in the Sydney Medical Programme. The items are measured on a 5-point scale, with 1 indicating strongly disagree, 2 indicating disagree, 3 indicating neutral, 4 indicating agree, and 5 indicating strongly agree. Exploratory factor analysis was implemented for the 16 question items for the pioneer testing for junior doctor population in this study, resulting in the identification of 2 common factors, labelled as ‘teaching attitude and skills,’ comprising 10 items, and as ‘clinical practice modelling,’ comprising 6 items. Confirmatory factor analyses using the maximum likelihood method were performed for 2 individual subconstructs and revealed good model fits. The subconstructs of ‘teaching attitude and skills’ and ‘clinical practice modelling’ had Cronbach’s α values of .940 and .870, respectively. Detailed information is presented in Supplementary Table A2. The factor scores (i.e. 2 composite scores) for two subconstructs of ‘teaching attitude and skills’ and ‘clinical practice modelling’ during the junior doctors’ workplace training were calculated using regression methods, with the results used for further analysis.

Opportunity (O) from AMO framework

This study investigated clinical unit cultures as opportunity indicators to gain insights into how the routine practices and management within clinical units influence junior doctors with consideration of their roles as trainees and employees. The measure of clinical unit cultures was adapted from the Organisational Culture Assessment Instrument (OCAI), which is grounded in the Competing Values Framework (CVF) for effective organisations (Cameron and Quinn 1999). The measure comprised 24 items covers the dimensions of dominant characteristics, leadership, employee management, organisational glue, strategic emphases, and criteria for success, scored on a 5-point scale, with 1 indicating strongly disagree, 2 indicating disagree, 3 indicating neutral, 4 indicating agree, and 5 indicating strongly agree (Cameron and Quinn 1999). This means of CVF-based culture classification was considered relevant to the present study because of the competitive and changing environment, resource limitations, organisational bureaucracy, and the presence of teamwork or multidisciplinary attributes in the healthcare sector (Lin et al. 2012). Because the primary focus of junior doctors’ training is their role as a foundation employee within a clinical unit, in the present study, the questionnaire items were designed to target the clinical units where they were undergoing medical training. Because the individual junior doctors’ ratings of the OCAI reflected their subjective ‘feelings’ based on their experiences regarding the clinical units where they received training, this study measured the concept of ‘clinical unit climates’ rather than ‘clinical unit cultures’ (Glisson and James 2002; Schneider et al. 1996).

This study investigated the junior doctors’ perceptions of the unit cultures at their clinical workplaces by implementing exploratory factor analysis for the pioneer testing for junior doctor population in Taiwan. This study retained factors with loadings greater than 0.4 and carefully considered the meanings of items, and 2 common factors were identified. The factors were ‘flexible and discretion-oriented’ unit cultures, comprising 10 items, and ‘stable and control-oriented’ unit cultures, comprising 11 items. Confirmatory factor analyses with the maximum likelihood method were performed and the results revealed good model fits. The subconstructs of ‘flexible and discretion-oriented’ and ‘stable and control-oriented’ unit cultures had Cronbach’s α values of .948 and .926, respectively. Detailed information is presented in Supplementary Table A3. The factor scores (i.e. 2 composite scores) for ‘flexible and discretion-oriented’ and ‘stable and control-oriented’ unit cultures during the junior doctors’ workplace training were calculated using regression methods. To prevent potential multicollinearity between these factor scores, we established a factor score of 0 as the cutoff value and employed a binary coding scheme for further analysis. A factor score > 0 was coded as ‘1 (high),’ and a factor score < 0 was coded as ‘0 (low).’ Detailed information is presented in Table 2.

Table 2. Junior doctors’ perceived motivation, opportunities, and well-being during workplace training: Comparisons of individual constructs over 3 years.

Constructs and variables / Mean [Freq (%)]	Item #	Year (1) (n = 322)	Year (2) (n = 404)	Year (3) (n = 200)	ANOVA^a
Motivation (M)
Educational Supervision
Relation-oriented	8	3.415	3.192	3.130	Year (1) > (2), (3)
Profession-oriented	7	3.781	3.618	3.618	Year (1) > (2), (3)
Clinical Supervision
Teaching attitude and skills	10	3.930	3.872	3.842	NS
Clinical practice modelling	6	4.193	4.122	4.095	NS
Opportunity (O)
Flexible and discretion-oriented unit culture	10	3.521	3.465	3.464	NS
Low		[147 (46)]	[178 (44)]	[93 (47)]
High		[175 (54)]	[226 (56)]	[107 (54)]
Stable and control-oriented unit culture	11	3.749	3.674	3.608	NS
Low		[133 (41)]	[175 (43)]	[95 (48)]
High		[189 (59)]	[229 (57)]	[105 (52)]
Well-Being
Burnout	9	2.655	2.778	2.820	Year (1) < (2), (3)
Compassion satisfaction	10	3.519	3.364	3.328	Year (1) > (2), (3)
Job performance	4	3.848	3.786	3.736	NS

Note. See Tables A2–A4 for detailed information on each item. Items were rated using a 5-point scale. The n’s indicate the number of survey responses for individual years.

^a Comparisons are presented in columns, with statistical significance indicated at p < .05; NS indicates no statistically significant difference between factors over 3 years.

Junior doctors’ well-being at clinical workplaces

The junior doctors’ well-being was evaluated by assessing their burnout, compassion satisfaction, and self-reported job performance during workplace training.

Levels of burnout and compassion satisfaction among the junior doctors were measured using the Professional Quality of Life Scale, Version 5 (ProQoL-5; Stamm 2009), which has been increasingly employed in studies on physicians (e.g. Bhutani et al. 2012; Buselli et al. 2020; Carmassi et al. 2022; El-Bar et al. 2013; Ghazanfar et al. 2018; Gleichgerrcht and Decety 2013, Latsou et al. 2022; Low et al. 2018; Varrasi et al. 2023; Wu et al. 2017). The ProQoL-5 is used to evaluate the negative emotions associated with perceived difficulties and feelings of hopelessness in managing work tasks, which are indicative of burnout. In addition, it is used to assess the positive feelings associated with a person’s ability to help others, which are termed compassion satisfaction. Ten items are used to evaluate each dimension (Stamm 2010). The items are scored using the following 5-point scale: 1 (never), 2 (seldom), 3 (sometimes), 4 (often), and 5 (always) (Stamm 2010). High scores indicate high perceived burnout or compassion satisfaction. We implemented exploratory factor analysis of burnout with 10 items and deleted 1 item with a factor loading smaller than 0.4 for the pioneer testing for junior doctor population. Confirmatory factor analysis of the 9 items using the maximum likelihood method further revealed a good model fit. In addition, compassion satisfaction with 10 items was assessed using exploratory factor analysis for the pioneer testing for junior doctor population, followed by confirmatory factor analysis with the maximum likelihood method. The results of both indicated a good model fit. The Cronbach’s α values were .821 for burnout and .955 for compassion satisfaction. Detailed information is presented in Supplementary Table A4. The factor scores (i.e. 2 composite scores) for burnout and compassion satisfaction during the junior doctors’ workplace training were calculated using regression methods, with the results used for further analysis.

The junior doctors’ job performance was measured using 4 items adapted from the General Nordic Questionnaire (QPSNordic) developed by Dallner et al. (2000). The questionnaire was validated for diverse work organisations. The selected items were used to evaluate the junior doctors’ proficiency in their work, including the quality of their work output, the quantity of accomplished tasks, the doctors’ problem-solving capabilities at work, and the doctors’ ability to maintain positive relationships with coworkers in a work setting. The items were measured on a 5-point scale (1: very little or not at all; 2: quite a little; 3: somewhat; 4: quite a lot; 5: very much). Exploratory factor analysis was implemented, which resulted in the identification of 1 common factor for the pioneer testing for junior doctor population. Confirmatory factor analysis with the maximum likelihood method was performed and the results indicated a good model fit. The Cronbach’s α value was .883. Detailed information is presented in Supplementary Table A4. A factor score of the junior doctors’ job performance as a composite score during their workplace training was calculated using regression methods for further analysis.

Personal and medical professional backgrounds of junior doctors

Aspects of the junior doctors’ personal backgrounds that were considered to potentially be associated with their well-being were incorporated as control variables in this study. These variables comprised sex (Bartholomew et al. 2022), age (Carr et al. 2018; Oskrochi et al. 2016; Sim et al. 2004), marital status (Sim et al. 2004), religion (Chan and Huak 2004; Phua et al. 2005), and financial well-being (Revythis et al. 2021; Zhou et al. 2020). Because medical schools in Taiwan underwent a reform that reduced the duration of the academic programme from 7 years to 6 years in 2013, 2 cohorts of medical students graduated from 1 of 2 academic programs of medical schools in June 2019. This unique circumstance enabled us to investigate 2 distinct cohorts enrolled in 2 academic medical systems through a natural experiment within the realm of medical education in Taiwan. Therefore, a variable denoting the year of study was established to distinguish the junior doctors who graduated from the 7-year or 6-year academic systems. In addition, another variable was created show the medical schools where the junior doctors studied (coded as Medical School A and Medical School B) in this study. Detailed information is presented in Table 1.

Statistical analyses

Descriptive analyses were implemented to analyse all studied variables. Means and standard deviations were calculated for continuous variables, and frequencies and percentages were calculated for categorical variables. If variations were evident in the junior doctors’ repeated responses regarding burnout, compassion satisfaction, and job performance over the 3-year observation period, 2-level or multilevel analysis was considered to be an option for data analysis. However, the intraclass correlation coefficients (ICCs) for the junior doctors’ burnout (ICC = .003), compassion satisfaction (ICC = .004), and job performance (ICC = .003) during the individual surveys were <.05, which is the minimum cutoff point for initiating multilevel analysis (Bliese 2000; Cohen 1988). Thus, multilevel effects were ignored for this dataset.

The unit of analysis in this study was the 926 responses obtained from 107 junior doctors over a 3-year period. The primary objective of this study was to discern critical factors influencing the junior doctors’ workplace socialisation across consecutive career years (i.e. across the stages of career progression) so as to provide guidance for medical educators and healthcare managers to offer step-by-step assistance for junior doctors. Therefore, this study segmented the 926 responses by year, resulting in 322 responses (from 103 junior doctors) in year 1, 404 responses (from 107 junior doctors) in year 2, and 200 responses (from 63 junior doctors) in year 3.

One-way analyses of variance (ANOVAs) were implemented to assess differences in the junior doctors’ motivation (i.e. educational and clinical supervision), opportunity (i.e. ‘flexible and discretion-oriented’ and ‘stable and control-oriented’ clinical unit cultures), and well-being (i.e. burnout, compassion satisfaction, and job performance) across 3 years of clinical training. When employing multivariate analyses, studies, particularly those conducting social science and behaviour research, must consider the ratio between the number of variables and the sample size. To address this consideration, path analyses, a modified structural equation model (Stepanek et al. 2019), were implemented. Path analyses were implemented separately for each year of the study (i.e. year 1, year 2, and year 3). The standardised coefficients obtained from the path analyses were used to evaluate the effect of the exogenous variables on the endogenous variables, with statistical significance indicated at p < .05. Good model fit was evaluated against established criteria, such as chi-squared statistic (χ²/dF) value between 1 and 3, a goodness of fit index (GFI) > 0.9, an adjusted GFI (AGFI) > 0.9, a root mean square error of approximation (RMSEA) < 0.08, a Tucker-Lewis index (TLI) > 0.9, a comparative fit index (CFI) > 0.9, an incremental fit index (IFI) > 0.9, and Hoelter’s critical N > 200 (Arbuckle 2017; Hooper et al. 2008). Throughout the modelling process, this study iteratively refined the path model to ensure that the final model effectively matched the data (Xing and Yan 2022). Statistical analyses were implemented using SPSS version 25 and SPSS AMOS version 25 (IBM, Armonk, NY, USA).

Results

Personal and undergraduate academic backgrounds of the studied junior doctors

This study enrolled 107 junior doctors, comprising 67 men (63%). Of the junior doctors, 105 were single (98%), and the average age was 26 years at the beginning of the study. Among the participants, 71 had no religious beliefs (66%), and the participants had an average self-reported financial well-being rating of 3.333 (5-point scale with endpoints ranging from 1 to 5). With respect to the junior doctors’ academic medical background, 62 of the junior doctors (58%) had graduated from Medical School A, 56 (52%) had graduated from 6-year medical programmes, and most (71%) had average academic performance scores ranging from 80 to 89 (100-point scale with endpoints ranging from 0 to 100). In addition, the average score for perceived preparedness for clinical practice upon graduation from medical school was 3.715 (5-point scale with endpoints ranging from 1 to 5). Detailed information is provided in Table 1.

Junior doctors’ perceived motivation, opportunities, and well-being during workplace training over 3 years

The results of the analysis of the junior doctors’ 926 responses are presented across the 3 individual years in Table 2. Regarding the degree of motivation experienced by the junior doctors during workplace training, which was measured using a 5-point scale (endpoints ranging from 1 to 5), the findings reveal that the average score for relational-oriented educational supervision was 3.130 to 3.415 and that for profession-oriented educational supervision was 3.618 to 3.781. These scores were higher in the first year than in the second and third years (p < .05). In addition, the ratings for clinical supervision with respect to teaching attitude and skills averaged between 3.842 and 3.930, and the ratings for clinical supervision with respect to clinical practice modelling averaged between 4.095 and 4.193, with no statistically significant difference noted across the 3 years (p > .05).

With respect to opportunities for junior doctors during workplace training, as assessed using clinical unit cultures measured on a 5-point scale (endpoints ranging from 1 to 5), the results reveal that flexible and discretion-oriented unit cultures were rated from 3.464 to 3.521 on average, and stable and control-oriented unit cultures were rated from 3.608 to 3.749 on average, with no significant difference noted across the 3 years (p > .05).

With respect to the junior doctors’ well-being during clinical training, as evaluated using burnout, compassion satisfaction, and job performance measured on a 5-point scale (endpoints ranging from 1 to 5), the results revealed an average burnout rating of 2.655 to 2.82 across 3 years. These scores were lower in the first year than in the second and third years (p < .05). In addition, the average compassion satisfaction ratings ranged from 3.328 to 3.519 across the 3 years. These scores were higher in the first year than in the second and third years (p < .05). The junior doctors’ average self-reported job performance ratings ranged from 3.736 to 3.848, with no significant difference noted across the 3 years (p > .05).

Hypotheses testing examining the determinants of junior doctors’ well-being during workplace training based on AMO framework over 3 years

To test our hypotheses, this study implemented multivariate path analyses for each of the 3 years. For the first year of the junior doctors’ clinical training (Table 3), the findings reveal that several factors were associated with their reduced burnout, including the ability factor of higher perceived preparedness for clinical practice (std β = −.356, p < .001); and the motivation factors of higher levels of profession-oriented educational supervision (std β = −.109, p < .05) and clinical supervision with respect to teaching attitude and skills (std β = −.220, p < .001) and with respect to clinical practice modelling (std β = −.194, p < .001). Multiple factors were associated with junior doctors’ increased compassion satisfaction included the ability factor of higher perceived preparedness for clinical practice (std β = .366, p < .001); and the motivation factors of higher levels of relational-oriented educational supervision (std β = .128, p < .01) and clinical supervision with respect to teaching attitude and skills (std β = .196, p < .001) and with respect to clinical practice modelling (std β = .201, p < .001). In addition, multiple factors were associated with junior doctors’ improved job performance included the ability factor of higher perceived preparedness for clinical practice (std β = .384, p < .001); and the motivation factors of higher levels of clinical supervision with respect to teaching attitude and skills (std β = .165, p < .01) and with respect to clinical practice modelling (std β = .204, p < .001). Overall, the first-year path analysis modelling exhibited a good fit to the data.

Table 3. First-year path analysis of determinants of junior doctors’ well-being: Ability-motivation-opportunity framework (n = 322).

Constructs and variables / Std Estimates	Burnout	Compassion Satisfaction	Job Performance
Ability (A)
Academic performance	−0.047	0.010	0.064
Preparedness for clinical practice upon graduation	−0.356***	0.366***	0.384***
Motivation (M)
Educational Supervision
Relation-oriented	−0.066	0.128**	−0.040
Profession-oriented	−0.109*	0.093	0.073
Clinical Supervision
Teaching attitude and skills	−0.220***	0.196***	0.165**
Clinical practice modelling	−0.194***	0.201***	0.204***
Opportunity (O)
Flexible and discretion-oriented unit cultures	−0.097	0.084	0.009
Stable and control-oriented unit cultures	0.039	0.021	0.001
Control Variables
Personal Background
Age	0.048	−0.024	−0.092
Sex (default: male)	0.047	−0.201***	−0.06
Marital status (default: single)	−0.091	0.078	0.017
Financial well-being	0.044	−0.021	−0.003
Religion (default: no)	0.018	−0.064	−0.058
Academic Background
Medical programme [7 years (default) vs 6 years]	−0.059	0.095	−0.087
Medical school [A school (default) vs B school]	0.066	−0.127**	−0.034

Note. The model fit showed Bollen-Stine χ2 = 94.156, df = 93, Normed chi-squared statistic (χ²/dF) = 1.012, GFI = 0.931, AGFI = 0.873, CFI = 0.999, IFI = 0.999, TLI = 0.998, RMSEA = 0.006, and Hoelter’s N(CN)=318.553.

* p < .05, **p < .01, ***p < .001.

For the second year of the junior doctors’ clinical training (Table 4), the findings reveal that several factors were associated with junior doctors’ reduced levels of burnout, including the ability factor of higher perceived preparedness for clinical practice (std β = −.207, p < .001); the motivation factor of higher levels of profession-oriented educational supervision (std β = −.142, p < .01) and clinical supervision with respect to teaching attitude and skills (std β = −.156, p < .001); and the opportunity factors of flexible and discretion-oriented (std β = −.111, p < .05) and stable and control-oriented (std β = −.108, p < .05) unit cultures. Multiple factors were associated with junior doctors’ increased compassion satisfaction included the ability factor of higher perceived preparedness for clinical practice (std β = .198, p < .001); the motivation factor of higher levels of relational-oriented (std β = .113, p < .05) and profession-oriented (std β = .086, p < .05) educational supervision and clinical supervision with respect to teaching attitude and skills (std β = .123, p < .01) and clinical practice modelling (std β = .098, p < .05); and the opportunity factors of flexible and discretion-oriented (std β = .126, p < .01) and stable and control-oriented (std β = .147, p < .001) unit cultures. In addition, multiple factors were associated with junior doctors’ improved job performance, including the ability factor of higher preparedness for clinical practice (std β = .239, p < .001); the motivation factors of higher levels of profession-oriented educational supervision (std β = .149, p < .001) and clinical supervision with respect to teaching attitude and skills (std β = .123, p < .01) and clinical practice modelling (std β = .114, p < .05); and the opportunity factor of stable and control-oriented unit cultures (std β = .106, p < .05). Overall, the second-year path analysis modelling exhibited a good fit to the data.

Table 4. Second-year path analysis of determinants of junior doctors’ well-being: Ability-motivation-opportunity framework (n = 404).

Constructs and variables / Std Estimates	Burnout	Compassion Satisfaction	Job Performance
Ability (A)
Academic performance	0.037	0.037	0.064
Preparedness for clinical practice upon graduation	−0.207***	0.198***	0.239***
Motivation (M)
Educational Supervision
Relation-oriented	−0.071	0.113*	0.034
Profession-oriented	−0.142**	0.086*	0.149***
Clinical Supervision
Teaching attitude and skills	−0.156***	0.123**	0.123**
Clinical practice modelling	−0.072	0.098*	0.114*
Opportunity (O)
Flexible and discretion-oriented unit cultures	−0.111*	0.126**	0.056
Stable and control-oriented unit cultures	−0.108*	0.147***	0.106*
Control Variables
Personal Background
Age	0.010	0.082	0.012
Sex (default: male)	0.094*	−0.273***	−0.108*
Marital status (default: single)	−0.078	0.033	0.023
Financial well-being	−0.286***	0.245***	0.219***
Religion (default: no)	−0.056	0.024	0.030
Academic Background
Medical programme [7 years (default) vs 6 years]	−0.003	0.070	0.076
Medical school [A school (default) vs B school]	0.103*	−0.180***	−0.093*

Note. The model fit showed Bollen-Stine χ2 = 99.269, df = 94, Normed chi-squared statistic (χ²/dF) = 1.056, GFI = 0.938, AGFI = 0.887, CFI = 0.996, IFI = 0.996, TLI = 0.994, RMSEA = 0.012, and Hoelter’s N(CN)=383.198.

* p < .05, **p < .01, ***p < .001.

For the third year of the junior doctors’ clinical training (Table 5), the results reveal that multiple factors were associated with their reduced burnout, including the ability factor of higher academic performance (std β = −.111, p < .05); the motivation factors of higher levels of relational-oriented educational supervision (std β = −.220, p < .001) and clinical supervision with respect to teaching attitude and skills (std β = −.193, p < .01) and clinical practice modelling (std β = −.221, p < .001); and the opportunity factor of flexible and discretion-oriented unit cultures (std β = −.242, p < .001). Multiple factors were associated with junior doctors’ increased compassion satisfaction, including the ability factor of lower academic performance (std β = −.159, p < .01); the motivation factors of higher levels of relational-oriented educational supervision (std β = 0.258, p < .001) and clinical supervision with respect to teaching attitude and skills (std β = .141, p < .05) and clinical practice modelling (std β = 0.223, p < .001); and the opportunity factor of flexible and discretion-oriented unit cultures (std β = .235, p < .001). In addition, multiple factors were associated with junior doctors’ improved job performance, including the ability factors of lower academic performance (std β = −.184, p < .01) and higher preparedness for clinical practice (std β = .224, p < .001); the motivation factors of higher levels of relational-oriented (std β = .247, p < .001) and profession-oriented (std β = .187, p < .01) educational supervision and clinical supervision with respect to teaching attitude and skills (std β = .129, p < .05) and clinical practice modelling (std β = .204, p < .001); and the opportunity factor of flexible and discretion-oriented unit cultures (std β = .158, p < .01). Overall, the third-year path analysis modelling exhibited a good fit to the data.

Table 5. Third-year path analysis of determinants of junior doctors’ well-being: Ability-motivation-opportunity framework (n = 200).

Constructs and variables / Std Estimates	Burnout	Compassion Satisfaction	Job Performance
Ability (A)
Academic performance	0.111*	−0.159**	−0.184**
Preparedness for clinical practice upon graduation	−0.009	0.105	0.224***
Motivation (M)
Educational Supervision
Relation-oriented	−0.220***	0.258***	0.247***
Profession-oriented	−0.077	0.07	0.187**
Clinical Supervision
Teaching attitude and skills	−0.193**	0.141*	0.129*
Clinical practice modelling	−0.221***	0.223***	0.204***
Opportunity (O)
Flexible and discretion-oriented unit cultures	−0.242***	0.235***	0.158**
Stable and control-oriented unit cultures	0.048	0.102	−0.036
Control Variables
Personal Background
Age	−0.105	0.118	−0.092
Sex (default: male)	0.300***	−0.391***	−0.134*
Marital status (default: single)	−0.059	−0.041	−0.109
Financial well-being	−0.172**	0.082	0.032
Religion (default: no)	−0.128*	0.048	−0.084
Academic Background
Medical programme [7 years (default) vs 6 years]	−0.010	−0.060	−0.194**
Medical school [A school (default) vs B school]	0.137*	−0.127*	0.128*

Note. The model fit showed Bollen-Stine χ2 = 100.087, df = 93, Normed chi-squared statistic (χ²/dF) = 1.076, GFI = 0.904, AGFI = 0.823, CFI = 0.992, IFI = 0.993, TLI = 0.987, RMSEA = 0.020, and Hoelter’s N(CN)=186.197.

* p < .05, **p < .01, ***p < .001.

In addition, our study reveals several personal characteristics critically influenced the junior doctors’ well-being, with sex (Tables 3–5) and financial well-being (Tables 4–5), and education institutions (Tables 3–5).

Discussion

This prospective cohort study used the AMO framework to explore the factors affecting junior doctors’ well-being during the 3-year period after their graduation in June 2019 from 1 of 2 the studied medical schools in Taiwan. This study collected data on ability indicators (i.e. academic performance and perceived preparedness for clinical practice upon graduation), motivation indicators (i.e. educational and clinical supervision), opportunity indicators (i.e. clinical unit cultures), and well-being indicators (i.e. burnout, compassion satisfaction, and job performance). Data were collected using web-based, self-administered structured questionnaires distributed at 3-month intervals between September 2019 and July 2022. This study enrolled 107 junior doctors (926 responses) and implemented multivariate path analyses for each of the 3 years to test the study hypotheses. In accordance with the AMO framework, our study revealed that the ability factor of junior doctors’ preparedness for practice was associated with their well-being throughout the 3 years following their graduation. The motivation factors of educational and clinical supervision during workplace training were associated with junior doctors’ well-being to varying extents throughout the 3 years following their graduation. The opportunity factor of clinical unit cultures emerged as an influential factor started in the second year, and flexible and discretion-oriented unit cultures positively and critically affected the junior doctors’ well-being lasted to the third year.

Effects of ability factors on junior doctors’ well-being

The ability measures in our study were centred around academic performance, which was considered as an objective indicator evaluated by academic faculty for the academic formal curricula at medical schools. In addition, self-reported preparedness for clinical practice was considered a subjective indicator, reflecting the confidence levels of the junior doctors at the time of graduation. Notably, the junior doctors’ academic scores in medical school, which were based on formal curricula, were not significantly associated with junior doctors’ well-being (i.e. burnout, compassion satisfaction, and performance) during the first 2 years after graduation; however, they were associated with increased levels of burnout and reduced levels of compassion satisfaction and job performance at the third year (p < .05). Studies have revealed inconsistent results regarding the effects of junior doctors’ academic performance. For example, a long-term cohort study focusing on medical graduates from Johns Hopkins University (1948 to 1964) revealed that the academic research performance and experience of medical students could be used to predict their academic achievements, such as their faculty ranking and the number of citations for their published papers over a 20-year timeframe (Brancati et al. 1992). However, a survey of the career trajectories of London medical students 20 years after graduation revealed contrasting results. The study asserted that A-level grades failed to predict medical students’ subsequent pursuit of higher academic degrees, research achievements, and levels of stress or burnout (McManus et al. 2003). A study of medical students at the University of Calgary in Canada obtained similar results, contending that school performance (e.g. GPA, internship reports, and scores on the Medical Council of Canada Qualifying Examination Part 1) was a poor predictor of the first-year performance of resident doctors (Woloschuk et al. 2010). A 10-year longitudinal study involving Polish medical students revealed that students with relatively poor grades in medical school were more satisfied with their quality of life, experienced less work pressure, and had a higher income after graduation; however, they were less satisfied with their jobs. Another study asserted that students with higher grades in medical school became doctors with greater medical competencies but also experienced higher levels of anxiety and more substantial psychological work burdens (Walkiewicz et al. 2012). The present study lacks sufficient information to provide comprehensive explanations for the effects of academic performance on the career development of medical doctors. This aspect could be explored further in future research.

Our study, nevertheless, revealed that self-reported preparedness for clinical practice among newly graduated junior doctors was positively associated with levels of workplace burnout, compassion satisfaction, and job performance during the 3-year period. Junior doctors’ self-reports regarding their preparedness for clinical practice are often employed to identify areas in which further educational intervention may be required (Monrouxe et al. 2017). We propose that attention be directed towards clinical training, particularly the aspects of clinical training that received lower scores in our study, as indicated in Supplementary Table A1. In addition, it was argued that patient views play a crucial role in preparing junior doctors for practice, and fostering more dialogue between patients, medical doctors, and educators is essential to clarifying expectations and confidentiality concerns related to patient care (Kostov et al. 2018).

Effects of motivation factors on junior doctors’ well-being

In addition to the significant attributes of personal capability and confidence (i.e. ability indicators), our study highlighted the crucial role of motivation within the AMO framework, as measured through educational supervision and clinical supervision, across a 3-year period of observation.

Educational supervision, assessed through mentorship, is viewed as critical to the professional success of physicians, physicians-in-training, and junior faculty in academic practice (Gillespie et al. 2012). Our study revealed that a focus on relational-oriented mentoring yielded positive outcomes for compassion satisfaction of first-year and second-year junior doctors. Moreover, this form of mentoring became especially crucial for third-year junior doctors, significantly influencing their well-being with respect to burnout, compassion satisfaction, and job performance. The findings of the present study also reveal that the influence of profession-oriented mentoring, especially its beneficial effect on second-year junior doctors’ well-being, cannot be overlooked. Although relational-oriented mentoring continued to have an effect across the 3-year period (Tables 3–5), our study revealed a decline in junior doctors’ perceived educational supervision during their second and third years postgraduation (Table 2), deserves noted for the future.

Our study results also reveal that clinical supervision focused on teaching attitudes and skills and that focused on clinical practice modelling played critical roles in shaping the junior doctors’ well-being across the 3 years, with the roles being particularly influential in the first and third years. Notably, these factors exerted a stronger influence than educational supervision did. We posit that as junior doctors navigate new career stages such as their first year as general medicine resident trainees and their third year as specialist resident trainees, they encounter clinical practices that present new challenges (i.e. clinical work patterns). Previous study reported that the cortisol levels of junior doctors can fluctuate in response to changes in work patterns (Smith et al. 2006). We might speculate that the need for clinical supervision becomes a priority, crucially affecting and improving junior doctors’ well-being when the junior doctors initiated new professional stages at the first and third years, respectively, as PGY general medical residents and as specialty medical residents. Therefore, fostering a constructive relationship between trainees and advisers, including mentors and clinical teachers, requires dedicated engagement (Lin et al. 2015) and ongoing re-evaluation (Gillespie et al. 2012) over the course of junior doctors’ career progression. We also propose that self-reported preparedness for clinical practice (Table A1), which was identified as an ability indicator within the AMO framework could be integrated into the audits implemented by educational and clinical supervision faculty to enable ongoing guidance for and monitoring of junior doctors’ progress.

Effects of opportunity factors on junior doctors’ well-being

Regarding clinical unit cultures as opportunity indicators within the AMO framework, our study findings reveal that these cultures were not related to the junior doctors’ well-being (i.e. burnout, compassion satisfaction, and job performance) during their first year postgraduation (p > .05, Table 3). This finding could be attributed to the first-year junior doctors primarily assuming the role of trainees rather than practitioners, which might have limited their direct involvement with and exposure to the actual medical environment and led to clinical unit cultures not influencing the well-being of the first-year junior doctors.

Our results, however, reveal a noteworthy trend for the junior doctors’ second and third years postgraduation. Our study reveals that flexible and discretion-oriented unit cultures exerted potentially beneficial effects, reducing burnout and increasing compassion satisfaction among the doctors during the junior doctors’ postgraduate second-year training (p < .05, Table 4) and were associated with multiple positive outcomes in the junior doctors’ workplace life, namely, reduced burnout, increased compassion satisfaction, and improved job performance during the third-year training (p < .05, Table 5). A comparison of the standard coefficients across the determinants revealed that flexible and discretion-oriented unit cultures had more beneficial effects (i.e. relative high coefficient values) than the other determinants did on the third-year junior doctors’ levels of burnout (std. β = −.242, p < .001, Table 5) and compassion satisfaction (std. β = .235, p < .001, Table 5). It was argued that adopting management strategies focused on fostering connections, providing support, and optimising teamwork within interprofessional care teams may help address problems related to employee burnout, turnover, and isolation (Park et al. 2023). Therefore, flexible and discretion-oriented unit cultures, which are characterised by an emphasis on support through participation, cooperation, mutual trust, teamwork, and fostering innovation by promoting the pursuit of new information, creativity, openness to change, and anticipation (Van Muijen et al. 1999), might enhance junior doctors’ perceptions of organisational support so as to reduce levels of burnout and enhance compassion satisfaction during workplace training and practice. We might suggest that unit supervisors must be flexible and create a trainee-friendly workplace climate based on the trainees’ needs across their career progression.

In addition, our study results reveal that stable and control-oriented unit cultures were associated with potentially beneficial effects, such as reduced burnout, increased compassion satisfaction, and improved job performance among junior doctors during their second year of postgraduate training (p < .05, Table 4). Previous study revealed that hospital behavioural control as the principal method to manage junior doctors for operational efficiency (Cogin et al. 2016). Stable and control-oriented unit cultures were characterised as climates that prioritise adherence to rules, respect for authority, rational procedures, and the division of work tasks. In addition, these cultures emphasise goal orientation, focusing on performance, accomplishment, and accountability (Van Muijen et al. 1999). Such cultures play a crucial role in delineating the duties and roles of junior doctors, thereby alleviating stress related to unclear demands and tension (Vance et al. 2019). Nevertheless hospitals implement rules to promote adherence to established procedures, it was argued that excessive and inappropriate use of behaviour controls could restrict healthcare managers’ ability to motivate and engage their staff and lead to short-term cost-cutting objectives but not long-term effectiveness (Cogin et al. 2016). Given this perspective, the lack of association between stable and control-oriented unit cultures and the junior doctors’ well-being at the third year in our study is reasonable.

Other factors related to junior doctors’ well-being

It has been noted that employee well-being in the workplace is not determined exclusively by external conditions related to the working and organisational environment; but stemming from individual traits and behaviours (Biggio and Cortese 2013). In our study, several personal characteristics critically influenced the junior doctors’ well-being, with sex and financial well-being exerting notable influences.

Our study findings reveal that female junior doctors experienced lower levels of well-being across the 3 years after graduation. The junior doctors had lower levels of compassion satisfaction during the first year of training (std β = −.201, p < .001, Table 3) and experienced higher levels of burnout (std β = .094, p < .05, Table 4), lower levels of compassion satisfaction (std β = −.273, p < .001, Table 4), and poorer self-reported job performance (std β = −.108, p < .05, Table 4) during the second year of training, than their male counterparts did. During the third year of training (Table 5), the female junior doctors continued experiencing higher levels of burnout (std β = .300, p < .001), lower levels of compassion satisfaction (std β = −.391, p < .001), and poorer self-reported job performance (std β = −.134, p < .05) than their male counterparts did. Previous studies have revealed several potential factors that could contribute to female junior doctors having lower wellness levels. These factors encompass women differing priorities and concerns from those of male junior doctors regarding factors such as performance feedback, overtime, and weekend duties (Ochsmann 2012). In addition, there are differences in work life, family conflict (Guille et al. 2017), imbalance (Rich et al. 2016), experiences of workplace microaggression (Thum et al. 2021), macroaggressions due to misogyny (Richards and Wohlauer 2021), sex-based role misidentification (Berwick et al. 2021), instances of bullying (Terry and Williamson 2022), and sleep quality concerns (i.e. insufficient sleep, not waking up refreshed, or excessive sleepiness; Gander et al. 2010). Furthermore, female junior doctors were found to be less confident in their clinical competency (Nomura et al. 2010) and to have less autonomy while conducting operations during their training than males (Meyerson et al. 2017, 2019). Although our data do not provide a definitive explanation to female junior doctors facing greater challenges than their male counterparts do, our study’s findings, highlighting the consistently poorer well-being of female junior doctors compared with their male counterparts across postgraduate years 1 through 3, underscore the urgency of addressing sex-related concerns in medical education and healthcare management in future studies.

Our findings reveal that better perceived financial well-being among junior doctors was strongly associated with reduced levels of burnout (std β = −.286, p < .001), increased levels of compassion satisfaction (std β = .245, p < .001), and better perceived job performance (std β = .219, p < .001) during the second year of training (Table 4); it was associated with reduced levels of burnout (std β = −.172, p < .05) during the third year of training (Table 5). Financial well-being has garnered attention within research on medical education focused on medical students and postgraduate doctors because of its potential effects on factors such as specialty selection (Smith et al. 2015) and intentions to leave the profession (Lambert et al. 2018). A systematic review and meta-analysis of studies on trainee physicians revealed an association between financial anxiety or worries and an increased likelihood of burnout and stress (Zhou et al. 2020). Our data revealed effects of junior doctors’ financial status on their workplace well-being during their second year of training, which may be explained by the assignment of fewer routine salary-related tasks due to the COVID-19 pandemic (Seifman et al. 2022) or experience increasing financial needs or burdens due to personal life planning, including marriage or parenthood. These possibilities warrant further exploration.

This study recruited graduates from 2 medical schools. To account for the potential effect of educational institution, we incorporated medical school as a confounding variable in our analysis. Our findings reveal that the junior doctors from different educational institutions exhibited differences in workplace well-being over 3 years with respect to burnout, compassion satisfaction, and job performance. Studies have proposed several strategies to bridge the gap between school education and real-world workplace training, including evaluating the ability of educational institutions to effectively prepare their graduates for medical practice, as indicated by their curricula and teaching methods, through retrospective evaluations by junior doctors (Bleakley and Brennan 2011; McRae 2016; Ochsmann et al. 2011; Rotstein et al. 2020). Studies have also proposed pregraduation orientation initiatives, such as student assistantship programmes that provide hands-on opportunities for refining clinical skills and assuming responsibilities under supervision (Chow et al. 2022), as well as prevocational medical training for junior doctors provided by training institutions (Gleason et al. 2007). Furthermore, the importance of incorporating entrepreneurship, leadership, and management skills into medical education has gained recognition, particularly because of the breadth and complexity of contemporary hospital systems (Myers and Pronovost 2017). These interventions could help mitigate disparities in academic effectiveness across educational institutions and require ongoing monitoring, accreditation, and enhancement efforts.

In addition, it should be noted that our study period was from August 2019 to July 2022, during which the world experienced the global COVID-19 pandemic. However, the study context, Taiwan, was one of the most unaffected countries worldwide (Bliźniewska-Kowalska et al. 2021; Chuang et al. 2022) due to prudent action, a rapid response, and an early-stage containment policy, coupled with transparency and public trust with solidarity (Awandare et al. 2020). Examples include commencing onboard quarantine measures for all direct flights arriving from Wuhan, China; activating airport quarantine measures; wearing face masks and hand-washing to prevent infection; ensuring adequate domestic supplies of surgical masks and personal protective equipment (Feng et al. 2020); introducing digital technology incorporating big data; quarantining COVID-19 cases; and implementing traveling and gathering restrictions (Lai et al. 2023). Moreover, the Taiwan Health and Welfare Minister has hosted a daily televised press conference to provide the public with case-by-case updates on the current status of the outbreak, seeking to maximise transparency and awareness and thereby minimise public anxiety since 22 January 2020 (Wu et al. 2020). Nevertheless, the utilisation of health care utilities in Taiwan did change, including emergency and outpatient visits, inpatient hospitalisations, and specific services; for instance, HIV and cancer screening and prevention services were significantly reduced during the pandemic (Lai et al. 2023). An international survey covering the USA, China, Saudi Arabia, Taiwan, and others revealed that physician trainees’ exposure to patients with COVID-19 was significantly associated with higher burnout rates (Cravero et al. 2021). Therefore, given the AMO framework on examining junior doctors’ workplace well-being in this study, future studies across international contexts should note that the COVID-19 pandemic as a potential confounder and keep monitoring its associated impact and appropriately responses in the post-pandemic period.

Study limitations

This study has several limitations. First, the small sample size of junior doctors who graduated in June 2019 who were recruited from only 2 medical schools could limit the generalisability of our study findings to nationwide. However, our longitudinal data (926 responses in our study) met the requirements for the individual analytical years (322 responses in Year 1, 404 responses in Year 2, and 200 responses in Year 3) by 10 times the minimum amount of the parameters found in the path analysis (Klein 1998); that is, 18 (composite) variables in our study. In addition, the issue about the potential challenge of sample attritions for the panel studies should be concerned. Second, a cross-sectional design of hypothesis testing was applied individually across 3 years, which limits our ability to establish causal relationships. Third, the measurement of junior doctors’ well-being with respect to burnout, compassion satisfaction, and job performance was based on self-reported subjective indictors, which may have introduced bias. However, previous studies have argued that even when measures are subjective or self-reported, using multiple measures tends to result in enhanced consistency and alignment with objective performance measures, which may be unavailable or undisclosable (Chakravarthy 1986; Dess and Robinson 1984). Finally, we explored the junior doctors’ well-being through the AMO framework, which characterises workplaces from a positive organisational psychology perspective and focuses on factors (e.g. AMO indictors in our study) that promote the well-being and health of individuals within a workplace (Bakker and Schaufeli 2008) from an education perspective. However, Herzberg’s ‘motivation-hygiene’ theory for management in the health care sector suggests that in work environments, ‘hygiene’ factors (e.g. reasonable salary and incentives, job insecurity, physical working conditions, Byrne 2006) could also contribute to junior doctors’ feelings of being overwhelmed and dissatisfied with the workplace training. In addition, future studies should employ the AMO framework examining junior doctors’ workplace well-being employed in the study to investigate various international contexts and continue to monitor junior doctors’ professional development in the post COVID-19 pandemic times.

Conclusions

This longitudinal 3-year study traced the trajectory of junior doctors’ medical careers. Guided by the AMO theory, this study investigated the distinct and critical indictors that influence workplace well-being in the 3 years following junior doctors’ graduation from medical schools. Our findings can provide a reference for medical educators and healthcare managers, enabling them to understand junior doctors’ workplace socialisation process and provide step-by-step support and guidance to enhance junior doctors’ workplace well-being. Additional studies must investigate the specific challenges encountered by female junior doctors as they progress in their careers and undergo advancing training, with a focus on providing valuable assistance to improve their overall well-being within the medical profession.

Authors’ contributions

All authors contributed to the study design and to the analysis and interpretation of the data. BYJLacquired the data, YKL drafted the manuscript, and all authors made substantial contributions to the manuscript. All authors reviewed, commented on, and approved publication of the manuscript.

Ethical approval

This study was approved by Research Ethics Committee of National Taiwan University in Taiwan, ROC (201811ES026). All participants provided written statements of consent.

Acknowledgements

The authors are grateful for the junior doctors’ participation in our study. We also would like to thank Wallace Academic Editing (https://www.editing.tw/) and Editage (https://www.editage.com/) for English language editing.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Additional information

Funding

Taiwan Ministry of Science and Technology Council provide the funds for this study’s academic and administrative processes and publication [MOST 108-2410-H-182-011-SS3 & NSTC 111-2410-H-182-001-SS3]. The funding body played no role in the design of the study and collection, analysis, and interpretation of data and in writing the manuscript.

Notes on contributors

Yung Kai Lin

Yung Kai Lin, MD, EBMA, Chief, Department of Surgery, Jen-Ai Hospital, Taichung, Taiwan, ROC.

Chia-Der Lin

Chia-Der Lin, MD, PhD, Vice Dean, College of Medicine and Professor, School of Medicine, China Medical University, Taichung, Taiwan, ROC, Chief, Department of Otorhinolaryngology Head & Neck Surgery, China Medical University Hospital, Taichung, Taiwan, ROC.

Blossom Yen-Ju Lin

Blossom Yen-Ju Lin, PhD, Professor, Department of Medical Humanities and Social Sciences, School of Medicine, Chang Gung University, Taoyuan, Taiwan, ROC.