Working conditions in primary care: a qualitative interview study with physicians in Sweden informed by the Effort-Reward-Imbalance model

Background Many problems with primary care physicians’ psychosocial working conditions have been documented. Many studies on working condition have used the Effort-Reward-Imbalance (ERI) model, which posits that poor health and well-being may result from imbalances between the level of effort employees perceive that they put into their work and the rewards they receive. The model has not been used in qualitative research or applied to investigate primary care physicians’ working conditions. The aim of this study was to apply the ERI model to explore the perceived efforts and rewards by primary care physicians in Sweden and approaches they take to cope with potential imbalances between these efforts and rewards. Methods The study has a qualitative design, using semi-structured interviews. A purposeful sampling strategy was used to achieve a heterogeneous sample of primary care physicians who represented a broad spectrum of experiences and perceptions. We recruited 21 physicians; 15 were employed in public health care and 6 by private health care companies. Results The analysis of the interviews yielded 11 sub-categories: 6 were mapped to the efforts category, 3 were attributed to the rewards category and 2 were approaches to coping with effort/reward imbalances. Many of the statements concerned efforts in the form of high workload, restricted autonomy and administrative work burden. They also perceived resource restrictions, unpredictability of work and high expectations in their role as physicians as efforts. Three types of rewards emerged; the physicians found their job to be stimulating and meaningful, and the work climate to be supportive. The physicians coped with imbalances by means of job enrichment and using decisional latitude. Conclusions Primary care physicians perceive numerous types of efforts in their job, which is consistent with research concerning work stress and associated consequences, such as poor subjective health and well-being. There are also rewards according to primary care physicians, but the findings suggest a lack of reciprocity in terms of efforts and rewards although firm conclusions cannot be drawn since the study did not investigate the magnitude of the various efforts and rewards or the effectiveness of the approaches the physicians use to cope with imbalances. The ERI model was found to be useful to explore physicians’ primary care work and working conditions but its applicability likely depends on the type of work or professions being studied.

Introduction xed response options in a questionnaire, could yield new insights into their views on efforts and rewards in their primary care work. A qualitative approach could also facilitate new insights into how perceived effort/reward inequity might be overcome.
Addressing these knowledge gaps, the aim of this study was to apply the ERI model to explore the perceived job-related efforts and rewards by primary care physicians in Sweden and approaches they take to cope with potential imbalances between these efforts and rewards. It is important to investigate their attitudes towards primary care work, both the work itself and working conditions, to gain a better understanding of what types of changes might be needed to improve working conditions in primary care to reduce work-related health problems and to make primary care work a more appealing career option.

Study design and setting
The study has a qualitative design, using semi-structured interviews. A qualitative approach with interviews was considered relevant to gain a deeper understanding of primary care work and working conditions based on physicians' experiences and perceptions.
The Swedish health care system consists of 21 regions providing health care for the Swedish population of more than 10 million funded primarily by taxes. All residents are insured by the state with equal access to health care for the whole population. Fees are low and regulated by law [29,30].
Primary care is rst-line care in Sweden and is responsible for the delivery of basic medical treatment, preventive work and rehabilitation. There are approximately 1200 primary care units in Sweden of which 43% are privately owned. The private health care companies are usually contracted to regions and the out-of-pocket fees for their patients are equal to that of publicly funded health care [29]. Primary care units typically employ physicians, nurses, physiotherapists and psychologists, although there are variations in the composition of the workforce among different units [31].

Recruitment of participants
A purposeful sampling strategy was used to achieve a heterogeneous sample of participants [32]. We recruited (1) physicians who were employed in publicly funded primary care units and in private health companies; (2) physicians who were employed in primary care units that differed with regard to geographic location; (3) physicians who were specialists and residents in primary care; and (4) physicians who currently worked with conventional face-to-face patient consultations although they may also be active in digital consultations, which have increased rapidly in Sweden in recent years [19]. The objective of this sampling strategy was to recruit physicians who represented a broad spectrum of experiences and perceptions of relevance for exploring the working conditions in primary care.
We recruited 21 primary care physicians for the interviews 15 were employed in public health care and 6 were employed by private health care companies. To recruit participants working in public care, we contacted all 21 regions in Sweden by examining the regions' websites to identify the person who seemed to be responsible for digital consultation in the region because we wanted to involve physicians who had experience with both digital and conventional face-to-face consultations. We sent an e-mail to this person, brie y informing them about our study and asking for physicians from the region to participate. We did not receive any response from 8 regions; 4 regions agreed to participate and provided contact information for physicians who had worked with digital consultation. We approached 29 primary care physicians from the 4 regions, and 15 who ful lled the four purposive sampling criteria (see above) agreed to participate. To recruit participants working in private health care, we approached 7 private companies. Of these, 5 agreed to participate in our study. We approached 12 physicians from these companies, and 6 who ful lled the four purposive sampling criteria agreed to participate. All the participants had some experience of employment in publicly funded health care.
The research was conducted in accordance with the Declaration of Helsinki. The study was approved by the Regional Ethics Review Board in Linköping (2019 − 01910). Transcripts are stored in the authors' password-protected computers and no unauthorized persons have access to the data.

Data collection
The authors developed a semi-structured interview guide to capture the physicians' perceptions and experiences concerning their psychosocial work environment. The interview guide was assembled by the research team behind the study, based on the existing literature on psychosocial work environments.
The questions concerned the physicians' conventional face-to-face patient consultations and their work with digital consultations [19].
The questions that were analysed in this study were the following: Why did you choose the physician profession and to work as a primary care physician?
What is most important for you in your role as a physician? How do you perceive the work situation to be at your primary care unit, in terms of working conditions, workload and expectations on you? How exible do you perceive your job to be? Would you like it to be different in any way and, if so, how? What support do you receive from the management or manager? How is the collegial support and collaboration at your primary care unit? Is there anything you would like to see more or less of? How do you maintain the balance between your private and working lives?
Numerous probes and follow-up questions were also asked, e.g. what the physicians considered to be the best feature of their work, how satis ed they were with the current work environment and how their working conditions had changed over time. The aim was to capture relevant aspects of the physicians' working conditions. We pilot tested the interview guide in 2 interviews, which indicated that further questions regarding aspects of digital work needed to be incorporated into the interview guide. Despite this, the rst 2 interviews included relevant information and were therefore included in the analysis. the participant best. Before the interviews were conducted, the participants signed informed consent stating that their con dentiality was guaranteed and that no one other than the interviewer would know their identity. To the other researchers, the participant was known only by initials and other demographic, nonidentifying data. No participant withdrew participation during or after the interviews.
Only the participant and interviewer were present during the interviews to allow the participant to speak freely. The participants did not have any previous relationship with the researchers except for the 2 participants in the pilot interviews (known to HF) and one participant (known to HF). The rst 3 interviews were transcribed verbatim by HF and the remaining interviews were transcribed by a professional transcription agency. All transcripts were carefully examined by HF to ensure accuracy. The interviews took place from April to October 2019.

Theoretical framework
We used the ERI model as a framework for a qualitative directed content analysis [33] of the interviews with regard to efforts and rewards experienced and/or perceived by the physicians and their approaches to coping with imbalances that may exist between efforts and rewards. Efforts refer to job-related factors that are imposed on the employee and make work demanding, e.g. time pressure due to a heavy workload, interruptions while performing the job, a great deal of job responsibility and pressure to work overtime. Rewards can be job-related factors such as receiving adequate salary, good promotion prospects, secure employment, a position that adequately re ects a person's education and training, respect from superiors and/or other relevant persons, adequate support in di cult situations and being treated fairly at work. The ERI model posits that individuals use different approaches to cope with effort/reward imbalances, referred to as over-commitment, to modify deleterious effects on health and well-being, e.g. sacri cing a great deal for one's work and seeking approval [34].
In this study, efforts were work-related characteristics (e.g. terms, responsibilities and circumstances) that were perceived to have a negative impact on the physicians' job satisfaction, rewards were characteristics that were perceived to positively in uence job satisfaction and approaches to coping with effort/reward imbalances were personal strategies used by the physicians to improve job satisfaction. Job satisfaction is the positive and negative attitudes employees have towards their work or individual aspects of the work, encompassing both the work itself and the working conditions [35].

Data analysis
Participants' responses concerning job-related efforts and rewards and their approaches to coping with effort/reward imbalances were analysed using directed content analysis, applying the ERI model to develop the initial coding scheme [33]. All authors read all transcripts to obtain an understanding of the whole and examined the ERI model because the model provided a framework for the analysis. Each category of the ERI model (i.e. efforts, rewards and coping approaches) covered one aspect only, i.e. the requirement of uni-dimensionality [36].
In the rst step, PN coded the transcripts by identifying participants' statements that were related to one of the three aspects. The statements were grouped into meaning units (i.e. constellations of statements that relate to the same central meaning), which were assembled into sub-categories that shared content associated with any of the three ERI components.
The sub-categories were created to be internally homogeneous and externally heterogeneous and were intended to be mutually exclusive, i.e. the requirement of mutual exclusiveness [36]. Each sub-category was given a name to provide a concise description of what it refers to and a description was generated to provide information about what is meant by a given sub-category [36].
In the next step of the analysis, PN mapped each sub-category onto one of the three pre-determined categories (efforts, rewards or coping approaches). This step involved all authors reading and re ecting on the three ERI-related categories and the proposed sub-categories, including their names, descriptions and associated quotations. These ndings were discussed at several Zoom meetings (the analysis was carried out during onsite workplace restrictions due to the coronavirus pandemic) and via emails. This process continued until consensus was reached on the categories and sub-categories.
Representative quotations from participants were selected by PN and HF and were then discussed with the rest of the team before the nal quotations were agreed upon. Quotations are marked from physician #1 to physician #21 in the Results.

Results
The characteristics of the 21 participants are shown in Table 1. Seventeen of the participants were employed by primary care units in 4 regions and 11 worked in 5 different private companies. Participating regions and companies were in central and southern Sweden. Ten of the participants had received their medical training in Sweden and 11 had undergone medical education abroad.  (29) *In Sweden, resident physicians have nished medical school, possess a medical licence and for 5 years, they provide health care and are learning to become specialist physicians (in this case, specialists in primary care medicine).
The analysis of the interviews yielded a total of 11 sub-categories: 6 were mapped to the efforts category, 3 were attributed to the rewards category and 2 concerned approaches to coping with effort/reward imbalances (Fig. 1).

Efforts
High workload Physicians described their workload as high, which had a negative impact on their job satisfaction. Many also said that the work burden was uneven, with busy, intense days interchanged with calmer periods. Stressful workdays could affect them long after work was over for the day. It was even argued that the high workload was too exhaustive to work full-time as a clinician in primary care. Reasons for the high and/or uneven workload, according to the physicians, included high staff turnover that resulted in understa ng in relation to the number of patients listed at the unit and a general trend towards increased responsibilities for primary care to perform tasks previously handled in secondary care. It is so intense that you cannot bear it. But we have many who do something else besides, work at the university with teaching or research or so. [#20] There were also physicians who described their workload in a more favourable light, emphasizing that the amount of work was usually reasonable although it could still vary a great deal from day to day or with regard to longer time periods.
[ Restricted autonomy The physicians' job satisfaction was negatively in uenced by what they perceived as restricted autonomy. They described having limited decisional latitude and in uence over their work. Many of the statements speci cally concerned dissatisfaction with the regimented nature of the work and the lack of exibility it allowed. The lack of independence was attributed to the governance of health care; many physicians were negative about higher management and political levels dictating terms for primary care. Physicians expressed that they felt controlled and believed many tasks that were imposed on them detracted from their desired focus on caring for patients.
We work so damned unstructured because we cannot control ourselves. All these ideas from above make it harder to manage. The big change was 15 years ago, when 'silo governance' was introduced. Previously I was more self-governing, then they introduced the silos and they must have statistics. [#4] It's quite in exible. The schedule is set, it is often fully booked and it is di cult, especially when you have other assignments and would need to take time off. [#8] The administrative work tasks imposed on the physicians led to many complaints and had a negative impact on their job satisfaction. Their main concern was that this type of work was time-consuming for which they did not seem motivated, and it ultimately detracted from a desired patient focus. The physicians mostly spoke about the consequences of the administrative work burden, but one of them blamed New Public Management principles for this development.
I would like to spend a little less time on administration. It's a lot to sit and write letters or write medical certi cates and stuff like that. That's the part you really want to shorten. [#7] You meet patients, but then every patient also requires administration, so it is usually di cult to keep up with it. There is often too little time for the administrative tasks because we don't have enough doctors. [#16]

Resource restrictions
Job satisfaction was also negatively affected by perceived resource limitations. The physicians mentioned restrictions concerning technology, transfer of data and information as well as with the localities. They believed that these problems could have a negative impact on the quality and effectiveness of the work they perform.
We are very vulnerable to IT problems, which we had here this morning. There were several employees who could not log in at all to the system. It is di cult to catch up later. These are probably our biggest challenges, IT problems and congestion in the [primary care unit] premises. [#19] We do not have such a good structure for information transfer; a lot of mails that come in duplicate. Yes, it's di cult. [#21] Unpredictability of work The physicians' job satisfaction was negatively in uenced by di culties they perceived with regard to planning work ahead and being prepared for unexpected events that might occur. The physicians accepted the inherent "putting out res" nature of much health care, yet they were dissatis ed with the focus on the short-term as it impeded their ability to plan and perform their work as well as they would like.
When we work in health care, care in general, planning is very di cult. You notice this quite clearly because there is a lot that happens unforeseen in terms of patient ow. And we are quite vulnerable when there are fewer staff due to diseases and so on. [#3] There are both calm days and stressful days, but I would say that days are mostly stressful. If bookings are wrong, it affects the day a lot. Then it will be stressful of course. [#9] High expectations Some of the expectations the physicians associated with their role as physicians had a negative impact on their job satisfaction. Although they recognized that they should be held accountable for decisions they make, they believed it was often di cult to live up to moral, ethical and patient safety ideals under less than optimal working conditions. Managers, colleagues and patients all contributed to the high expectations the physicians felt in their work.
There is an expectation from my employer that I will do a good job and that the patients, above all, will be satis ed. There is a lot of focus on this and I think it is very good. And of course there is an expectation that I will 'produce'. [#11] There are more expectations from the patients. The patients are more well-read and I think that is good, I like that the patients have read online and that makes them better prepared, but they also can make justi ed demands. It can be perceived as pressing for some, that they know what they need. [#21]

Stimulating work content
The stimulating content of the work the physicians do positively in uenced their job satisfaction. They particularly appreciated the interesting challenges provided with the variation and breadth of tasks in primary care work, which required them to develop and use many different skills and abilities. Meeting, getting to know and following patients over time were mentioned as inspirational and satisfactory aspects of their work.

Meaningfulness of work
Physicians considered their work to be highly meaningful, which contributed positively to their job satisfaction. They recognized that the work they perform as physicians is of great importance for patients who seek primary care for help with their illnesses. Having the skills and ability to make a difference by helping patients and achieving patient bene ts was important for the physicians' sense of meaningfulness. With few exceptions, the physicians did not mention nancial aspects as being relevant for the meaningfulness of work.
The patient contact is probably the most important because that is why I became a clinician. The patient contact gives a lot back. [#14] [Most important in my work] is the patient contact, to have a valuable and good time together with the patient that is valuable for the patient. But it also rewarding for me as a doctor that I can help the person who is seeking help from me. [#18] Supportive work climate Working in a supportive climate at a primary care unit had a positive impact on the physicians' job satisfaction. Collaborating with other physicians and staff from other professions, receiving support from colleagues and interacting with and receiving feedback from patients were important aspects of the favourable work climate. The opportunity to speak informally with and ask other physicians was also appreciated. Some physicians claimed that job satisfaction was primarily due to the social relationships at work.
The best thing about my job, hand on heart, may not be the medical work itself, but it is probably this togetherness we have with my employees and colleagues and with other professions. [#3] It has become a good learning climate. There are many who are willing to share their knowledge so we can increase our overall competence level. [#19] Approaches to cope with effort/reward imbalances

Job enrichment
The physicians utilized various opportunities to enrich their job as a way to cope with imbalances between efforts and rewards, thus improving their job satisfaction. This involved initiatives to diversify their work tasks, develop new competences and take on responsibilities beyond the normal job in primary care. Such initiatives could reduce the workload because it gave them a break from regular clinical patient work in the primary care unit. Job decisional latitude Using job decisional latitude to in uence one's own work schedule, i.e. when and how much to work, provided the physicians with another means to reduce an overbearing workload and try to avoid over-commitment in their work, thus improving their job satisfaction. Several physicians described the attainment of a good work/leisure balance as a di cult struggle.
You have to ght to catch up and I'm used to it now. I can set limits, but [to maintain a decent work/leisure balance] that's tough. [#4] You have to work actively to get it [balance between work and leisure]. The job can devour all your time, that happens fast, because you get a lot of assignments all the time. So somewhere you proactively just have to prioritize well and above all try to make time for your own life so that you do not get stressed by your job. [#6]

Discussion
This study sought to explore primary care physicians' perceived job-related efforts and rewards as well as their approaches to coping with potential imbalances between efforts and rewards. We used the ERI model as a framework for the analysis of interviews with the physicians. Most of the sub-categories that emerged from the analysis could be mapped to the efforts category of the ERI model. This nding suggests a lack of reciprocity in terms of high costs and low gains, which elicits negative emotions according to the ERI model. In the long run, an imbalance between efforts and rewards at work increases the susceptibility to poor health and well-being [26], which means that our ndings provide a plausible explanation for work stress and associated consequences, such as poor subjective health and well-being, that have been shown in previous primary care research [3][4][5][6][7][8][9][10][11].
Most of the physicians lamented about their high workload although there were also those who described the work burden in more neutral terms, emphasizing that their work was also characterized by calmer periods. The overall ndings concerning the work burden are consistent with other studies that have documented problems with high workload in primary care in many countries [24,37,38]. High workload in primary care has been attributed to many factors, including ageing populations, changing disease patterns in the population and evolving societal norms and values in society, some of which have yielded higher expectations for access to primary care, improved patient experience and increased patient involvement in care decision making [19,[39][40][41][42]. The workload has also been affected by a shift in tasks from secondary to primary care, which has not always been accompanied by su cient resources. Primary care increasingly manages conditions previously handled by secondary care, e.g. palliative care and chronic disease, and patients are discharged to primary care more quickly than before [43].
Many of the physicians' statements concerned restricted autonomy and the burden of administrative work. Again, these ndings are consistent with many international studies concerning physicians' working conditions [6,[44][45][46]. Issues related to limited autonomy and administrative work burden for health care professionals have often been attributed to New Public Management (NPM) principles because physicians and other health care professionals are expected to document their work, take on administrative tasks and participate in management-led quality improvement initiatives to achieve organizational goals [24,47,48]. There has been a lively public debate in Sweden on NPM, with many physicians critiquing core NPM principles and highlighting the consequences for health care professionals [49][50][51]. In response to the criticism of NPM principles, the Swedish government has recently introduced the concept of "trust-based governance", intended to reduce the administrative burden and "letting professionals be professional" by allowing them to focus on their core activities, primarily patient work [52,53]. This initiative is new and there are no research studies on the concept yet to examine whether or how it can be realized in practice.
Three reward factors emerged from the analysis. The physicians found the content of their work to be stimulating, their job to be meaningful because their work is important for patients and the work climate to be supportive. These ndings are aligned with other studies of the physician profession in other countries, many of which have shown the relevance of physicians' personal sense of competence [54], collegial relationships [55] and patient interaction [56] for their job satisfaction.
Our ndings can be considered in relation to the Cognitive Evaluation Theory [57]. This theory proposes that autonomy, competence and social relatedness are three prerequisites for more intrinsically motivated behaviours, i.e. behaviours performed for their inherent satisfaction rather than because of external pressure or rewards [58]. Importantly, intrinsic motivation has been linked with higher job satisfaction [59][60][61]. The physicians in our study provided many statements that underscored the relevance of competence and social relatedness for their job satisfaction. However, the Cognitive Evaluation Theory [57] implies that their perceived lack of autonomy limits the extent to which their behaviours are intrinsically motivated, thus restricting their job satisfaction.
Two of the three reward factors in our study, stimulating work content and meaningfulness of the work, are consistent with so-called motivating factors in Herzberg's Two Factor Theory. The theory posits that the presence of motivators such as achievement, recognition, responsibility and advancement create satisfaction by ful lling individuals' needs for meaning and personal growth [62]. The third reward factor in our study, a supportive work climate, is considered a hygiene factor in Herzberg's theory. The presence of such factors does not necessarily build motivation. Rather, hygiene factors operate primarily to dissatisfy employees when they are not present [63].
Overall, we found the ERI model to be useful to explore physicians' primary care work and working conditions and to identify effort and reward factors as well as approaches to managing effort/reward imbalances of relevance for their job satisfaction. Most of the effort and reward factors that emerged from our analysis are in line with aspects addressed in questionnaires based on ERI. However, our study yielded few statements by the physicians about issues such as their income, employment security, job promotion prospects or whether the position adequately re ected their education and training, all of which are aspects included in ERI-based questionnaires [34]. The paucity of these aspects in our interviews suggests that the applicability of the ERI model depends on the type of work or professions being studied.
Few physicians mentioned income as a reward that contributed to their job satisfaction. We did not speci cally ask the physicians if they considered their income to be a reward or an incentive that could offset the efforts they perceived. Speaking openly about one's wage is usually considered inappropriate in Sweden. This reluctance has been attributed to the so-called Jantelagen, i.e. a widely held attitude of disapproval towards expressions of individuality or personal success [64]. Further, while physicians' income may be a motivating factor for choosing the occupation, the salary is unlikely to be a decisive factor for job satisfaction. Salary is a hygiene factor in Herzberg's Two Factor Theory, meaning that it can lead to dissatisfaction but when fully catered for, is not su cient to satisfy employees [63].
Coping with effort/reward imbalance in the ERI model is described as over-commitment, which means sacri cing a great deal for one's work and/or seeking approval [34]. Items in the ERI-based questionnaire on over-commitment concern being overwhelmed by work, having di culties switching off work and sacri cing too much for the job. Such consequences were mentioned by the physicians in our study, but not as approaches to handle effort/reward imbalances. Instead, the physicians viewed job enrichment, e.g. working with digital patient consultations, as an approach to manage such imbalances.
Another coping approach was to use decisional latitude to achieve a more reasonable workload. Both approaches reduced the physicians' total workload, enhanced their autonomy and improved their work/life balance. The results suggest that the responsibilities are placed on the physicians themselves to deal with imbalances in their work situation.
There are a number of limitations which should be considered in interpreting the ndings of this study. A qualitative approach using interviews was chosen because it enables a deeper understanding of how primary care physicians make sense of efforts and rewards as well as how they cope with potential imbalances between the two aspects. Participation in the interviews was voluntary. The study participants actively expressed their interest in participation, which implies that they were interested in the topic. The transferability of our results is limited to primary care settings in Sweden and the study ndings cannot be directly transferred to international settings. Regardless, the study results may be relevant for other settings because the sample was adequate [65].
We sought analytical generalization rather than statistical generalization by comparing ndings with comparable empirical research and relevant theories.
There are also considerable strengths to the study. The credibility of the study was enhanced by the multidisciplinary research team as this composition of researchers facilitated different perspectives on the investigated issue [32]. The team was comprised of a behavioural economist (PN), a physician (HF), a political scientist (IS), a registered nurse (KS), a behavioural scientist (CE) and a public health researcher (JS). The relatively high number of interviews (n = 21) was another strength of the study. We used quotations from 15 different participants, something which added to the transparency and trustworthiness of the ndings. The fact that the participants came from different geographic regions of Sweden and from both public and private organizations was another strength. Furthermore, both men and women of different ages were included and they differed with regard to previous experiences from primary care work.

Conclusions
This study of primary care physicians suggests a lack of reciprocity in terms of high costs and low gains, providing a plausible explanation for work stress and associated consequences that have been shown in previous primary care research. We found the ERI model to be useful to explore physicians' primary care work and working conditions and to identify effort and reward factors as well as approaches to managing effort/reward imbalances of relevance for their job satisfaction.