“Patients, not doctors, get sick”: A study of fifteen Swedish physicians on long-term sick leave

Abstract

Based on in-depth interviews with 15 physicians on long-term sick leave in the north of Sweden, the aim of this paper is to describe and understand what happened before, during and after sick leave. A constructivist grounded theory approach was used to construct a framework inductively from the stories of the physicians. As a result of hierarchical coding procedures in terms of initial, focused and theoretical coding, the circular or helical road back to well-being was constructed. This core category, formed in response to structure and to the culture of the medical profession, is constituted of five categories, including dealing with changing working conditions, resisting sick and patient role, resisting sick leave, spending time on sick leave and, finally, returning to work. Not only were the physicians expected to work very hard, but also they were expected to take charge, and did take charge, of their own treatment and rehabilitation. They reinforced these phenomena, by employing strategies to deal with high workload and to avoid sick leave. None of them later returned to full-time work. Notwithstanding that, some aspects of the culture of the medical profession are destructive; in this case, it promoted a strong will and motivation to return to work.

• Physicians
• working conditions
• sick leave
• sick role
• well-being
• grounded theory

Introduction

Many times the initial reaction of people, including members of the research community, to this study of physicians on long-term sick leave was that of bewilderment and surprise. Comments like “I didn't know they exist” or “I've never heard of that” were common. At first, these reactions puzzled me, but a brief look at previous research showed that there are gaps to be filled. During the last decade, as a result of dramatically increasing sick leave numbers in the Swedish work force, there is a growing body of research on and interest in sick leave related to working conditions (Allebeck & Mastekaasa, 2004). However, even though there is research indicating that, like for many other categories, the work situation of Swedish physicians has become characterized by high workload, high demands, less control and influence (Arnetz, 2001), the medical profession has not been the target for most of these studies. This should also be considered in relation to the fact that it is well-known that, for example, down-sizing efforts and mergers, which are common within the health care sector (Lindgren, 1999; Kivimäki, Vahtera, Pentti & Ferrie, 2000; Kullén Engström, Rosengren & Hallberg, 2002; Brown, Arnetz & Pettersson, 2003; Fältholm, 2005), contribute to stress and to risks of sickness and sick leave (Karasek & Theorell, 1990).

As I progressed with this study, however, it became clear that, also to the physicians themselves, the roles of doctor and patient do not combine. Additionally, the literature review that I eventually conducted shows that members of the medical profession do not expect to get sick themselves; that they are reluctant to accept personal illness (McKevitt & Morgan, 1997; Thompson, Cupples, Sibbett, Skan & Bradley, 2001); that they do not seek care to the same extent as other groups (Forsythe, Calnan & Wall, 1999; Töyry et al., 2000), and that self-treatment is common (Baldwin, Dodd & Wrate, 1997; Töyry et al., 2000). It also indicates that they resist being on sick leave (McKevitt & Morgan, 1997; Forsythe et al., 1999) and that they tend to work when sick (Baldwin et al, 1997; Kivimäki et al., 2001; Cupples, Bradley, Sibbett & Thompson, 2002; Tillett, 2003), unlike other categories with high income and similar status (McKevitt, Morgan, Dundas, Holland, 1997), but like other categories within the health care sector (Aronsson, Gustafsson & Dallner, 2000). Previous research shows that this is the result of perceived pressure to be at work (Töyry et al., 2000; Aronsson et al., 2000; Thompson et al., 2001; Kivimäki et al., 2001; Perkin, Higton & Witcomb, 2003) and that members of professional categories that are not easily replaced by others when absent show a higher degree of sick presence than others (Aronsson et al., 2000).

Further, notwithstanding that doctors are members of a professional category that enjoys power, status and prestige (Freidson, 1970), the medical profession, and in particular that of general practice, has undergone profound changes in its authority, autonomy, status and political power. Even though there is a debate on whether or not medical practitioners are losing their privileges (Lupton, 1997), this development has been described in terms of deprofessionalization and disempowerment (Riska & Wegar, 1993; Lupton, 1997; Einarsdottir, 1997; Pringle, 1998; Barnett, Barnett & Kearns, 1998; Nordgren, 2000). While deprofessionalization reflects changes in professional authority, monopoly of knowledge and autonomy, the concept of disempowerment, which is strongly related to deprofessionalization, can be used to describe the process of reduction of influence inside the political policy process where decisions about the health system and professional privileges are taken (Nordgren, 2000).

Another aspect that might be related to the concept of disempowerment is the level of control over work itself. According to Karasek and Theorell (1990), who developed the demand/control model, the medical profession is characterized not only by high psychological demands, but also by high levels of control. In their view, this means that physicians, although there is a discussion of deprofessionalisation, still have a high-prestige occupation carrying out so-called active jobs, receive high income, the highest psychological reward from work and the highest job satisfaction (Karasek & Theorell, 1990). It also means that high levels of control counteract the negative effects of high demands, creating good conditions for healthy work.

Recent changes in the nature of medical work, however, have raised issues of concern of lack of control over work itself (Riska, 2001), and there is reason to relate this to the erosion of the traditional power of the medical profession. Lupton, for example, argues that the deprofessionalization debate needs to move beyond its focus on macro-structural and policy issues to the everyday experiences of doctors at work (Lupton, 1997). One such area, she argues, is the nature of doctor-patient interaction. For example, she shows that doctors’ view of patients can be described in terms of a shift from the “dependent patient discourse” to “the consumerist discourse” (Lupton, 1997). She also argues, however, that this does not necessarily involve a loss of professional status or authority and that it is too simplistic to argue that patients and doctors are engaged in struggles of power. Fairclough (1997), however, argues that the doctor-patient relation can be analysed and described in terms of a shift from the “medical discourse” into the “counselling discourse”, entailing greater informality, more democratic interaction, sharing of control and reduction of power asymmetries.

As assumptions of power asymmetries have been used to analyse the doctor-patient relationship, it might be expected that doctors who become patients would have a mutual relationship with their treating doctor (McKevtt & Morgan, 1997). As the patient-doctor encounter and the power asymmetry related to it can be understood in terms of an encounter between the medical discourse and the lay discourse, or between specialized and lay knowledge (Fairclough, 1997; Gwyn, 2002), the question is what happens when both patient and doctor draw upon the medical discourse. Intuitively, it would seem that the prerequisites for reducing power asymmetry, for empowerment and models of shared decision-making and patient-centred medicine (Gwyn & Elwyn, 1999; Luftey, 2005), are better than in the “normal” encounter. Drawing on Freidson's account of Szaz and Hollander's patterns of interaction in treatment, it might be concluded that the doctor-patient relation in this case can—and ideally should be—characterized according to “the model of mutual cooperation”, as there are prerequisites for the collaborative status needed; for the patient to not merely accept the authority of the doctor, but for each to accept the other as an equal in the search for a solution to the problem (Freidson, 1970).

However, as pointed out above, doctors tend to ignore their own sickness and that of colleagues, as sickness of their own is difficult to accept; and it is the sickness of patients that is in focus (McKevitt & Morgan, 1997; Forsythe et al., 1999; Thompson et al., 2001; Cupples et al., 2002; Tillett, 2003). Not only to the public, but also to the doctors, the roles of patient and doctor seem incompatible (McKevitt & Morgan, 1997; Thompson et al., 2001). In other words, drawing on Parson's conception that sick people cannot simply be ill, they have to adopt a sick role (Parsons, 1951/1991), this adoption might be more difficult for members of the medical profession than for other groups. Further, when adopting the sick role, it is expected that sick people seek competent help for their illness and that they cooperate with attempts to get them well (Parsons, 1951/1991). Freidson describes this in terms of adopting a patient role (Freidson, 1970). Contrary to what one might believe, however, because of the doctor's insider status and access to the medical system, research suggests that being a member of the medical profession and surrounded by colleagues that could provide care, does not make it easier to seek and get care and to behave like an “ordinary patient” (McKevitt & Morgan, 1997). Thus, in spite of the prerequisites for a mutual relationship referred to above, it does not seem to emerge.

Looking back at the research on sick physicians presented above, with few exceptions most of it is quantitative, based on surveys and questionnaires aiming at generating statistical analyses on sick physicians. Although some of the studies are qualitative and relate phyzicians’ sick leave behaviour to cultural and organizational factors (McKevitt & Morgan, 1997; McKevitt et al., 1997; Thompson et al., 2001), none of them describes the whole sick leave process of physicians. Questions never answered are: what happens before and during sick leave and what happens when the physicians return to work? What strategies do the physicians employ in order to deal with the work situation, and to avoid sick leave? How do they spend their time on sick leave and what happens during rehabilitation and when they return to work?

In 2003, representatives of the County Council of Norrbotten, the provider and financer of health care in the northernmost region of Sweden, noted that, in contrast to previous years with increasingly high numbers of nurses on long-term sick leave, the number of physicians on long-term sick leave had increased faster than any other group. A study was initiated and funded by the county council, aiming at exploring and describing possible reasons for this troublesome and costly development and to generate a knowledge basis for action. I was engaged to carry out the study, and later used the data, consisting of 15 in-depth interviews with physicians on sick leave, for the analysis of the study presented in this article. In this article, therefore, applying a constructivist grounded theory approach; I will try to answer the questions presented above.

Method

Participants and procedure

At the time of the data collection in 2003 and 2004, 36 physicians in the region were on long-term (> 90 days) sick leave, but since the county council, for legal and ethical reasons, was prevented from making data on these individuals public, a strategic selection of interviewees, based on, for example, gender, specialty, age, diagnosis, etc. was not possible. Instead, the employer recruited the interviewees by sending a letter to all physicians on long-term sick leave in the region, asking whether they were interested in participating. This way of recruiting respondents raised ethical considerations, such as whether the physicians would feel forced to participate, but as the letter made it clear that participation was voluntary and that they had the right to withdraw at any time during the project, it was deemed reasonable. Fifteen of the physicians agreed to participate, and out of these, nine were general practitioners employed at primary health care centres and the remaining six were other types of specialists employed at three of the hospitals of the region. Out of the 15 interviewed physicians, six were women, and this reflects the proportion of women doctors on sick leave as well as the proportion of women physicians in the region. The ages of interviewees varied between 42 and 61.

Based on their wishes, the participants were interviewed either in their respective homes, in my office at the university, at their place of work or—in one case—by telephone. The interviews lasted from one and a half to three hours. All of the interviews were audio taped and transcribed and coded for tentative categories during initial coding, which began already during the collection of data in accordance with a grounded theory approach (Glaser & Strauss, 1967).

In-depth interviews

In the process of interviewing the physicians and when analysing their stories, results other than the ones that I was initially looking for were generated. These results were not only found in what was said, but also in how it was said, indicating that I apply a constructivist, rather than an objectivist approach to grounded theory (Charmaz, 2006). Since, even though there are no firm lines between them, an objectivist “… attends to data as real in and of themselves and does not attend to the process of their production” (Charmaz, 2006, p. 131), while a constructivist studies how participants construct meanings and actions in specific settings. This means that, not only were my previous personal and professional experiences sources of theoretical sensitivity (Strauss & Corbin, 1998), but also what happened during the interviews also contributed to my ability to, in the act of theorizing, “reach down to fundamentals, up to abstractions, and probe into experience” (Charmaz, 2006, p. 135). It also means that, during the interviews, the participants and I were engaged in a reflective process that might be labelled “action learning” (McGill & Brockbank, 2004).

Data analysis

In addition to concurrent data collection, coding and analysis, one of the major characteristics of grounded theory is hierarchical coding processes, meaning that initial or open coding is followed by focused or selective coding (Hallberg, 2006). In this case, this resulted in the establishment of a core category as well as a number of categories and subcategories (Strauss & Corbin, 1990, 1998). Additionally, in my effort to order the data of this study, it became clear that it was necessary to integrate structure and process and in order to code for actions and processes, like Charmaz, I “… have developed subcategories of a category and showed the links between them as I learned about the experiences the categories represent. The subsequent categories, subcategories, and links reflect how I made sense of the data” (Charmaz, 2006, p. 61). This means that the last coding procedure was inspired by the theoretical coding suggested by Charmaz (2006).

As a result, the core category a road back to well-being as well as its categories and subcategories were constructed inductively from the stories of the participants. For example, throughout the process, memo-writing and constant comparisons were carried out. Further, the literature review presented in the introduction was not the result of studies of previous research carried out before the collection of data. Instead, it was driven by and evolved from the analysis of the interviews. As the line between a literature review and a theoretical framework often blurs (Charmaz, 2006), the introduction above contains both a literature review and a presentation of theoretical concepts that are used later in the analysis.

In accordance with the use of literature put forward by Strauss and Corbin (1990, 1998), the initial interview schedule, based on a brief literature review contained issues like organizational changes, working conditions, professional identity and development, sick leave and rehabilitation practices. As interviews and coding were carried out concurrently, a framework containing concepts like disempowerment, demand, control, sick and patient role, and doctor-patient relationship emerged, reflecting the stories of the physicians.

Results

The core category of this study is a road back to well-being, formed in response to structural aspects, such as the changing working conditions and organization of health care, and to the culture of the medical profession. It starts with changing working conditions and the discovery of illness and ends with the return to work and practically the same working conditions. Therefore, it is not a straight road back to health and well-being and it is characterized by strategies employed by the physicians to deal with their work situation, to avoid accepting sick and patient role, to manage sick leave and finally to be able to return to work and to well-being. This means that all of the strategies employed by the physicians were the result of their wishes to be able to practise as physicians and that work itself was a prerequisite for well-being, though not necessarily for health. This is illustrated by what one of the physicians said when asked to describe what it was like to return to work: “I'm actually much better off now, after getting sick, compared to when I was healthy”.

It also means that while aspects of the culture of the medical profession are destructive, as they might lead to and prolong sick leave, other aspects promote a strong will and motivation to return to work. To the physicians in this study, who all eventually managed to return to work, though none of them on a full-time basis, being able to work again meant they had come far along the road back to well-being. Thus, the wish to continue working or to return to work or—in other words—to enter the road back to well-being is something that permeates all constituting categories and subcategories.

As shown in Table I, the core category is constituted by five categories: dealing with changing working conditions, resisting sick and patient role, resisting sick leave, spending time on sick leave and, finally, returning to work. In the following sections, I will describe these categories as well as their subcategories, also listed in Table I. Throughout the article, these categories and subcategories are illustrated by selected excerpts from the interviews.

Table I.  Table showing the core category, its categories and subcategories.

Core category	A road back to well-being
Categories	Dealing with changing working conditions	Resisting sick and patient role	Resisting sick leave	leave	Returning to work
	Increased workload	Denying sickness	Fearing loss of legitimacy	Sleeping during sick leave	Being in charge of one's own rehabilitation
Subcategories	Disempowerment	Being treated by a colleague	Working through sickness	Reflecting on one's life situation	Fighting for a lost identity
	New types of patients	Being in charge of one's own treatment	Postponing and limiting sick leave		Coping with workload and pressure
		Dealing with colliding identities			Secluding oneself from the organization

Dealing with changing working conditions

By way of introduction, the physicians were asked what initially had made them choose to become physicians and to describe their work situation. Although their stories varied, they had common features, such as the changing character of medical practice, which permeated most stories. Later in the interviews, the participants were also asked whether their work situation had been a contributory cause to sickness and sick leave. Six of them, with diagnoses such as burnout, depression or chronic fatigue syndrome, stated that work itself had been the most important factor. Six others said that their diagnoses were not only related to work, but that work had had some kind of importance, at least concerning the time of getting sick and the length of the sick leave period. In their view, their diagnoses were stress related, such as heart disease or high blood pressure. Three of those interviewed stated that work had had nothing to do with their sickness. However, regardless of to what extent the physicians thought that work had had an impact on their sickness and/or sick leave, they made clear what aspects of their work situation that did have an impact and in what ways they affected sick leave behaviour and possibilities of work rehabilitation. Dealing with changing working conditions is a category that describes these aspects.

Practically all of the participants said that they had always worked very hard, because they had always liked what they were doing. Deeply troubled, they now talked about their increased workload, which is one of the subcategories of dealing with changing working conditions. Unlike previous changes, the physicians could not control these structural changes. For example, in the region, as a result of the building of a new hospital and a changed division of labour between non-institutional and institutional care, there had been a transfer of tasks from the hospitals to primary care institutions for certain types of patients; but also a transfer of more advanced medical care from the smaller hospitals to the largest and newly built one. These structural changes had implied an increased workload for physicians within primary care, as patients (except acute cases) were now always first referred to the closest primary health care unit. In order for patients then to be referred to the hospital for more specialized care, there were much higher demands on examinations, tests and treatment before access was given to the hospital. This means that resources that even before these changes were scarce, were now becoming even scarcer, which contributed to an increased workload:

As soon as there is down-sizing, patients are sent to us, but we are not given more resources. Hospitals manage to save money by transferring tasks to the primary health care, but we cannot send our patients anywhere else.

Many of the general practitioners also said that working at a health care centre “is like working on a conveyor belt” with numerous patients arriving and without any possibility to control what kind of or how many patients that were referred to them. Nevertheless, the picture is more complicated. At the same time as the general practitioners perceived work as routine, they were able to carry out more and more complicated medical interventions because of the medical and technological development. However, this contradiction was probably more connected to the frequency of the tasks than to their complexity, an aspect that in its turn was related to the lack of physicians within primary health care:

When there is a general scarcity of doctors, it's even harder to recruit doctors to health care centres. It is more or less piecework. Everyone understands that there are a number of patients coming here every day and that they must be taken care of regardless of whether you are seven or two doctors. If there are only two of you, it means a lot more work.

For the physicians working at the smaller hospitals of the region, that had been subject to mergers, downsizing and/or a transfer of tasks, the development was also described as problematic. As a result of a centralization of institutional care to the newly built and largest hospital of the region, their work was degraded; wards were closed down, they were no longer to carry out more advanced surgery, there was a drain of competence, leaving the remaining physicians with more routine type of surgery and with smaller possibilities of recruiting and supervising medical students or young physicians in training. As one of the participants said, “Patients are still as important, but the possibilities to develop professionally are gone … I wouldn't recommend a young colleague to come here, because there are no opportunities here any more.”

In addition, at the newly built and largest hospital, resources had become scarcer and because of mergers, downsizing measures had been taken and workload increased. Among the interviewees, there was also widespread questioning of work organization and management. Bitterness, frustration and anger toward the management of the county council were exposed in many of the interviews, even though it seems that the hospital physicians, more often than the general practitioners, had been involved in internal struggles of power because of organizational changes. As one of the interviewees put it:

Doctors still enjoy very high status, but the power connected to this status is different. Doctors of today are integrated into a political and administrative structure that has reduced us to wage-earners on the conveyor-belt in a fundamentally different way. When you describe what it was like fifty years ago, when you were master in your own house, that doesn't exist today. The degree of control is so much lower.

Therefore, in addition to dealing with high workload, reduced autonomy and status, the fact that the physicians were employed by a politically governed organization meant that they also had to deal with aspects of disempowerment, rendering this into the second subcategory of dealing with changing working conditions. Political decisions resulting in a transfer of tasks from hospitals to primary health care had been taken in addition to decisions on shutdowns, mergers, downsizing or of transfer of care from one hospital to another. According to both groups of physicians, neither politicians nor management was familiar with the conditions “in the field”, which created a distance between those that take decisions and those that implement the decisions taken. Many of them said they had been very active in the development of the organization and had tried, but in most cases not succeeded, in stopping the implementation of decisions taken by politicians.

The organizational changes described above also led to changes in types of patients and degree of authority and control in relation to patients. Although work at the health care centre was described as “work on the conveyor belt”, “sicker” patients were said to be treated at both the primary health care centres and smaller hospitals because of the general medical and technological development and of the downsizing of institutional care in the region. Patients were said to suffer from more than one disease and there were more old people due to demographic changes of Sweden. A growing number of patients also suffered from psychiatric disorders, which had led to a higher frequency of diagnoses such as burnout, depression and chronic fatigue syndrome. As a consequence, a larger number of patients were on long-term sick leave, and the physicians, being the ones that certify illness, said that in the debate about this being huge societal and economic problem in Sweden, they were regarded as part of the problem and even blamed for it. The fact that many of them were sick-listed for the same diagnoses that they should treat, meant even further pressure.

Further, the doctors also talked about more demanding patients. In addition to having small possibilities to manage their own time, they also said that they had limited possibilities to control what type of patients that were referred to them, especially at the health care centres. This led to the feeling that it was “not the sickest, but the angriest”, that got access to information on new types of treatment and to care. According to the interviewees, it was common that patients “ordered” a special diagnosis or treatment, because of easily accessible information on the Internet, but also that their expectations had become unrealistic. This means that the physicians had to play the role of educator and had to spend time explaining why “knowledge” about a special treatment or cure was not applicable to the patient in question:

Of course, it takes a lot of time and energy to have to sit and explain, and inside, you fly into a temper with the person that gave them that information. If you have five such patients in a day, all of your consulting hours and perhaps even your lunch break are spent just explaining …

It should also be noted, however, that most physicians interviewed argued that, since lay knowledge in the past was often based on popular belief and even delusion, the patients’ increased level of knowledge gives a better basis for discussions.

The last category of dealing with changing working conditions is thus new types of patients.

Resisting sick and patient role

Having described in what ways the physicians were dealing with the sickness of patients, I now move forward to how they were dealing with their own sickness. Denial, resistance and fear characterize the stories. Ignoring signs of sickness and only later reluctantly admitting sickness were parts of the process. This means that the first subcategory of the category of resisting sick and patient role is denying sickness. However, even though there were elements of sometimes denying symptoms that should have been easy to diagnose, it was not always a matter of denying what was obvious to everybody else. Many of the physicians had suffered and still suffered from more than one sickness or symptom and described very complex courses of events. Very few can be described as perfectly healthy individuals suddenly falling ill, but in spite of this (or perhaps because of this) falling ill or acknowledging illness was often very dramatic. The ones that seemingly abruptly were stricken by some kind of somatic disease did not have much choice. They were taken to hospital, operated on and they woke up the next day as patients. For some of the ones that collapsed because of burnout, the process was similar. They broke down and they were forced to leave the workplace. Also for others, with less dramatic courses of disease, becoming aware of the disease was often dramatic. For one of the physicians the awareness of extremely high blood pressure was the result of helping a colleague to test new technology. Another one consulted a colleague for a “trivial matter” and this contact led to a much more severe diagnosis than expected. In yet other cases, like a snowball, the insight to one disease came slowly, and once the snowball started rolling, it was easier to acknowledge other symptoms and/or diseases. When asked about the reasons for difficulties in gaining insight to personal illness, one of the physicians stated, “patients, not doctors, get sick”. The phenomena of viewing sickness as belonging to patients and of denying and ignoring sickness in colleagues did not help the sick physicians. They were surrounded by colleagues that “helped” them deny illness.

In spite of their access to knowledge and expertise, practically all of the participants spoke of difficulties to seek help. To them, it seemed bad enough to have to expose weakness and vulnerability to anybody, and even worse to have to consult a colleague. As one of the interviewees put it:

It is difficult to turn to colleagues that you work with on a daily basis and to leave yourself in their charge. I think that as a doctor, you're very independent and you think that you always know best. Maybe I'm simplifying, but it is not easy to seek help. I think this contributed to me falling ill, that I didn't seek help in time.

Yet, many of the physicians in this study eventually had to be treated by present or former colleagues, by somebody they used to supervise or went to medical school with. Being treated by a colleague is, therefore, the second subcategory of resisting sick and patient role. Some of the physicians, however, avoided this by referring themselves directly to doctors in other parts of Sweden, even though it did not entirely solve the problem, as, being a member of the medical profession, even a physician to whom they had no relation was considered a colleague. Others solved the problem by establishing a contact with a colleague they trusted, but with whom they did not work, as they did not want to burden their colleagues with their own problems, problems that might lead to sick leave and a load on the already overloaded colleague. The fact that it was often a close colleague that “discovered” that the physician was sick, thus seldom led to a long-term physician-patient relation.

The fact that the treating doctor and the patient/doctor, regardless of whether they were close colleagues or not, were both able to draw upon the medical discourse, however did not seem to facilitate the sharing of decisions. Instead, being patients and doctors, the interviewees were largely expected to answer for their own treatment, which is not the same thing as shared decision-making or empowerment. Being of central importance to the category of resisting sick and patient role, being in charge of one's own treatment is its third subcategory.

Considering the fact that being in charge of their own treatment is a way of dealing with the resistance to adopting an ordinary sick and patient role, it might be a paradox that the participants said they did not want to “be their own doctor”, mainly because of the notion that “all objectivity is lost when you are to deal with your own health”, as one of them put it. Whether they turned out to be their own doctor or not, most of them seem to have been very active as patients. This activity, however, took place outside the doctor-patient encounter. Many of the interviewees had spent a lot of time improving their own knowledge about their own diagnosis, or the speciality associated with it, in many cases in order to find a somatic or physiological explanation to burnout syndromes or other psychiatric disorders. In addition to the substitute of some other doctor taking responsibility for their care, the stories indicate that this seems to have reduced their feelings of shame and embarrassment. Further, though the physicians eventually became very active in their own treatment, initially it was not easy to accept the fact that they were in need of treatment, as it was seen as a sign of weakness and failure:

It is a strange self-image that you have, that medicine is for others and not for me. Do I really need it? It is some kind of failure and rejection. It is a proof of limitation and weakness, simply a violation.

This attitude and lack of compliance to treatment further illustrates some of the seemingly contradictory results of this study. The fact that the doctors wanted to be treated like any ordinary patient, at the same time as they resisted medication and treatment can be understood as the needs of the patient colliding with the identity of the doctor, an identity built upon the notion of caring for and taking decisions on treatment for others, but that does not allow personal illness. This means that dealing with colliding identities is the last subcategory of the category of resisting sick and patient role.

Resisting sick leave

Eventually the physicians anyhow accepted both sickness and sick leave, though very reluctantly, because of feelings of shame and failure, not only in relation to colleagues, family and friends, but in relation to patients. Some of them said they did not want the patients to find out about their sickness and sick leave and were afraid that their legitimacy as physicians and their ability to help others would be questioned, had it become known that they could not manage their own health. One of them, who had avoided sick leave by working part time and had worked at the health care centre every second week, said:

I hoped my patients thought I worked elsewhere when I was not here. Because it still feels like a gigantic failure, not to be able to handle working a lot and just clench my teeth and just go on.

This and numerous of other statements made by the physicians during the interviews illustrate the first subcategory of resisting sick leave, namely fearing loss of legitimacy. The second subcategory, which might be regarded as a strategy to resist both the sick role and sick leave, is working through sickness. As put by one of the physicians: “Then I steeled myself and went on. I didn't think I would have to be on sick leave, I never even imagined I would, and thought that this will sort itself out in the end”.

Many of the physicians interviewed said that, prior to the sick leave in question, they had not been on sick leave at all and that they had become used to “working their way through sickness” instead of calling in sick. “Then it became clear to me, that all they had on me was two days of sick-leave. Statistically I must have suffered from ‘imaginary health’”, was one way of putting it.

Working through sickness, however, was not only a strategy to avoid sick leave or a consequence of difficulties in adopting the patient role. The interviews also show that there was a pressure to be present at work and not to burden their colleagues by being absent:

You haven't had the liberty of being at home when you are sick, and worked through sickness. There has not been anyone to replace you and you don't want your colleagues to have to do your work since there isn't anyone to take your place. I guess there is reason to say that we have neglected our health badly. And it lies within the profession to deny your own sickness.

When asked about what made them so stubbornly resist sick leave, some of the physicians talked about the culture of the medical profession. Already at medical school, weakness and sickness were not tolerated. It meant working extremely long hours and always performing extremely well in order to be considered for attractive positions in the future.

Since all of the physicians anyhow ended up being on sick leave, working through sickness merely postponed what seem to have been inevitable. Postponing and limiting sick leave thus constitutes the last subcategory of resisting sick leave; because of a wish to be loyal to colleagues and patients, the physicians employed different strategies in order to reduce high workload and to avoid long-term sick leave. Prior to sick leave, some of them had chosen to work part-time, to work for the Medical Association, to have time to improve their knowledge or to be able to reflect on their situation. The women physicians said they also worked part-time in order to be able to combine work and family life, while the male physicians working part-time stated reasons that aimed at reducing high workload and pressure. Other strategies were to change position, to be transferred to another work setting or to give up being chief physician, head of clinic or health care unit.

When sick leave in spite of these types of strategies was impossible to avoid, there were negotiations with the certifying doctor to make it as short as possible. As a result of their strong wish to return to work and sometimes in opposition to the certifying doctor, some of the physicians managed to reduce the time on 100% sick leave, though not the ones that had fallen too seriously ill or collapsed and could not even return to the workplace: “I could not accept being fully on sick leave, I have resisted this a long time, I still don't like it, but I can see that I have become much better”, said one of the participants.

Spending time on sick leave

At the time of the interviews, none of the physicians were still fully on sick leave, which means that all of them had returned to work, though only partially (10–75%). Surprisingly few of the physicians, however, told of any type of treatment that had successfully eliminated the cause for their sick leave. For those with a diagnosis such as burnout or something similar, the treatment seems to have been limited to staying away from the workplace, resting and exercising. When I asked how they had spent their time when they were still fully on sick leave, many of them said that they had been sleeping a lot, especially in the beginning, as the period prior to sick leave had been a period of constant fatigue. Many of them had also been encouraged to take exercise by taking long walks, for example. As put by one of the physicians:

I slept. It was like letting out the tiredness of my whole life, like giving up and realizing that this is how bad it is. So I slept for 18 months, 12–14 hours a day, even though part of it was due to the medication. I spent time outdoors as well.

Being such an essential part of how time was spent during sick leave, sleeping during sick leave constitutes one of the subcategories of the category of spending time on sick leave. In between spending time sleeping and resting, like the participant quoted above, many of the physicians started thinking about their previous lives, rendering reflecting on one's life situation into the second subcategory of spending time on sick leave. For example, some of them said that when looking back at the period prior to sick leave, which in many cases was several years, it became clear that it had been characterized by stress and fatigue, but that this had become part of normal life. As put by one of the interviewees: “Now that I'm better, I realize that this went on for years. But you adapt … you can push yourself very hard.”

In order to reflect more systematically upon their life situation, some also went to therapy, while others said they could not even handle having to show up at a certain place at a certain time once a week. Not much time was spent on other activities, such as catching up with family and friends, as so much time was spent on sleeping. As soon they were well enough, they went back to work on a part-time basis, which meant that their part-time sick leave was spent on sleeping in order to recover between working periods.

Returning to work

The process of returning to work started already during sick leave, which means that, although the lines between it and spending time on sick leave blur, it is one of the categories. Some of the physicians that went to therapy said they were referred to a therapist, but many of them had to find one themselves. Surprisingly enough—or perhaps not so surprisingly considering the culture of the medical profession—it was the same for those with somatic diseases. It is true that many of them went to surgery or were subject to medication, but also for these physicians, somebody with the overall responsibility for their treatment and rehabilitation was lacking. Unlike other types of patients, it seems they were expected to treat and rehabilitate themselves, or at least take the initiative to get a certain treatment. As I described earlier, many of the participants seemed desperately to need somebody to talk to and “gave” me the role of therapist in some of the interview sessions, which was something that, in addition to what they actually said, made me think that I was the only one who had ever taken the time to sit down and listen to their stories.

In spite of this, all of the physicians eventually managed to return to work, at least partially. Their stories, however, gave me reason to believe that this was not the result of treatment and rehabilitation measures taken by those in charge, as many of the interviewed said that they did not get the support and help from management they were entitled to. Just as they had been in charge of their medical treatment and rehabilitation, they themselves had arranged for part-time work, change of workplace or change of tasks. Being in charge of one's own rehabilitation is therefore the first subcategory of returning to work.

Closely related to the subcategories of dealing with colliding identities and reflecting on one's life situation described above, the second subcategory of returning to work is fighting for a lost identity, as all of the physicians fought their way back to—if not health and well-being—work. Initially, getting in touch with the participants of this study was not easy, and at first, this puzzled me. I had expected that since all of them were on sick leave, this would not be a problem. Numerous calls to their home numbers, in some cases, were pointless. As it turned out, I had to call them at their workplace, as all of them had managed to return to work, at least on a part-time basis. In addition, when I finally realized that I had to try to reach them at work; it still was not easy to get hold of some of them. Contrary to my conception of people on sick leave, it became clear that I was dealing with very busy people. Although many of them, at least initially, thought of getting sick and having to be on sick leave in terms of a failure and even a threat to their authority, identity and legitimacy as doctors, some of them said that it had brought about new attitudes to life and work in general. But even though it in some cases was a question of appreciating other things in life, work remained the most important part of their lives. During one of the interviews, one of them asked, “What would I do if I didn't have work to go to? How would I make time pass?” And when I asked about whether they had found new hobbies or areas of interest, the answers were consistently in the negative. To the participants, it was difficult to imagine life beyond the identity as a doctor. When we discussed the possibility of returning to work on a full-time basis, it became clear that almost all of them desperately wanted to work full-time, but that almost none of them thought this would ever be possible. Still, they had fought their way back to a lost identity, that of a doctor.

The third subcategory of returning to work is coping with workload and pressure, because also after returning to work, the physicians used different strategies to cope with the workload and pressure. Working hours were planned and scheduled in consideration of perceived needs of both patients and colleagues. For example, part time working hours were scheduled so that meeting hours and other “non-productive” activities were not included. One of the doctors said: “I can't sit here and say that I'm sick ….We've got to help each other. If not, somebody else will get sick.”

Though this pretty well illustrates the attitudes of most of the physicians in the study, some of the general practitioners said that, after returning to work, they were “protected” by their co-workers and not allowed to fall back into their old habits, to ensure that they did not have to be 100% on sick leave again:

I'm very much supported by the nurses. They are really tough when they plan my working hours so that I'm given that space. They know that I can't take that much more.

However, while the general practitioners were protected and supported when trying to find flexible solutions for working hours and schedules, the hospital physicians were not given the same support by their colleagues. Indeed, most general practitioners returned to the same health care centre when returning to work, while the hospital physicians more often moved to another ward, unit or even hospital and by doing so, changed tasks. After returning to work, both groups seem to have concentrated on encountering, examining and treating patients, even though the majority of them stopped being on emergency duty, in spite of a culture in which being on duty is an essential part:

When we were in medical school, it was drummed into our heads that as a doctor you're supposed to handle being on emergency duty three days in a row without being tired. You played some kind of heroic role, almost like Superman.

The interviews thus show that when returning to work, most of the participants started to focus merely on patients, leading to the subcategory of secluding oneself from the organization. Many of them had played prominent parts in the development of their respective organization, but out of fear of bringing back to life conflicts that sometimes had contributed to the sick leave and/or fear of the pressure that comes with a wider commitment, they gave up being chief physicians or heads of clinics or health care centres. This way, they no longer had to be responsible for the implementation of organizational changes, downsizing, mergers and transfer of tasks. Also in this context, lack of support can be one of the reasons. “You're supposed to be God and to be able to handle also this without getting the proper resources or support” as one of the physicians, who previously had worked as head of department, put it.

Finally, although there was a general agreement that all categories were subject to an increased workload, many of the interviewees said, “… at the end of the day, the others can put work aside and go home. As a doctor, you can't do that, and have to stay until all of the work is done”. This means that although parts of the commitment to the overall organization had vanished during the sick leave process, the loyalty to patients remained unbroken.

Discussion

Looking back at the stories of the sick physicians, a picture comes into my mind. The physicians walking along a road lined with pain, shame and loss, but also with commitment, loyalty and motivation. The road, however, is not straight, but has the form of a circle, or rather a helix, which means that, in the end, the physicians are in a sense back where they started, a bit more bitter and vulnerable, but also much more reflecting. Although they work part-time after the long walk, they are just as loyal, love work just as much and have to cope with the same working conditions as before leaving. Not entirely, though, as they themselves have employed strategies to prevent yet another walk along the road of sick leave. This means that the categories and their subcategories accounted for above, would, should I be inclined to draw a picture of how they are related to one another, be placed at different positions along this circle or helix, one stage leading to and reinforcing the other. Thus, it is a picture of a circular or helical process, a process formed in response to structure, such as the changing working conditions and organization of health care, and in response to the culture of the medical profession itself.

The question is what we—researchers and practitioners—might learn from the study of this “road” or process. Beginning with the method used, the constructivist approach to in-depth interviews generated knowledge on the culture of the medical profession that otherwise would have been difficult to acquire, illustrated by the fact that, during many of the interviews, my “needs” and motives as a researcher were overridden by the needs of the participant for somebody to talk to. For example, the fact that some of the interviews were carried out in my office or in the home of the participant, and that they sometimes were disrupted, because I had to fetch a glass of water or a face tissue for the sick doctor to be able to continue telling his or her story, reinforced the feeling that I was put in a position of “lay therapist” and the interviewee in a position of a “patient” of mine. During the first interviews, I looked upon parts of the stories as something I had to put up listening to, in order to get the information the financer of the project and I were looking for. As I went along, however, it became clear that the way that the physicians talked of themselves and their lives and that the roles taken by us as researcher/”lay therapist”/fellow-being and informant/”patient”/human being, respectively, were in themselves sources of results and inquiries. For example, the interview situation also gave rise to questions on my position as a researcher, on power relations in the interview situation and on ethical aspects of these types of studies, as when the interviewees exposed their vulnerability, I felt that my competence as “lay therapist” was limited and had to rely on my position as fellow-being. However, even though it is important to distinguish between research and therapeutic interviews, respectively, as it has been argued that only the former aims at acquiring knowledge (Kvale, 1996), this situation thus generated results on the culture of the medical profession and its attitudes to weakness, sickness and disease.

In addition to methodological aspects, there are theoretical aspects of this study that might provide new knowledge. Looking back at the categories presented above, however, it is clear that their contribution to theory is of substantive nature and that further study is necessary in order to develop theory of generic or formal quality (Glaser & Strauss, 1967). In the meantime, these categories and extant research and theory might be juxtaposed and in the theoretical framework presented by way of introduction, there are certain concepts that might be used in order to deepen the analysis and to set the direction for future research. When it comes to the subcategory of dealing with new types of patients, for example, the ways that ‘the consumerist discourse’ might be related to the disempowerment of the medical profession is intriguing. In the description above, it seems that the new types of patients and their access to new types of information led to reduced control and authority. Previous research indicates that, for example, lay knowledge on the Internet can be seen as a threat to the autonomy of the medical profession and that it can change the relationship between the profession and its clients (Hardey, 1999). According to Hardey, it could also, however, be seen as a resource to promote partnership, which, notwithstanding its somewhat contradictory results, is confirmed by this study, meaning that new types of patients are not necessarily a source of disempowerment and reduced control.

Further, above I raise the question of what happens when both patient and doctor are members of the medical profession and when both draw upon the “medical discourse” (Fairclough, 1997). Looking at previous research, it would seem that these prerequisites lead to an adoption of “the model of mutual cooperation” (Freidson, 1970). Instead, the category of resisting sick and patient role and the subcategories of having to be treated by a colleague, being in charge of one's own treatment, and dealing with colliding identities are based on stories of neglect, loneliness and deprivation. If drawing on research by McKevitt and Morgan (1997), these stories show that while the sick doctors to some degree want to be treated like any ordinary patients, they are expected to take, and take, the role of the patient in control. Nevertheless, instead of cooperation, this means leaving them without adequate counselling, information and emotional support. Instead of encounters where both parties draw on the medical discourse, the physicians seek knowledge on their own, in their struggle to get well and to return to work.

The category of resisting sick leave and the subcategories of working through sickness and postponing and limiting sick leave might be seen in the light of the different types of conceptions of illness originally specified by Herzlich (Gwyn, 2002). Based on their stories, it is quite clear that by the physicians of this study, and possibly by the medical profession as a whole, illness—and in this case, of course personal illness—is regarded as destructive, as opposed to illness in some cases also being regarded as a liberator or even an occupation. Illness is often regarded as destructive by people who are particularly active in society and consider themselves indispensable. As a result, according to Freidson (1970), taking time off for illness is unthinkable, whatever the state of health. This means there are other categories that are inclined to work through sickness, but previous research shows that the reasons behind this phenomenon might vary. Compared to other groups, attitudes to illness and culture of work seem to play a more important role for physicians (McKevitt et al., 1997).

In addition, the categories of spending time on sick leave and returning to work might be related to extant research and theory. For example, being in charge of one's own rehabilitation is no more an example of “mutual cooperation” than the type of doctor-patient relation accounted for above. Neither the treating doctors nor management, who are all members of the same profession, take responsibility for treatment and work rehabilitation of the sick physicians and seem to believe that absence from the work setting is all that is needed in order for the physicians to get well, in combination with their own treatment and rehabilitation efforts. Moreover, when the physicians return to work, they rely on them to arrange the work situation so that work rehabilitation is possible. The conception of it as something unthinkable and destructive is shared by the sick doctor and those who should be in charge, rendering it even more difficult for the sick doctor when dealing with colliding identities than perhaps it would be for members of other professions.

Not only are they expected to work extremely hard, but also they continue to be regarded as high achievers, even when it is a question of getting well and returning to work. The sick physicians themselves reinforce this type of behaviour, first by not realizing that they are sick and, second, by hesitating to seek care and to be on sick leave. They literally work until they drop and there is no choice. Moreover, because they seek help so late, it might both worsen their condition and prolong the sick leave period. Very few of them later return to full-time work. At the same time, however, even though aspects of this culture are destructive, it also has advantages. As the identity of the physician is so closely connected to the profession, it promotes a strong will and motivation to return to work. However, the stories upon which the findings of this article are based give reason to question whether it also promotes health and well-being among physicians. When the road from dealing with changing working conditions to returning to work is completed, the question is how the physician is to deal with the identity of still being partly on sick leave. How might this new identity take form in the shadow of the identity as a doctor? Moreover, how might they be prevented from having to enter once again the road to sickness? In other words, what might be learned on a more practical level?

According to Swedish law, the employer is responsible for work rehabilitation. This means that, alongside the treating doctor's decisions on treatment and medical rehabilitation, and in cooperation with the sick employee, the employer is to establish a plan for the rehabilitation back to work of the sick employee. According to the rehabilitation policy of the County Council of Norrbotten, this plan might include measures to reduce risks leading to sick leave, such as change of tasks, change of workplace, improvement of work environment etc. This plan is to be established already after one month of sick absence and then continuously revised during the sick leave. For both the management of health care organisations, for those in charge of implementing rehabilitation policies like the one referred to above and for doctors treating other doctors, the findings of this study point at the importance of starting to view the physician like “any other employee” or “any other patient”. It should, however, be noted that previous research, rather than advocating one type of decision-making, emphasizes the need for flexibility and individual differences in the physician-patient encounter (Charles, Gafni & Whelan, 1999) and that physicians treating other physicians should be sensitive to their unique needs (Taub, Morin, Goldrich, Ray & Benjamin, 2006). Further, this study shows that it is important for members of the medical profession to take an interest in its development.

Previous research shows that in the Swedish health care system, both nurse managers and clinical directors are subject to high demands that are not buffered by the right kind of support (Lindholm, 2006). Aspects of the work situation, such as an increased workload and the contradictory development of deskilling or work “on the conveyor belt” in combination with higher demands for technological and medical skills can be described in terms of increased demands in combination with lower levels of control. In addition, there is also a reduction of influence on the organization and the distribution of resources, as well as demands on managing the implementation of organizational changes such as mergers, downsizing and transfer of tasks. These issues need to be dealt with by the management of most health-care organizations in order to prevent increasing sick leave numbers in physicians. Furthermore, research shows that physicians’ well-being is important to the quality of medical care (Firth-Cozens, 2001).

This study is limited to the stories of 15 participating physicians in the north of Sweden. However, it might encourage both researchers and practitioners to focus on the whole sick leave process rather than merely investigating what working conditions that cause sick leave and merely developing effective rehabilitation measures when sick leave is already a fact.

This study was funded by The County Council of Norrbotten, Sweden. I should like to thank all participating physicians for their stories. I also wish to thank Associate Professor Peter Waara and Professor Ewa Gunnarsson as well as the editor and the anonymous reviewers of the International journal of qualitative studies on health and well-being for their helpful comments on the article.