Narratives of burnout and recovery from an agency perspective: A two-year longitudinal study

Contents lists available at ScienceDirect Burnout Research journal homepage: www.elsevier.com/locate/burn Narratives of burnout and recovery from an agency perspective: A two-year longitudinal study Stela Salminen a,⁎ , Elena Andreou a , Juha Holma a , Mika Pekkonen b , Anne Mäkikangas a,c a Department of Psychology, University of Jyväskylä, Jyväskylä, Finland b Peurunka Rehabilitation Centre, Laukaa, Finland c Faculty of Social Sciences (Psychology), University of Tampere, Finland ARTICLE INFO Keywords: Interview Narrative analysis Spheres of meaning Rehabilitation Occupational health ABSTRACT Purpose: To provide knowledge about the recovery process during rehabilitation and two years later by exploring the manifestation of agency and spheres of meaning in the narratives of participants in a national rehabilitation course. Material and methods: The subjects of the study were four participants in a national rehabilitation course, whose burnout levels had decreased between the initial and follow-up periods of the course. Semi-structured interviews on two occasions and an electronic questionnaire 1.5 years post rehabilitation comprised the main material. In addition, the BBI -15 (Bergen Burnout Indicator) and DEPS-screen were used. Results: Thematic narrative analysis revealed highly individual and heterogeneous paths of recovery. The first parts of the narratives demonstrated a high degree of proxy and collective agency. As recovery progressed personal agency was strengthened. The spheres of meaning participants attached to their stories also varied, with the rehabilitation course and the professionals involved being viewed as morally good. Three major common themes were identified: 1) the benefits of the rehabilitation course; 2) supervisor support; and 3) personal factors. Conclusions: Burnout and recovery are not the consequence of work-related or individual-related factors alone, but rather the outcome of a combination of these. Regardless of common factors, the process and the degree of recovery may vary. The most beneficial path was observed when personal agency was high and was reinforced by a supportive job environment and favourable personal factors. 1. Introduction Burnout is a work-related stress disorder, which affects a large number of people of working age in industrial countries. In Finland, in the latest national representative survey conducted in 2011, 2% of men and 3% of women in the workforce suffered from severe burnout, while the percentage of those with mild burnout was significantly higher: 23% and 24% for men and women respectively (Duodecim, 2015). In occupational health psychology, burnout is typically described as a reaction to long-lasting untreated occupational stress, which is characterized by three symptoms: exhaustion, cynicism and reduced professional efficacy (Maslach, Jackson, & Leiter, 1996). It is not classified as a medical disease, but instead an additional code (ICD-10: Z73.0) (Duodecim, 2015), which indicates the presence of a problem related to life control, is added to the diagnosis. The relatively high prevalence of burnout underlines the need for a better understanding of its antecedents, development processes and, above all, the paths to recovery. In response to this need, this study sought to capture subjective experiences from the onset of burnout to recovery. Since burnout investigation began, a large body of research –mainly quantitative –has focused on the causes, symptoms and consequences of burnout (Schaufeli & Enzmann, 1998; Schaufeli, 2000). Although no single psychological theory offers a full explanation of burnout, it is commonly agreed that the causes of burnout can be divided into three categories (Schaufeli & Enzmann, 1998;Schaufeli, Maslach, & Marek, 1993): individual (burnout is regarded as the outcome of intrapersonal factors); interpersonal (burnout is seen as the result of difficult relations with others at work); and organizational (burnout is viewed as a mismatch between the person and the job). Consequently, burnout interventions have focused on these same categories of causes (Schaufeli & Enzmann, 1998). While rehabilitation interventions have been found to exert a positive effect on burnout reduction (Norlund et al., 2011; Stenlund, Ahlgren et al., 2009;Stenlund, Birgander, Lindahl, Nilsson, & Ahlgren, 2009;Stenlund, Nordin, & Järvholm, http://dx.doi.org/10.1016/j.burn.2017.08.001 Received 28 March 2017; Received in revised form 9 August 2017; Accepted 10 August 2017 ⁎ Corresponding author. E-mail address: [email protected].fi(S. Salminen). Burnout Research 7 (2017) 1–9 2213-0586/ © 2017 The Authors. Published by Elsevier GmbH. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/BY-NC-ND/4.0/). MARK 2012), particularly in the components of exhaustion (Hätinen et al., 2009;Hätinen, Kinnunen, Pekkonen, & Kalimo, 2007;Hätinen, Mäkikangas, Kinnunen, & Pekkonen, 2013) and cynicism (; Hätinen et al., 2007), the focus tended to be on investigating correlations between variables rather than individual developmental trajectories of burnout and recovery (Mäkikangas & Kinnunen, 2016). The few qualitative studies on rehabilitation have highlighted an important concept –that of agency –and argued that the strengthening of agency is one of the primary goals in rehabilitation interventions (Järvikoski, Martin, Autti-Rämö, & Härkäpää, 2013). Theoretically, three different modes of agency are distinguished: personal, proxy, and collective (Bandura, 2000). Personal agency implies taking control or exercising influence with the aim of achieving or producing given effects. In many cases, however, individuals are unable to directly impact their social environment. Instead they employ a mediated form of agency, proxy agency, in which they turn to others in possession of the appropriate resources, knowledge or means to act on their behalf to produce the outcomes they desire. The third mode, collective agency, emphasizes the fact that many outcomes are achievable only through joint effort. At the core of it is people’s shared belief in their power to achieve desired outcomes. Although other studies on burnout rehabilitation do not refer specifically to the concept of agency, they identify similar themes and categories. In a study of an established rehabilitation intervention in Finland, utilizing the same baseline interview data as in the present study (Salminen, Mäkikangas, Hätinen, Kinnunen, & Pekkonen, 2015), the importance of the accumulation of support, awareness and approval led clients to the revelation that they are primarily responsible for their own well-being. The overarching theme of My well-being in my own hands, which emerged in the analysis, strongly resembles the concept of agency (Bandura, 2000). Another study, conducted in Sweden, of patients’experiences of rehabilitation consisting of two different intervention groups –one with cognitively-oriented behavioural rehabilitation and QiGong (a programme combining tension-relieving movements, relaxation and meditation), and the other with QiGong alone –produced similar results (Fjellman-Wiklund, Stenlund, Steinholtz, & Ahlgren, 2010). During the recovery process, patients emphasized the beneficial effects of affirmation and support from rehabilitation professionals and group members, in enabling them to restore control over their lives. Despite their contribution to the understanding of the mechanisms of change, a common feature of these studies is the use of a cross-sectional design, which does not yield sufficient information about participants’post-intervention experiences. To our knowledge, there are no published qualitative studies which have tracked the development of the recovery process over time. To fill this gap, the present study is a pilot research comprising four narratives, which sets out to investigate the entire process from the onset of burnout to recovery at different time points during and after rehabilitation. The main focus was on the content of the narrative, i.e., on what participants told about the various stages of the process. As the reinforcement of agency is an important goal in rehabilitation and its significance has been demonstrated in previous studies (Järvikoski et al., 2013; Salminen et al., 2015), a further aim was to explore the role of the three modes of agency (personal, proxy and collective) in more depth along with their impact on recovery within this sample. Narrative inquiry was chosen as the primary method of investigation, as recovery paths are typically individual and heterogeneous (; Hätinen et al., 2013). 1.1. Narrative analysis Narratives enable the exploration of human experience in a holistic manner by preserving its rich and complex nature (Bell, 2002) and recognizing lived experience as a source of knowledge and understanding (Clandinin, 2013). Within the framework of narrative research, a plethora of different definitions, approaches, and specific methods exists (Pinnegar & Daynes, 2007). For the purposes of this study, thematic narrative analysis, as described by Catherine Riessman (2008), was applied. It is one of the most commonly used types of narrative analysis, and one in which exclusive emphasis is placed on the content of the story told. While preserving the features of the individual narratives, this type of analysis enables identification of common thematic elements across the participants’accounts (Riessman, 2004). This type of analysis, with its exclusive focus on narrative content, was considered particularly suited to capturing both participants’experiences over time and the meanings they attribute to the process of recovery. At the core of narrative inquiry lie questions about what the story reveals about the person and his/her world and what light interpretation of the story can shed on the cultural and social milieu in which the narrative is produced (Patton, 2015). In constructing their stories, individuals both incorporate their understanding of the surrounding world into the fabric of their narratives and engage in a sense-making process. Several spheres of meaning –emotional, explanatory, moral and ethical –can be discerned in narratives (Hänninen & Koski-Jännes, 1999) and these are explored in this study. Emotional spheres of meaning refer to the emotional underpinning of a narrative –whether the goals of the main character are achieved or thwarted, whether the story has a positive (happy) or negative (tragic) ending, what its high and low points are (Gergen, 1988). The explanatory perspective is related to the manner in which events are linked with each other as causes and consequences. The moral sphere of meanings deals with the allocation of responsibility for events and whether there are elements of accusation or excuse in the narrative (Baumeister & Newman, 1994). The ethical aspect articulates the deeper resonating elements of good and bad as well as right and wrong in life. Thus, in addition to the manifestation of agency and the exploration of the entire process of recovery, these four spheres of meaning are all discernible in the participants’narratives. Their identification provides a yet deeper level at which subjective experiences can be captured, while simultaneously revealing important perceptions about the broader social and cultural milieu in which the narratives are produced. 2. Material and methods 2.1. Study design, participants and data collection This study comprises the narratives, collected on three occasions, of four participants in a Finnish national rehabilitation course. These courses were state financed, i.e., the Finnish Social Insurance Institution funds rehabilitation services and provides income during participation in the course. Participants either apply or are referred by their occupational health care service for a burnout intervention. Applicants and referrals are screened by the local branch of the Social Insurance Institution and the final selection is made by a physician from the rehabilitation centre. These rehabilitation courses start with a 10-day period and end with a follow-up period, which takes place within at most 7 months after the first period. Rehabilitation includes an evaluation of an employee’s physical and psychological status by various professionals, including a physician, psychologist, physiotherapist, and social worker. The rehabilitation activities in burnout interventions focus primarily on enhancing individual resources and supporting coping strategies. This is achieved through various individualand group-level activities. Individual-level activities, which account for 11 h of the 70-h rehabilitation programme, include guidance and counselling with various rehabilitation professionals, tests and examinations, and specific tasks to be completed between the rehabilitation periods. Group-level activities, amounting to 59 h, comprise physical exercise, education on nutrition, ergonomics and health, and participatory group discussions. Participants for this longitudinal study were selected from various S. Salminen et al. Burnout Research 7 (2017) 1–9 2 rehabilitation courses that took place in 2012 at a rehabilitation centre in Central Finland (Salminen et al., 2015). Of the three rehabilitation courses with a follow-up period between August and December 2012, 15 eligible employees, whose burnout levels had decreased, were identified, and 12 participated in the first phase of the study (Salminen et al., 2015). The psychologist at the rehabilitation centre informed clients whose burnout score had diminished of the possibility to participate in the study. Their willingness to participate in the follow-up phases of the study was also inquired. The baseline levels of burnout and the subsequent change were measured with the Bergen Burnout Indicator (BBI-15) (Näätänen, Aro, Matthiesen, & Salmela-Aro, 2003). The BBI-15 comprises the three symptoms of burnout: exhaustion, cynicism and reduced professional efficacy. Total scores vary between 15 and 90, forming four different classes of burnout: severe, moderate, mild and no burnout. Eligibility was determined if the burnout symptoms had decreased by at least one class and in at least one sub-dimension. The first interviews, which provide accounts of burnout development and experiences of the rehabilitation course were conducted after the end of the course follow-up period in the autumn of 2012 on rehabilitation centre premises (Salminen et al., 2015). Participants were approached 1.5 years later (in Spring 2014) via an electronic questionnaire sent to their private e-mail addresses. The follow-up questionnaire measured participants’burnout and depression score with the BBI-15 (Bergen Burnout Indicator, BBI-15, Näätänen et al., 2003) and with the DEPS-screen (Salokangas, Poutanen, & Stengard, 1995). The questionnaire also included open questions, in which participants were asked to evaluate, e.g., their manager’s activities, changes that had occurred at work and the impact different rehabilitation factors had had on their well-being. A reminder to those who had not answered was sent at the end of May 2014. Nine out of twelve participants sent back the filled-out questionnaire. Of the three participants who did not respond, one was on sick leave at the time of data collection and thus did not receive information about the questionnaire. The other two participants could not be reached. Of the nine respondents, eight gave their consent to be contacted for further inquiry. Of these eight, four were selected through purposeful sampling (Patton, 2015) for an in-depth interview with a view to collecting differing stories of burnout and recovery. The sampling was based on the degree of recovery (continued recovery or worsened burnout) and on whether job change had taken place (two of the participants had changed jobs and two were in the same job). They were approached by e-mail and their willingness to participate in a follow-up interview was elicited in October 2014. Sample age at the last data collection point varied between 41 and 64 years (mean 55.8 years). Two of the participants had a vocational school education, one had completed a professional polytechnic qualification and one had a university degree. Time with the current employer varied between 3 months and 40 years. All participants were currently working, but one was on part-time retirement. The four participants whose recovery process was explored in more detail were interviewed by means of a thematic interview by a Master’s student of Psychology in NovemberDecember 2014. The interviews were conducted in a quiet place in the participant’s home. Each interview lasted between 1 and 1.5 h. The interviews were transcribed verbatim; however, the precise length of pauses, stutters and other sounds, and non-verbal language were not written down as the main emphasis was on the content of the story told. The total length of the transcribed interview material from the first and third data collection points was 138 pages, while the questionnaire administered at the second data collection point was 12 pages in length. 2.2. Data analysis The accounts given at these three data collection points were regarded as three parts of the same person’s narrative. This approach was adopted to honour the longitudinal design of the study and, more importantly, to allow the tracking of change over time through comparison of each participant’s descriptions of burnout and recovery at the different time points, and of the different social conditions and places in which the data had been collected. This accords with the three commonplace features of narrative inquiry posited by Connelly and Clandinin (2006): attention to temporality, sociality, and place. Temporality refers to the continuity of experience and the fact that people and events are situated in time and are described with a past, present and future. The sociality commonplace reflects the notion that personal and social conditions shape the individual’s context. The third commonplace draws attention to the fact that all events occur in a concrete physical place. In line with Riessman’s (2008) presentation of thematic narrative analysis (p. 54), data were interpreted in light of the agency theme, which served as the theoretical framework for the study, i.e., the participants’accounts were viewed through the lens of agency and how its different modes were interwoven therein. In the analysis, the focus was on the content of the narratives, i.e., on what participants reported about the process, and not on how they reported about it. Furthermore, the choice of narrative analysis was justified as we wanted to view the narratives as whole entities instead of identifying categories as, for example, is done in other types of qualitative analysis. The narratives for each participant from the three collection points were first read to grasp the general meaning and spirit of the story. During successive readings we focused on uncovering the particular manner in which the participant reconstructed the formation of burnout, how change started to emerge in the story, particularly in the rehabilitation stage, and how the process had continued at 1.5 and 2 years after the end of the rehabilitation course. While reading, we attempted to discern all three modes of agency as well as the presence of weak agency or the absence of agency. Once we had obtained a better understanding of the story, we proceeded to identify its emotional, causal, moral and ethical aspects. These were based exclusively on the participants’subjective experiences and their interpretation of the events before and during the burnout stage and throughout the recovery period. Finally, we assigned a title to the narrative, which reflected its core meaning. When the narratives of all the participants were analysed, we isolated the most important common themes for closer examination. The common themes were recurring topics, events or factors, found in all four narratives, which were experienced as particularly impactful in the whole process, and were thus indicative of either successful or impeded recovery. The participants were identified by a pseudonym to guarantee their anonymity and avoid any risk of their being recognized based on the narrative presented in the study. 2.3. Credibility To increase the credibility of the study, the participants were originally selected from three different courses. The rehabilitation courses comprise a lot of group exercises, which aim to enhance group cohesiveness. However, this may lead to groupthink, and inhibit the expression of individual and differing opinions. By studying participants from different groups, we aimed to minimize this effect. To increase the credibility and interpretive potential of the study, data triangulation based on (1) a self-report questionnaire (BBI-15); (2) a written questionnaire at Time 2; and (3) semi-structured interviews at Time 1 and 3, was used (Denzin, 1970). 2.4. Ethical considerations The study was approved by the Ethical Committee of the University of Jyväskylä, Finland. The study was introduced to the rehabilitation centre clients by the psychologist, who provided them with written materials. The voluntary nature of participation, anonymity and confidentiality were emphasized before the start of the study and were mentioned at each subsequent stage of the data collection. The participants signed a consent form and were informed of the possibility to discontinue their participation at any time during the study without any S. Salminen et al. Burnout Research 7 (2017) 1–9 3 consequences. A consent for audio recording was also requested. Rehabilitation staffrepresentatives were not present at any point during the interviews. 3. Results 3.1. Sara, age 57, kindergarten teacher, divorced: “Change is the only constant” Sara’s narrative reveals the story of a woman who found herself grappling with massive and constant changes in the workplace (Table 1). As these changes spun out of her control, they affected her well-being and eventually led to very severe burnout. It is important to note that she did not perceive that her burnout was caused by the nature of her work, personal conflicts or team-related schisms, but as a result of the rapid accumulation of changes, which undermined her feeling of belonging, role clarity and professional identity.“What then started increasing my feelings of burnout was the constant, incessant change and the fact that there was no stability or clarity. I myself had the feeling sometimes that I don’t know who I am anymore, let alone at the time when I was my colleague’s supervisor for a year –a year and a half, and then all of a sudden I was an ordinary line worker alongside fellow workers.” Her personal agency was present in the burnout stage in her efforts to call the management’s attention to the detrimental effects of the constant organizational changes, but this mode of agency seemed to lack efficacy and failed to make an impact on the overall situation. “When we then started falling ill and taking sick leave absences and substitutes came in, and they started falling ill, then occupational healthcare started to take an interest in our problem, the thing is that we tried very hard to communicate this and spread the word that things are not all right here, but simply nothing happened.” The question of the accumulation of stress becomes particularly evident here: as Sara’s burnout co-occurred with physical ailments and family problems (divorce), she attributed her symptoms of exhaustion, loss of professional pride and cynicism to these problems and thus failed to recognize the primary cause of her burnout. “You are as burnt out as you could possibly be…you really need a break...I myself thought in a way that…that I have to demand more and more from myself, I have to just manage and cope with it, and so...I just didn’t realize how deep I was in my own burnout.” The high point in Sara’s narrative was at the end of rehabilitation, which she saw as good, useful and affirmative. She also praised the role of her daughter and pet dog, who gave her the gentle push she needed for recovery. Her personal agency remained somewhat in the background, whereas proxy agency (help from rehabilitation professionals as well as family members) and collective agency (help from the rehabilitation group) were boosted. After her first change of job, Sara continued on the road to recovery, but soon the same problems as in the previous workplace started to emerge. She had three different managers during the two years she spent in the second workplace. Thus, her account of the events during the 1.5 years after rehabilitation, was tinged with disappointment: replication of the same problems after her change of job following the rehabilitation course led to severe burnout, as also identified by her score on the burnout scale (BBI-15). Eventually, after this first job change she moved to her third workplace, shortly before the interview two years after the rehabilitation course. The third sub-narrative was evaluative and retraced the full spectrum of events and emotions accompanying her burnout, rehabilitation and present state of well-being. Sara also discussed the effects of personnel turnover and the lack of stability in the workplace, thus attaching a morally negative meaning to the latter. It is noteworthy that despite the disappointment caused by her first job change her narration of events was informed by a certain degree of awareness and serenity. She was able to offset the full development of burnout by demonstrating proactivity, exercising personal agency and changing jobs again. The emotional tone in the third sub-narrative was cautiously optimistic and pronouncedly realistic, and Sara was not resigned to helplessness or hopelessness. Personal agency was manifested in her intentions to keep looking for a satisfying job: “It hasn’t at any point entered my mind, oh, what did I do, I mean I haven’t thought, did I make the right decision or anything like that...I haven’t really…and I saw that my coming to X. (her new workplace) happened just at the right point, it was good…[…]. My earlier workplaces were somehow...I always had to start from scratch and again from scratch and create routines…and I would have liked to create something, to drive something forward, but the reality of work somehow devoured so much. […] Now I feel I want to look at all the possibilities…’cos you know this job can be done in so many different ways…and especially here in [this] area there are so many ways…and it’s an experience for me.” The fact that Sara had moved to a new workplace prior to the interview explained why the narrative remained open-ended. She recognized that her recovery was a long process and was aware that she was not yet fully recovered. She hoped for a better outcome this time, and allowed herself to explore further opportunities, should the right moment arise. Sara’s personal agency, although existent, was rendered ineffective and futile by poor management of problems in the workplace. Proxy and collective agency were the most actively exercised during rehabilitation, which in turn boosted her personal agency. The main cause for developing burnout were the constant organizational changes, which undermined Sara’s feeling of stability, blurred her role clarity and robbed her of her sense of belonging and control. The narrative blamed the multitude of, in her view, pointless changes in the workplace which were not handled adequately by the management. Credit was given to family members (daughter and dog), who provided support without passing judgement. On a deeper, value-driven level, Sara saw the attitude of disrespect and disregard for her suffering as ethically bad even when these were openly raised. What Sara emphasized as ethically good was the support, despite all her suffering, provided by occupational health care, the rehabilitation professionals and family members. 3.2. Diana, age 61, reception nurse, married: “Eyes set on retirement” Diana’snarrative represents the path of an employee who suffered from severe burnout, which could not be completely treated, and who expects retirement to resolve the problem (Table 2). For Diana, the accumulation of stressors leading to exhaustion was something that started a long time ago, when the number of reception nurses remained unchanged while the amount of work steadily increased. Before the onset of her burnout Diana described her working conditions as inadequate and the organizational climate as Table 1 Open-ended narrative “Change is the only constant”. Massive constant changes −> very severe burnout −> rehabilitation 1st period −> change of job −> rehabilitation follow-up period −> improved well-being −> exacerbation of the work situation −> change of job −> mixed feelings/cautious optimism −> contemplation of change −> open-ended narrative Emotional tone: mixture of optimism and disappointment Explanatory: constant changes, no firm ground Morally blamed: the lack of stability undermines one’s foundations; problems are not taken seriously by management Morally praised: daughter and dog Ethically good: presence of support Ethically bad: being treated without respect; being robbed of professional dignity S. Salminen et al. Burnout Research 7 (2017) 1–9 4 unsupportive. Employees’needs were disregarded and the management failed to react to inequalities. Her hopes of improvement were shattered when, contrary to promises, an extra person was not hired, which led her to see herself as a victim of severe injustice. It was at this point that Diana felt she could no longer cope, and she started complaining of stomach aches and sleeplessness, which in turn led to increased exhaustion and sensitivity. “And then the moment came…that I couldn’t cope anymore. I had somehow managed up to when they promised again and again extra help, but then in the winter it just came to the point where I couldn’t cope anymore. I had come up against a wall.” An additional burdening factor mentioned by Diana was her sick mother who required daily care and taxed her resources. Through exercising proxy agency, with the help of occupational health care professionals and the rehabilitation course, Diana was able to moderate her burnout symptoms and achieve an improvement in her well-being. This was translated into increased energy levels and the motivation to engage in more physical exercise. However, Diana’s focus regarding her means of recovery lay mainly in reducing her workload and being granted part-time retirement. She recognized the impact of her physical ailments on her well-being, and thus her efforts were directed towards minimizing strain rather than achieving full recovery in order to return to her pre-burnout work capability. This slightly passive approach to recovery presumably led to an interesting discrepancy between her measured levels of burnout and the burnout she reported at the second time point. Although Diana’s overall burnout score diminished between the first and second data collection time points, she nevertheless reported an increase in burnout symptoms. Such a mismatch could perhaps be accounted for by her approaching full retirement age as well as the various physical ailments she suffered from. Presumably, she invested less effort in recovery, i.e., she exercised personal agency on a smaller scale, and instead focused her attention on retirement. The last part of the narrative at two years post-rehabilitation was more contemplative and oriented towards the family. Both sides of the burnout and recovery process were voiced. The overall tone became more philosophical as realisation of one’s age and finiteness grew. In relating her expectations of the future, Diana expressed a wish to be healthy and able to spend time with her children and grandchildren. Hence, her personal agency was more directed towards reaching and maintaining good physical health so as to be fit to serve her family members. “I’m waiting for retirement and then I hope I’ll be healthy again, so that we can travel and have hobbies, and then of course that I’ll have the energy to take care of my grandchildren and play with them…this is a wonderful time that I’m looking forward to.” In Diana’s narrative, proxy agency is the most prominent mode of agency and the mode exercised most actively across the study period. Diana holds experts’tips and recommendations in high regard and relies strongly on their knowledge and encouragement. Occupational health care and rehabilitation professionals along with her rehabilitation group played a significant role in the process of recovery and they remained instrumental in maintaining it. Personal agency was also identifiable in her account, although to a lesser degree. It was reactivated during and after the rehabilitation course, and was primarily visible in Diana’s attempt to maintain a healthier lifestyle through sports and balanced nutrition. The overall emotional tone was of slight resignation and fatigue, although in the last part hope and expectations for the future were also voiced. The narrative did not incline towards a positive or a negative end, but could rather be categorized as suspended. The cause of burnout was perceived as work overload and the lack of an adequate organizational response thereto, combined with personal ailments. From a moral perspective, the rehabilitation programme and the peer support were praised whereas the disregard for employees’suffering and the injustice inexperienced were harshly blamed. On a deeper level of values, Diana viewed her work ethic and her connection to her family members as good and the injustice perpetrated toward a particular employee group as ethically evil. 3.3. Paula, age 41, management secretary, married: “The successful job changer” Paula’s narrative describes the burnout and recovery process, starting from breakdown to the regaining of joy and clarity by means of a mental shift (Table 3). She depicts the accumulation of an excessive workload, to which she responded with habitual conscientiousness, and the implementation of managerial changes. Paula realised that the amount of work was becoming intolerable, but despite her attempts to bring the problems of understaffing to the management’s attention, her concerns were not adequately responded to and nothing changed to lighten the burden. Simultaneously, she had health problems which caused fatigue and prevented her from recognizing that she had symptoms of burnout. She also started suffering from sleeplessness and anxiety over unfinished work until her breakdown, which happened “overnight”. With regular help from occupational health care professionals (a physician, a health nurse and a psychologist), in other words, through exercising proxy agency, she was able to articulate her problems and become aware of her burnout. The identification of the problem triggered an initial reaction of fear: “It was awfully scary…you know, all of a sudden I can’t manage anymore. But the scary thing was that I had been walking with my eyelids shut, in a tunnel. And it (burnout) had been there for a long time, but I hadn’t realized it myself. So…fear was the biggest thing, I guess.” Paula demonstrated initiative and sought the services of a private therapist to gain an outside perspective on her state. She also spoke about the enormous gratitude she felt towards her dog, who was by her side, non-judgemental and supportive, throughout the entire process. After several stints of sick leave, she returned to work part-time, but continued on partial sick leave. Specific changes regarding the division of labour were introduced and agreed upon in writing, and an additional person was hired to lighten the workload. Table 2 Narrative “Eyes set on retirement”. Inadequate work conditions and chronic overload −> severe burnout −> rehabilitation −> Harnessing burnout −> Subjectively perceived exhaustion −> part-time retirement −> looking forward towards full retirement Emotional tone: resignation and fatigue Explanatory: work overload and lack of adequate response thereto; physical ailments Morally blamed: injustice Morally praised: rehabilitation professionals and part-time retirement Ethically good: own work ethic Ethically bad: perceived injustice Table 3 Narrative “The successful job changer”. Excessive workload −> Unawareness of the symptoms −> Utter exhaustion and breaking down overnight −> Occupational health care support −> Stints of sick leave and commencement of recovery −> Return to work −> Rehabilitation −> Processing −> Job change −> Revival −> Maintenance of balance Emotional tone: revival Explanatory: excessive workload combined with extreme conscientiousness Morally blamed: work ethic in the childhood family Morally praised: rehabilitation Ethically good: mental maturation Ethically bad/evil: perceived injustice S. Salminen et al. Burnout Research 7 (2017) 1–9 5 For Paula, rehabilitation came after the acute stage of burnout had been overcome and the initial phases of recovery had commenced. She believed this enabled her to benefit from it the most and she described the experience thus: “[What did rehabilitation mean for you?] P: Pausing…and complete and thorough processing of these things. It has been…let’s see how I am, how I feel. Perhaps I had always pushed myself into the background before, to make room for other things. So [in rehab] I could be myself, for myself. ” Paula had to go through a profound and thorough process of reprioritizing her values: coming from a family with entrepreneur-mother, Paula believed that work had to be done and was top priority. Positioning her own well-being higher in the hierarchy required conscious effort and personal agency, but the outcome, already at the first time point, was improved work-life balance. Overall, despite the positive changes, Paula was aware that the process of recovery was slow and time-consuming, and she admitted the possibility of a renewal of burnout. As a preventive measure, Paula brought up “talking earlier” and “strongly desiring change”. Factors hindering recovery were related to the management, who instigated the subsequent changes. Problems with her indirect supervisor led to an increase in burnout symptoms. With the help of her direct line manager, who was supportive and helpful, she was transferred to another unit within the same organization between the first and second data collection points. At the second time point, Paula experienced a lot of positive changes. She felt accepted and welcome in the new team, enjoyed greater responsibility and was able to craft her job more freely. Despite the long distance between her home and the workplace, Paula felt that she was in charge of her life again, which in turn had a positive effect on her relationship with her husband and children. The emotional tone of the narrative was elated and hopeful. The amount of personal agency she exercised had grown and was visible in her motivation to engage in sports, eat healthily and maintain her well-being. She also actively revisited the lessons learnt during rehabilitation and could now see her burnout in a positive light: “Burnout can affect anyone. I am happy I went through that hell, because I was able to grow as a person.” Paula’s state two years after rehabilitation (at T3) was good, although at the time of the interview she was on a sick leave due to physical ailments. She was able to reflect more extensively on her change of job and the apprehension she had felt at that time. Her motivation to avoid burnout or be able to detect it in time was visible in her conscious monitoring of her state, in her efforts to keep her workload within manageable proportions and in her ability to listen to her body’s signals, all of which are manifestations of enhanced personal agency. Proxy agency was exercised through the supervision provided by her workplace, which served as a forum for self-reflection and additional monitoring of her well-being. Paula had adopted a new mindset through a mental shift: “I’ve realised I’m not omnipotent […] a kind of tranquillity has settled in…being merciful to oneself […]I’ve done so much…that I find joy in things in a new way or in a bigger way than earlier. Earlier I would have stared at the unfinished work…oh, there are still so many things to do, and I’ve only done this…now it’s the other way around: I have finished so much, and well, there are still unfinished tasks…but they don’t stress me...I’ll get down to them next.” A factor taxing her resources was her mother, who was recovering from cancer, but this was counterbalanced by an increased pool of physical, mental and interpersonal resources. However, she reported an enhanced maturity and prudence in accepting additional tasks. In summing up the main reasons for her recovery Paula gave credit to the rehabilitation programme for providing her with valuable knowledge about burnout and helping her adopt a mindful attitude and an ability to be present. She expressed gratitude to her husband, who had been a source of support during the burnout period, and was thankful for the opportunity to have a new rewarding job. Her vision of the future was positive and more family-oriented. Paula demonstrated a lot of personal and proxy agency, both at the beginning, prior to her development of burnout, during her sick leave and after rehabilitation. It is noteworthy, however, that personal agency, even when present, failed to lead to the desired results at the workplace. The restoration of balance was a slow process in which personal agency was supported through the deployment of proxy and collective agency. Her extensive use of occupational health care was essential not only in raising awareness and initiating a mental shift, but also in building robust personal agency in the form of assertiveness and self-mercy. Paula held the accumulation of excessive workload and the mindset of being in charge accountable for the development of her burnout. This combination generated a vicious circle of increasing extra duties and an inability to put a stop to it. Breaking point was reached “overnight”. She found fault with the work morality adopted in her childhood family, whereas she praised the rehabilitation programme. On a deeper level of values, “evil”was seen in the injustice present in the workplace, in which employees were not treated equally. The ethical virtue in turn can be seen in Paula’s mental maturation. Her mental awakening and strengthening of personal agency, which occurred through the reception of external help, transformed her from being a victim into being an active agent. 3.4. Astrid, age 64, director, married: “Turning being married to the job into a happy marriage” Astrid’s narrative is an example of successful recovery (Table 4). It demonstrates well the recovery of personal agency in a balanced manner, although it should be borne in mind that her initial burnout level was not high. Astrid felt the onset of burnout two years before the rehabilitation course, when the pace at work intensified and she started having problems sleeping. Astrid’s exhaustion was the main reason she sought help and participated in rehabilitation. The pressure caused by work was intensified by personal factors (her husband’s disease and aggravated poor health) and accelerated by her own health problems, which taxed her physical and mental resources. The onset of burnout also caused Astrid to rethink what work meant to her: although she was close to retirement she did not want to leave her life’s work unfinished. Her response to burnout was proactive and personal agency was discernible in several aspects: she sought means to alleviate the situation by learning to delegate, by accepting social support from friends and colleagues, and by listening to music. The programme gave Astrid many take-home messages and led her to make significant shifts in her thinking and in her perception of her own boundaries. She became aware of her tendency towards perfectionism and adopted a more merciful and self-approving attitude. Her well-being even started Table 4 Narrative “Turning being married to the job into a happy marriage”. Pressure at work and concerns about husband’s health −> exhaustion −> unawareness of her own state −> recommendation by a doctor −> rehabilitation programme −> raised awareness −> increased self-mercy and assertiveness −> mental shift −> re-ordering of priorities −> full recovery Emotional tone: happy and relieved Explanatory: own perfectionism Morally blamed: feeling of insufficiency Morally praised: rehabilitation and husband’s recovery Ethically good: re-ordering of priorities in life and being reconnected to her family Ethically bad: – S. Salminen et al. Burnout Research 7 (2017) 1–9 6 improving between the two periods of the programme through an accumulation of positive events (improvement in husband’s health, normalization of working hours, time spent with grandchildren) and personal agency (sports, conversations at home about her spouse’s disease, changes in nutrition). The work environment and the hiring of an extra person allowed for a more long-term planning and forecasting, which in turn exerted a beneficial effect on Astrid’s well-being. She also attributed the commencement of recovery to the peer support received from the rehabilitation group, which reinforced her proxy agency. A significant mental shift was discernible in her account: “When I came here it felt like being in a tunnel, with no light at the end of it. The walls were falling in. But now I see so many other possibilities, that work is not always the only right thing to do, that there are other things you can spend your time on. I don’t know, I don’t understand if I was really so burnt out that I didn’t see anything else. But now my mind has begun to clear in a totally different way” Astrid’s well-being was maintained after the rehabilitation programme and was detectable already at the second time-point. This was also confirmed by the BBI-15 measurement score, which showed no presence of burnout. The third sub-narrative was assertive, imbued with positive emotion, self-confidence, dedication and commitment to work. Work was perceived as manageable and enjoyable, and Astrid’s leadership was visible in decision-making, conflict resolution and her handling of minor problems in the workplace. Astrid perceived herself as fully recovered. There is also a sense of relief, as her husband’s suspected disease turned out not to be serious. As her recovery advanced, Astrid put renewed effort into her hobbies, which further accelerated recovery. Astrid’s personal agency was high, and she argued that recovery starts from oneself, but she also exercised proxy agency by relying on the occupational nurse’s recommendations: “Recovery starts from oneself...I’ve learnt work the hard way and also I use the word letting go and I have delegated to the others...I have just learnt it and come to the realisation that if I don’t let go then I won’tbe able to manage…and then there’s the occupational health nurse […] who always reminds me of my well-being…so this is also a means of help” She concluded her narrative in a slightly humorous way, which condensed the meaning of the transformation she had undergone: “When I turned 60 I said in my thank you speech that the theatre has been my second home and my husband corrected me saying that it was the first (laughter)…and then I pondered for the first time whether everything is in the right order (laughter), that indeed it has been like that...I don’t know if it’s inherited from home, but work has always been number one…and it has been extremely difficult to change this, but I think I have succeeded in it.” Astrid’s narrative demonstrates the power of proxy agency represented by the doctor, who suggested taking a distance from work, and the occupational health nurse, who supported her at a later stage. Astrid was able to avail herself fully of the lessons learnt in rehabilitation. Her narrative, describing her raised awareness, mental shift and a changed attitude towards herself and work also shows the strengthening of her already existing personal agency. Her motivation to manage at work until retirement was high, but this was handled in a balanced way, allowing for the incorporation of other meaningful and important aspects of life such as the family, social engagements and sports activities. The narrative is complete, and its emotional tone is one of happiness and relief. Her recovery, commenced two years earlier, continued on a stable path, along with a realignment of her values and priorities in life. The explanation for burnout was found in oneself: Astrid’s perfectionism and inability to delegate duties is perceived as the main reason for her burnout. She blames her own feeling of inadequacy and her desire to maintain control over everything, whereas she expresses her gratitude to the rehabilitation professionals and for the positive events related to a family member’s health. On the ethical level, Astrid did not identify anything as “evil”, but instead celebrated the re-ordering of her priorities in life and the re-emergence of connectedness with her family. 3.5. Common themes According to Riessman (2004, 2008), thematic narrative analysis enables common thematic elements to be tracked across participants’ accounts while at the same time preserving the story as an entity. Following Riessman’s guidelines (2008), we identified three common themes that were encountered across the narratives and which were particularly meaningful for all four participants. 3.5.1. The benefits of the rehabilitation course There was common agreement that the rehabilitation course was crucial for the recovery process. The course was regarded by all participants as highly beneficial and awareness-raising. The support and affirmation given by the rehabilitation group validated the experiences each participant had individually gone through. The knowledge and practical advice provided by the rehabilitation professionals, along with their non-judgmental attitude, helped participants articulate and comprehend their burnout situation and achieve bodily and mental awareness. Through rehabilitation and the exercise of proxy (via the professionals) and collective (via the group) agency, the participants’ personal agency was re-activated and strengthened, with its effects extending well beyond the duration of the course. Rehabilitation was also an object of the participants’sense-making process and came in for moral praise. 3.5.2. Supervisor support Supervisor support was of immense importance for recovery. An understanding, empathic supervisor was able to alleviate the symptoms of exhaustion and facilitate the re-adaptation to work by ensuring adequate working conditions and providing emotional support. An unsupportive or neglectful supervisor, on the other hand, rendered one’s personal agency completely ineffective (in Diana and Sara’s cases), led to an increase in burnout symptoms and the experience of injustice. Furthermore, such a supervisor could precipitate a change of job (as in Paula and Sara’s narratives). 3.5.3. Personal factors Personal factors such as divorce, the sickness of a close family member or of oneself had a substantial impact on both the onset of burnout subsequent recovery from burnout. The simultaneous presence of a disease prevented or obstructed the participants from registering or distinguishing the onset of burnout. The fear and anxiety as well as physical ailments caused by the disease in question led in general to a decrease in personal resources and an increase in reported symptoms of exhaustion and impaired work capacity, as seen in Paula’s and Diana’s stories. Crises such as divorce slowed down recuperation, as demonstrated in Sara’s narrative. The suspicion of a major disease in the participant’s partner, as in Astrid’s case, had a similar effect, causing the participant to experience an enormous emotional burden on while the removal of that threat had an immediate positive impact on the participant’s well-being. 4. Discussion The present study investigated the narratives of four participants covering the onset of burnout and recovery therefrom over the course of two years. The results demonstrated that recovery from burnout is a slow and individual process, which does not follow a uniform path. The most crucial factors for recovery in this particular sample were agency, participation in the rehabilitation course and the supervisor support. Agency bears a resemblance to taking control of one’s life, S. Salminen et al. Burnout Research 7 (2017) 1–9 7 being in charge and assuming responsibility, and studies have shown that it plays a significant role in rehabilitation (Järvikoski et al., 2013; Salminen et al., 2015). Agency is a robust predictor of well-being and adaptive psychological functioning across time (Gallagher, 2012), which emphasizes further the long-term effects of a strengthened sense of agency. Congruent with this finding, rehabilitation was instrumental in reinforcing personal agency through proxy and collective agency with a relatively long-term impact. Conversely, the phase of burnout was associated with ineffective personal agency and the absence of other modes of agency. It is noteworthy that even in the narratives of interrupted recovery and increased burnout personal agency remained present, although to a lesser degree than in the narratives of stable recovery. The combination of high personal agency with a supportive job environment and conducive personal factors provided the most favourable prerequisites for beneficial recovery. The rehabilitation course was perceived by all participants as having had an enormous positive effect on their recovery. It provided numerous healing experiences, affirmation and valuable knowledge instrumental for the early recognition of burnout symptoms. Rehabilitation interventions have been recognized as means for the validation and affirmation one’s experiences of burnout in previous research, although the overall benefit has varied depending on the structure and content of the intervention (Salminen et al., 2015; Fjellman-Wiklund et al., 2010;Söderberg, Jumisko, & Gard, 2004). Supervisor support or the lack thereof had a huge impact on the ease of making the return to work and was an essential factor influencing participants’perception of stress. A supportive manager had a positive impact on well-being by alleviating the symptoms of exhaustion and helping the recovering employee re-adapt to work. This relationship has been demonstrated in other studies as well: Aas and colleagues (Aas, Ellingsen, Lindøe, & Möller, 2008) showed that certain leadership qualities, e.g., the ability to make contact, being understanding, being empathic, facilitate the return to work after sick leave whereas lack of support leads to an increase in the frequency of sick leave (Vaananen et al., 2003). The narratives presented here were highly individual and heterogeneous, which bears testimony to the fact that burnout is a multiplex phenomenon as confirmed by other studies (Rutherford & Oda, 2014). Participants were able to voice their understanding of the path they had walked and attribute particular meanings to the events and experiences encountered. As expected, there were also certain similarities and common features, particularly in the phase preceding the onset of burnout, e.g., heavy responsibilities, bodily manifestations, fatigue and ultimately breakdown, which have been reported in other studies as well (Ekstedt & Fagerberg, 2005). The recovery phase manifested more variation as the participants adopted partially divergent strategies (e.g., change of job vs. no change; agentic versus passive/receptive approach) and were affected by factors beyond their control, e.g., adverse changes in the workplace and attitude of supervisor. The narratives presented here demonstrate that a change of workplace per se was not a decisive factor either contributing to or impeding healing, as found in previous research (Bernier, 1998). A significant factor was the amount of change in important areas such as supervisor support, and the possibility to influence one’s workload and implement changes aimed at diminishing work stress. Personal factors such as deterioration in one’s own or partner’s health, or disruption in the family (divorce or death of a close relative) also played a major role in the exacerbation of burnout and impeded the path to recovery. This demonstrates that burnout does not occur in isolation from other life events, but in parallel with them, which in turn presents the risk of burnout remaining unrecognized. Furthermore, recovery may be slowed by these personal factors even when the necessary changes are implemented in the workplace. 4.1. Methodological considerations To increase the validity and reliability of the findings, we applied different types of triangulation. Triangulation between methods was implemented by collecting data from semi-structured interviews and self-report data (i.e., BBI-15, Näätänen, 2003) at the first and second data collection points). Researcher triangulation was also employed. The first interviews in 2012 were conducted by three independent interviewers who were not involved in the rehabilitation course and thus provided a neutral perspective. The interviews at two years post rehabilitation were conducted and transcribed by the second author. The third author, who is a clinical psychology researcher, provided knowledge of and practical guidance on the application of qualitative methods. The fourth author, who is a physician and an expert in burnout, provided access to the rehabilitation course and participants and ensured a smooth data collection. The last author and the project leader, who is an experienced work and organizational psychologist as well as an occupational health researcher, provided expertise on the theoretical background and data analysis. All the interviews were audiotaped and transcribed verbatim to ensure maximum data completeness. The filled-out questionnaires were stored in accordance with the ethical guidelines of archiving in Finland, and the BBI-15 scores were calculated by the first author to determine participants’burnout category at the second collection point. Purposeful sampling (Patton, 2015) was applied in the study as we wanted to explore the effect of job change on subsequent recovery. Semi-structured interviews were regarded as a suitable method for capturing the experiences of participants, and the use of a questionnaire containing both structured and open questions was seen as an apt tool for reporting interim status. The use of narrative methods of analysis was justified due to the size of the sample and the longitudinal nature of the data. The use of narratives allowed a more integrated interpretation of the experiences reported by the participants (Bell, 2002) and a focus on the overall road from burnout to recovery or repeated burnout. Tracking the process of recovery over time offered a unique opportunity to uncover individual paths and sense-making processes. The present study has naturally certain limitations. First, as it was the outcome of purposeful sampling, the size of the sample was very small (Patton, 2015). A larger number of studied participants could reveal further common themes and demonstrate in greater depth the impact on recovery of changing vs. not changing one’s job. Additionally, all the participants were female and thus further studies with mixed samples are needed to validate the results across gender. However, the study participants were representative of the population typically attending such courses (Hätinen et al., 2009, 2007, 2013). The results cannot be generalized beyond the study participants owing to the size of the sample and the fact that participant selection was based on a decrease in burnout symptoms. On the other hand, such generalizations are also beyond the scope and purpose of qualitative studies. As this is the first study investigating long-term recovery from burnout by applying a narrative approach, it should be emphasized that the outcomes of the analysis may reflect characteristics of the burnout and recovery processes specific only to these four participants. 4.2. Conclusions In all, the narratives demonstrated that the process and the degree of recovery from burnout vary, as a multitude of factors –individual effort, rehabilitation, change of workplace, and supervisor support – shape and modify the path each individual takes. The most favourable path to recovery appeared to be activated when personal agency was strong and other life areas –job environment, supervisor, and family members –were conducive to recovery. It should be recognised, however, that the relevance of agency may be overemphasized as the study participants were initially selected on the basis of successful recovery, even if purposeful sampling, using the criteria of degree of recovery and job change, was applied at the last data collection point. Although the aim of the study was not to evaluate the effectiveness of the rehabilitation course, it is noteworthy that the course was essential both S. Salminen et al. Burnout Research 7 (2017) 1–9 8 in equipping the study participants with the tools to recognize and deal with burnout and in strengthening all three modes of agency, and thus its inclusion in burnout treatment should be considered. Future studies should aim at deepening the existing knowledge on the relationships between agency, job environment and long-term recovery in a larger and more diverse dataset, preferably including family members’subjective experiences of the entire process. Conflict of interest The authors declare that there are no conflicts of interest. Acknowledgment The study was funded by a grant to Stela Salminen from the University of Jyväskylä. References Aas, R. W., Ellingsen, K. L., Lindøe, P., & Möller, P. (2008). Leadership qualities in the return to work process: A content analysis. Journal of Occupational Rehabilitation, 18, 335–346. http://dx.doi.org/10.1007/s10926-008-9147-y. Bandura, A. (2000). Exercise of human agency through collective efficacy. Current Directions in Psychological Science, 9,75–78. http://dx.doi.org/10.1111/1467-8721. 00064. Baumeister, R., & Newman, L. S. (1994). How stories make sense of personal experiences: Motives that shape autobiographical narratives. Personality and Social Psychology Bulletin, 20, 676–690. http://dx.doi.org/10.1177/0146167294206006. Bell, J. S. (2002). Narrative Inquiry: More than just telling stories. TESOL Quarterly, 36, 207–213. http://dx.doi.org/10.2307/3588331. Bernier, D. (1998). A study of coping: Successful recovery from severe burnout and other reactions to severe work-related stress. Work and Stress: An International Journal of Work, Health and Organizations, 12,50–56. http://dx.doi.org/10.1080/ 02678379808256848. Clandinin, D. J. (2013). Engaging in narrative inquiry. Walnut Creek, CA: Left Coast Press. Connelly, F. M., & Clandinin, D. J. (2006). Narrative inquiry. In J. Green, G. Camilli, & P. Elmore (Eds.). Handbook of complementary methods in education research (pp. 375– 385). Mahwah, NJ: Lawrence Erlbaum. Denzin, N. (1970). The research act in sociology: A theoretical introduction to sociological methods. London: Butterworth & Co. Duodecim, 2015. http://www.terveyskirjasto.fi/terveyskirjasto/tk.koti?p_artikkeli= dlk00681 Accessed 04.03.2016. Ekstedt, M., & Fagerberg, J. (2005). Lived experiences of time preceding burnout. Journal of Advanced Nursing, 49,59–67. http://dx.doi.org/10.1111/j.1365-2648.2004. 03264.x. Fjellman-Wiklund, A., Stenlund, T., Steinholtz, K., & Ahlgren, C. (2010). Take charge: Patients’experiences during participation in a rehabilitation programme for burnout. Journal of Rehabilitation Medicine, 42, 475–481. http://dx.doi.org/10.2340/ 16501977-0534. Gallagher, M. W. (2012). Agency, optimism, and the longitudinal course of anxiety and well-being. Doctoral Dissertation. Dissertation Abstracts International: Section B: The Sciences and Engineering, 73(4-B). Gergen, M. (1988). Narrative structures in social explanation. In C. Antaki (Ed.). Analysing everyday explanation. A casebook of methods (pp. 94–112). London: Sage. Hänninen, V., & Koski-Jännes, A. (1999). Narratives of recovery from addictive behaviours. Addiction, 94, 1837–1848. http://dx.doi.org/10.1046/j.1360-0443.1999. 941218379.x. Hätinen, M., Kinnunen, U., Pekkonen, M., & Kalimo, R. (2007). Comparing two burnout interventions: Perceived job control mediates decreases in burnout. International Journal of Stress Management, 14, 227–248. http://dx.doi.org/10.1037/1072-5245. 14.3.227. Hätinen, M., Kinnunen, U., Mäkikangas, A., Kalimo, R., Tolvanen, A., & Pekkonen, M. (2009). Burnout during a long-term rehabilitation: Comparing low burnout, high burnout-benefited, and high burnout-not benefited trajectories. Anxiety, Stress and Coping, 22, 341–360. http://dx.doi.org/10.1080/10615800802567023. Hätinen, M., Mäkikangas, A., Kinnunen, U., & Pekkonen, M. (2013). Recovery from burnout during a one-year rehabilitation intervention with six-month follow-up: Associations with coping strategies. International Journal of Stress Management, 20, 364–390. http://dx.doi.org/10.1037/a0034286. Järvikoski, A., Martin, M., Autti-Rämö, I., & Härkäpää, K. (2013). Shared agency and collaboration between the family and professionals in medical rehabilitation of children with severe disabilities. International Journal of Rehabilitation Research, 36, 30–37. http://dx.doi.org/10.1097/MRR.0b013e32835692d3. Mäkikangas, A., & Kinnunen, U. (2016). The person-oriented approach to burnout: A systematic review. Burnout Research, 3,11–23. http://dx.doi.org/10.1016/j.burn. 2015.12.002. Maslach, C., Jackson, S. E., & Leiter, M. P. (1996). Maslach burnout inventory. Manual (3rd ed.). Palo Alto, CA: Consulting Psychologists Press. Näätänen, P., Aro, A., Matthiesen, S. B., & Salmela-Aro, K. (2003). Bergen burnout indicator (15th ed.). Helsinki: Edita. Norlund, S., Reuterwall, C., Höög, J., Nordin, M., Edlund, C., & Birgander, L. S. (2011). Work related factors and sick leave after rehabilitation in burnout patients: Experiences from the REST-Project. Journal of Occupational Rehabilitation, 21,23–30. http://dx.doi.org/10.1007/s10926-010-9250-8. Patton, M. Q. (2015). Qualitative research and evaluation methods. SAGE Publications. Pinnegar, S., & Daynes, J. G. (2007). Locating narrative inquiry historically. In D. J. Clandinin (Ed.). Handbook of narrative inquiry: Mapping a methodology (pp. 2007). SAGE Publications, Inc.. Riessman, C. K. (2004). Narrative analysis. In M. S. Lewis-Beck, A. Bryman, & T. Futing Liao (Vol. Eds.), The SAGE encyclopedia of social science research methods: Vol. 2, (pp. 705–709). Sage Publications, Inc. Riessman, C. (2008). Narrative methods for the human sciences. SAGE Publications, Inc. Rutherford, K., & Oda, J. (2014). Family medicine residency training and burnout: A qualitative study. Canadian Medical Education Journal, 5,13–23. Söderberg, S., Jumisko, E., & Gard, G. (2004). Clients’experiences of a work rehabilitation process. Disability Rehabilitation, 7, 419–424. http://dx.doi.org/10.1080/ 09638280410001663111. Salminen, S., Mäkikangas, A., Hätinen, M., Kinnunen, U., & Pekkonen, M. (2015). My well-being in my own hands: Experiences of beneficial recovery during burnout rehabilitation. Journal of Occupational Rehabilitation, 25, 733–741. http://dx.doi.org/ 10.1007/s10926-015-9581-6. Salokangas, R. K., Poutanen, O., & Stengard, E. (1995). Screening for depression in primary care. Development and validation of the Depression Scale, a screening instrument for depression. Acta Psychiatrica Scandinavica, 92,10–16. http://dx.doi.org/10. 1111/j.1600-0447.1995.tb09536.x. Schaufeli, W. B., & Enzmann, D. (1998). The burnout companion to study and practice: A critical analysis. Washington, DC: Taylor & Francis. Schaufeli, W. B., Maslach, C., & Marek, T. (1993). The future of burnout. In W. B. Schaufeli, C. Maslach, & T. Marek (Eds.). Professional burnout: Recent developments in theory and research (pp. 253–259). New York: Taylor & Francis. Schaufeli, W. B. (2000). Burnout. In Karwowski (Ed.). International encyclopaedia of ergonomics and human factors (pp. 382–386). London: Taylor & Francis. Stenlund, T., Nordin, M., & Järvholm, L. S. (2012). Effects of rehabilitation programmes for patients on long-term sick leave for burnout: A 3-year follow-up of the REST study. Journal of Rehabilitation Medicine, 44, 684–690. http://dx.doi.org/10.2340/ 16501977-1003. Stenlund, T., Ahlgren, C., Lindahl, B., Burell, G., Steinholtz, K., Edlund, C., et al. (2009). Cognitively oriented behavioral rehabilitation in combination with Qigong for patients on long-term sick leave because of burnout: REST—a randomized clinical trial. International. International Journal of Behavioral Medicine, 16, 294–303. http://dx.doi. org/10.1007/s12529-008-9011-7. Stenlund, T., Birgander, L. S., Lindahl, B., Nilsson, L., & Ahlgren, C. (2009). Effect of Qigong in patients with burnout: A randomized controlled trial. Journal of Rehabilitation Medicine, 41, 761–767. http://dx.doi.org/10.2340/16501977-0417. Vaananen, A., Toppinen-Tanner, S., Kalimo, R., Mutanen, P., Vahtera, J., & Peiro, J. M. (2003). Job characteristics, physical and psychological symptoms, and social support as antecedents of sickness absence among men and women in the private industrial sector. Social Science Medicine, 57, 807–824. S. Salminen et al. Burnout Research 7 (2017) 1–9 9