Long-term sick leave and work rehabilitation - prognostic factors for return to work
MetadataShow full item record
- Faculty of Psychology 
The main objective of this thesis is to examine individual prognostic factors for return to work (RTW) after work rehabilitation, for workers on long-term sick leave with common musculoskeletal and mental health complaints. The process of returning to work after long-term sick leave may be complex, and is often influenced by other factors than health complaints and diagnoses alone. The primary hypothesis in this thesis was that individual’s cognitions about health and illness would be central for returning to work or not, after work rehabilitation. A second hypothesis was that socioeconomic status (SES) through education or occupation would predict RTW after work rehabilitation. A third hypothesis was that the process of returning to work would be complex and differ between subgroups of work rehabilitation participants. Cognitions, such as illness perceptions and fear avoidance beliefs may be a matter of beliefs about cure, control, and expectancies, thus of coping. Coping, as defined in the Cognitive activation theory of stress (CATS), was applied in this thesis. In the CATS, coping is defined as positive response outcome expectancies, in contrast to negative response outcome expectancies (hopelessness) or no response outcome expectancies (helplessness). In Norway, comprehensive inpatient work rehabilitation may be offered to individuals on long-term sick leave. Participants in inpatient work rehabilitation programs typically have sick leave diagnoses related to musculoskeletal and mental health complaints, often characterized by non-specific conditions, mostly subjective health complaints, with few objective medical findings. Individuals with subjective health complaints may believe that their complaints are harmful and may therefore try to avoid activities they believe will harm them, such as work. Experiencing distress and poor functional ability may lead to vicious circles of hopelessness and helplessness, i.e. poor coping. Maladaptive illness perceptions and fear avoidance beliefs about work may contribute to prolonged disability and time out of work. The aim of work rehabilitation is to alter such vicious circles through positive experiences and cognitive processes, and facilitate RTW. This is done by interdisciplinary assessments, education, physical activities, and cognitive behavior modifications offered in a combination of individual and group-based sessions. In addition, collaboration with external stakeholders, such as health care providers, the employer, or the local social insurance office (NAV-office) are important elements during work rehabilitation. In this thesis, individual prognostic factors for RTW after work rehabilitation were investigated in three different samples of work rehabilitation participants. Predictive information was extracted from questionnaires and patient journals while information of work and sick leave were measured by self-reports and official register data of The Norwegian labor and welfare administration (NAV). The primary and secondary hypotheses were investigated in the first paper, where the aim was to examine whether health complaints, illness perceptions, fear avoidance beliefs, coping, and education predicted non-working 3 and 12 months after participating in work rehabilitation, and to assess the relative importance and interrelationship of these factors. Logistic regression analysis was conducted. The results showed that fear avoidance beliefs for work were the most important predictor for non-working both at 3 months, and at 12 months follow-up after participating in work rehabilitation. A multiple regression analysis displayed that almost half of the variance in fear avoidance beliefs for work were explained by the amount of musculoskeletal and pseudoneurological health complaints, i.e. tiredness, sadness/depression, and anxiety, and by illness perceptions and education. For illness perceptions, the components concerning perceived duration, consequences, and personal control of the illness were the most important. Coping did not contribute to explain any variance in fear avoidance beliefs for work. In conclusion, high levels of fear avoidance beliefs for work were a strong predictor for non-working after work rehabilitation. However, the intervening mechanisms between fear avoidance beliefs and subsequent avoidance behavior, in terms of avoiding the workplace when sick, are still poorly understood. The primary and secondary hypotheses were investigated in the second paper, where the aim was to test if fear avoidance beliefs for work would mediate the relationships between musculoskeletal and pseudoneurological complaints, functional ability, level of education, and number of days on sickness benefits during 3-year follow-up after work rehabilitation. Structural equation modeling (SEM) was used to test a predefined mediation model for direct and indirect effects between the hypothesized predictors and days on sickness benefits during follow-up. As hypothesized, fear avoidance beliefs for work mediated the effect of musculoskeletal complaints and education on sick leave during follow-up. There was however, no direct effect of musculoskeletal complaints on fear avoidance beliefs, as this relationship was fully mediated by poor physical function, in terms of moving ability and lifting/carrying ability. Fear avoidance beliefs for work did not mediate the relationship between pseudoneurological complaints or mental function, in terms of coping/interaction ability and sick leave during follow-up. Pseudoneurological complaints had a small direct effect, and length of previous sick leave had a strong independent effect on days on sickness benefits after work rehabilitation. In conclusion, the mechanisms involved in the process of returning to work are complex and involve several intervening factors including health and functional ability, education, previous sick leave, and fear avoidance beliefs for work. The second and third hypotheses were investigated in the third paper. Here the aim was to examine if gender, age, diagnosis, occupation, and length of previous sick leave predicted differences in the process of returning to work, in terms of being at work or registered with sickness benefits, and transitions in and out of work and sickness benefits, during a 4-year follow-up after work rehabilitation. Proportional hazard regression analysis was used to explore the probabilities of being at work, or of receiving sickness benefits, or disability pension, and differences in the transitions between any of these states during follow-up. Regression models based on transition intensities detected differences in the risk factors of entering and leaving a given state. For example among women, the lower probability of being at work than men, could be explained by a lower probability of transitions to work, and not by a higher probability of leaving work. In addition, the probabilities of being at work, and of receiving sickness benefits, and disability pension differed between men and women, age groups, diagnostic category, type of work, and previous history of sick leave. Being a female, having diagnoses other that mental and musculoskeletal, having bluecollar work, and receiving long-term sick leave before entering work rehabilitation, increased the risk of not returning to work and of receiving disability pension during follow-up. The use of novel statistical methods made it possible to understand more of the different patterns in or out of work or of receiving sickness benefits, and how the prognosis differed between groups. The results from this thesis show that the process of returning to work after long-term sick leave and work rehabilitation depends on the interplay between multifaceted prognostic factors related to the history of previous sick leave, age, gender, SES, health, function, and cognitions in terms of illness perceptions and fear avoidance beliefs for work. These findings may have implications for selection criteria into work rehabilitation, for tailoring actions during a work rehabilitation program, and may guide follow-up actions aiming at RTW in collaboration with stakeholders outside the work rehabilitation clinic.
Has partsPaper I: Øyeflaten I, Hysing M, Eriksen HR. Prognostic factors associated with return to work following multidisciplinary vocational rehabilitation. J Rehabil Med 2008;40(7):548-54. The article is available at: http://hdl.handle.net/1956/12830
Paper II: Øyeflaten I, Opsahl J, Eriksen HR, Braathen TN, Lie SA, Brage S, Ihlebæk CM, Breivik K. Subjective health complaints, functional ability, fear avoidance beliefs and days on sickness benefits after work rehabilitation – a mediation model. BMC Musculoskeletal Disorders 2016; 17:225. The article is available at: http://hdl.handle.net/1956/12728
Paper III: Øyeflaten I, Lie SA, Ihlebæk CM, Eriksen HR. Prognostic factors for return to work, sickness benefits, and transitions between these states: a 4-year follow-up after work-related rehabilitation. J Occup Rehabil 2014;24(2):199-212. The article is available at: http://hdl.handle.net/1956/9495