Introduction

Although schizophrenia historically was considered to have a deteriorating course (Albert & Weibell, 2019), recovery is considered a feasible treatment goal today (Liberman & Kopelowicz, 2005). Consequently, there is currently greater emphasis on functional outcomes (Bechi et al. 2017) such as employment and social functioning. Despite findings indicating that the majority of people with schizophrenia spectrum disorders (SSDs) want to work (Gühne et al., 2021; Westcott et al., 2015), employment rates remain low. International estimates range from 10 to 20% (Owen et al., 2016), and estimates from Norway are even lower at 4–13% (Evensen et al., 2015; Helle & Gråwe, 2007). To remedy this, vocational rehabilitation programs have been developed to help people with SSDs and other severe mental disorders (SMIs) gain and maintain employment. Such programmes include Individual Placement and Support (IPS) and the IPS-inspired Job Management Program (JUMP) where there is close collaboration between mental health workers and employment specialists to provide vocational support for people with SSDs (Sveinsdottir et al., 2020).

Research on the aforementioned vocational rehabilitation programmes have found that receiving such support has beneficial effects on vocational outcomes (Bond et al., 2020; Falkum et al. 2017). Several studies have also shown advantages of employment in itself for people with SSDs (Drake, 2018). The positive outcomes reported include increased likelihood of symptomatic recovery, financial freedom, social inclusion and increased quality of life (QoL) (Coombes et al., 2016; Raeside & McQueen, 2021; Schennach et al., 2012). A few studies also indicate that employment has a positive influence on self-esteem (Luciano et al., 2014; Negrini et al., 2014), and that competitive employment may be particularly favourable (Bond et al., 2001). Self-esteem refers to how individuals view themselves, both generally (global self-esteem) and in specific domains (Rentzsch et al., 2018). Findings suggest self-esteem is associated with well-being and may predict QoL in individuals with SSDs and other psychotic disorders (Eklund & Bäckström, 2005; Eklund et al., 2018). However, studies often include mixed samples including participants with SSDs together with participants with other diagnosis. Thus, more research is needed to determine the impact of employment specifically for people with SSDs.

Employment specialists play an important role in vocational rehabilitation programs by creating opportunities for their clients to attain and retain work. They develop relationships with employers to help their clients find work related to their preferences and provide support for clients and employers during the vocational rehabilitation (Marino & Dixon, 2014; Moen et al., 2021; Spjelkavik, 2012). As such, employment specialists are an important part of interdisciplinary collaboration in treatment of people with SSDs and are well equipped to assess individual resources and support needs related to employment. They are also in a unique position to observe the impact employment has on the individuals they are supporting.

The aim of this study was twofold. Firstly, we wanted to assess whether vocational activity influenced self-reported self-esteem among people with SSDs. We hypothesised that vocational activity had beneficial effects on self-esteem, and that this effect differed depending on type of vocational activity. Secondly, we wanted to investigate other potential benefits of vocational activity for this group as described by the employment specialists. To this end, we conducted qualitative analyses on post-intervention reports provided by the employment specialists for each individual participant.

Material and Methods

The Job Management Program

The Job Management Program (JUMP) study was a quasi-experimental study investigating the effects of a multisite vocational rehabilitation program provided in collaboration between Norwegian health and welfare services. Adults with SSDs from six counties in Norway were offered a comprehensive vocational rehabilitation program. Participants worked in work placements, sheltered work and/or ordinary employment for 10 months while receiving close support from employment specialists. Sheltered work refers to jobs in sheltered workshops that have reduced work demands. Work placements are in ordinary places of employment and include ordinary work demands, but are funded through employment schemes from the Norwegian Labour and Welfare Administration (NAV). The employment specialists were based in sheltered workshops and provided support for around ten participants each (Falkum et al., 2017; Lystad et al., 2017). The vocational rehabilitation program followed several of the IPS principles, such as following the place-then-train principle, but IPS fidelity was not assessed (Sveinsdottir et al., 2020).

Participants also received individual sessions of either cognitive remediation (CR-group) or cognitive behavioural therapy techniques (CBT-group) for six months. The CR intervention used a computerized training program and emphasised potential transfer values between training and vocational activity. The CBT intervention focused on work-related difficulties. Both interventions were delivered twice a week by the employment specialists, who had undertaken 40 h of training about various aspects of SSDs (e.g. symptoms and treatment) and either basic principles of CR or basic methods of CBT. Allocation to the interventions was randomized on county-level. Participants from counties Oslo, Oppland and Nord-Trøndelag received CBT, while participants from the counties of Vest-Agder, Telemark and Buskerud received CR. For more detail about the different interventions, see e.g. Falkum et al., 2017 or Lystad et al. 2017.

Participants

Participants were recruited between August 2009 and March 2013. They were referred from vocational services, local mental health centres or through self-referral. Inclusion criteria were having a diagnosed SSD, speaking Norwegian well enough to ensure valid neuropsychological test performance, no previous head injury with loss of consciousness for more than 10 min or requiring medical treatment and IQ of 70 or more. Items one to three on the Health of the Nation Outcome Scales (Wing et al, 1998) were used to assess risk of violence and suicide, and severe substance and/or alcohol dependence. A score of three or higher on either item lead to exclusion. Four participants had competitive employment at inclusion but were included because they needed support in order to maintain employment (Lystad, 2016). The subject flow is displayed in Fig. 1.

Fig. 1
figure 1

Subject flow in the JUMP-study

Full size image

Assessments and Measures

Clinical, cognitive and functional assessments were conducted at inclusion, post intervention (i.e. approximately 10 months after inclusion), and 2 years after inclusion (2-year follow-up). In addition, the employment specialists submitted reports to the research group for each participant post intervention. Data from all three measuring points are reported in the current paper, including the qualitative reports. Only measures relevant for the current paper are presented here. All clinical assessments were performed by trained and calibrated clinicians (Bull, 2016).

Clinical Characteristics

Diagnosis was determined according to the Diagnostic and Statistical Manual of Mental Disorders, fourth edition (DSM-IV) using the Norwegian version of the Mini International Neuropsychiatric Interview (M.I.N.I.) PLUS (Sheehan et al., 1998). M.I.N.I. PLUS is a more detailed version of the original M.I.N.I., and has good reliability and agreement with other diagnostic interviews (Sheehan et al, 1997; Lecrubier et al, 1997). The Longitudinal observation, Expert opinion, All Data (LEAD) (Kranzler et al., 1994; Spitzer, 1983) procedure was applied when necessary. Psychotic symptoms were assessed using the Structured Clinical Interview of the Positive and Negative Syndrome Scale for Schizophrenia (SCI-PANSS) (Kay et al, 1987). The structured interview has improved on the already satisfactory inter-rater reliability of the original version of PANSS (von Knorring & Lindström, 1994).

Duration of illness (DOI) was defined as the time from first contact with mental health services for psychotic symptoms (retrieved from hospital records) to inclusion to the study (Evensen, 2016). Main dosages of antipsychotic medication were converted to a measure of defined daily dose (DDD) (World Health Organization, 2011) for statistical purposes.

Functional Outcome

The Global Assessment of Functioning (GAF) scale—split version (Endicott et al., 1976) was used as a measure of general functioning (GAF-F) and psychological symptoms (GAF-S). The two scales are given a score between 1 and 100, with higher scores indicating higher level of functioning. The split version is based on the traditional GAF scale, which is an overall measure of impairment due to mental disorder using a single score. The Norwegian version of the split version of GAF has been validated against other clinical measures (Pedersen & Karterud, 2012).

Employment and educational history was gathered through structured interviews. Highest level of completed education is reported according to the Norwegian educational system, categorized as less than primary school (< 9 years), primary school (9 years), high school/vocational school (12 years), college (13–15 years), and university (17 + years). Employment history is reported as whether or not the individual had any experience with vocational activity before inclusion (yes/no) and months of previous working experience fulltime and part-time. Employment specialists collected information about current vocational status at all assessment points, and registered hours spent in vocational activity per week (Bull, 2016). Vocational activity was reported as no vocational activity, sheltered work, work placement, or competitive employment (including higher education).

Intellectual Functioning

Intelligence quotient (IQ) was assessed with the Wechsler Abbreviated Scale of Intelligence (WASI) two-subtest estimates of IQ (WASI, 2007). The Norwegian version appears to be a reliable tool for screening intellectual functioning. However, Norwegian norms are not available and use of US norms (as in the current study) may cause slight overestimation of intellectual functioning (Siqveland et al., 2014).

Self-Esteem

The Norwegian version of the Rosenberg Self-Esteem Scale (RSS) (Rosenberg, 1965) was used to measure global self-esteem. This is a 10-item self-report questionnaire where each item has four alternative responses ranging from strongly agree to strongly disagree. Half or the items are negatively worded and these scores were reversed before the sum scores were calculated. Each item is given a score from 1 to 4, therefore the total score ranges from 10 to 40 with higher scores indicating higher self-esteem (Rosenberg, 1965). The factor structure and internal reliability of the RSS has been replicated across various cultures and languages (Schmitt & Allik, 2005). Findings from the Swedish version suggest good internal consistency, validity, at least some sensitivity to change, and that it can be used to measure of global self-esteem (Eklund et al., 2018).

Statistical Analyses

IBM SPSS Statistics for Windows (version 27) (IBM Corp, 2020) was used for all statistical analyses. Levels of significance were set at p = 0.05. All tests were two-tailed. The effect sizes were interpreted by the guidelines developed by Cohen (1988). Thus, eta squared (η2) was categorized as large, moderate and small effects at the values 0.14, 0.06 and 0.01 (Pallant, 2020). A previous study found no significant baseline differences between the CR and CBT groups except for gender and DOI at baseline, nor any differences post intervention or at 2-year follow-up regarding how many had employment, type of employment or hours worked per week (Evensen et al., 2017). The two groups were therefore treated as one sample in the current paper.

Standard multiple regressions were conducted to assess whether degree of involvement in vocational activity predicted differences in self-esteem post intervention and at 2-year follow-up. RSS sum scores were used as dependent variables, while gender, age, education, baseline sum scores on the PANSS symptom scales and hours worked per week were included as independent variables.

To investigate whether different types of vocational activity affected self-esteem differently, a mixed between-within subjects analysis of variance (ANOVA) was conducted. The predictor variable was type of vocational activity post intervention. Gender, age, education and baseline sum scores on the SCI-PANSS scales were included as covariates, while RSS sum scores on the three measuring points were the dependent variables. The four participants with competitive employment at inclusion were excluded from this analysis to prevent potential effects of competitive employment at baseline from affecting the results.

Qualitative Analysis

Employment specialists in the JUMP study wrote a post-intervention report for each participant (N = 121) describing key events during the intervention, their experience of collaborating with different agencies, their evaluation of the JUMP method, the individual participants’ work affiliation and ability to work. They also included additional information that was important to support the participant (Template Appendix A). The discrepancy between the number of reports and the total number of participants is due to a combination of factors such as drop-out and clerical error.

The reports were anonymized and marked with participant id/numbers. They were investigated through a qualitative approach based on thematic analysis as described by Braun and Clarke (2006) and Joffe (2011). The reports were read several times and key words noted to generate the coding frame. The coding frame included information regarding benefits of the intervention, the intrinsic value of employment, the importance of receiving a salary, factors that contributed to a positive development and challenges during the intervention. With further analysis the included material was reviewed and reorganized into two meta-codes termed “beneficial results of vocational rehabilitation” and “processes of positive development”. Examples from the final coding frame are displayed in Table 1.

Table 1 Examples from the final coding frame used in the qualitative approach
Full size table

The included segments were assessed multiple times, searching for patterns. Experiences that were mentioned in several reports and were descriptive of multiple codes were regarded as themes. For example, the codes “acknowledgement from others” and “belonging to a group” both included segments describing that participants had beneficial interactions with other people due to vocational activity. This was included in the theme “positive interpersonal experiences”. The final themes were broadened skillsets, positive interpersonal experiences, increased well-being and increased self-esteem.

The first author conducted the initial coding. Multiple authors discussed how to interpret the themes. Translations of the included material from Norwegian to English were made after the qualitative approach through collaboration among the authors. Some of the segments were rephrased because they included expressions that are not used in English. The original quotes and corresponding translations are included in Appendix B.

Ethical Considerations

All participants provided written informed consent. The JUMP study was approved by the Regional Committee of Medical Research Ethics and the Norwegian Data Protection Authority. ClinicalTrials.gov Identifier: NCT01139502.

Results from the Statistical Analyses

Demographic and clinical characteristics of the sample are presented in Table 2.

Table 2 Demographic and clinical characteristics of the sample
Full size table

Table 3 shows mean RSS scores and standard deviations by vocational status for all measuring points. Because of missing data the sample sizes differs somewhat from what is reported in Fig. 2.

Table 3 Mean scores and related standard deviations on the Rosenberg Self-Esteem Scale at all measuring points, grouped by vocational status
Full size table
Fig. 2
figure 2

Number of participants in competitive employment, work placements, sheltered work or no vocational activity at the three measuring points

Full size image

Standard Multiple Regressions

In the standard multiple regressions there was a significant effect of hours worked per week on the RSS sum score post intervention (Beta = 0.19, B = 0.11, SE = 0.05, p = 0.029). The model was significant and explained 17.80% of the variance (F = 3.57, p = 0.002). None of the other independent variables contributed significantly to the model. At 2-year follow-up, the model was no longer significant (F = 1.74, p = 0.108).

Mixed Between-Within Subjects ANOVA

In the mixed between-within subjects ANOVA, the effect of vocational status post intervention on the RSS sum scores across the three measuring points was not significant F(6, 162) = 1.72, p = 0.119. Due to missing data, only 92 participants were included in this analysis. Of these, six participants had competitive employment, 43 had work placements, 31 had sheltered work and 12 had no vocational activity post intervention. Because methodological challenges related to different group-sizes may potentially cloud actual effects, we report the different trends between the groups regarding the development in self-esteem. Participants with competitive employment showed an increase in RSS scores throughout the study, while persons with work placements had an initial increase that stagnated after the intervention ended. The group with sheltered work had the smallest increase in RSS scores, and the participants with no vocational activity had a decrease in scores. Figure 3 illustrates the estimated marginal means for the RSS sum scores divided by vocational status post intervention.

Fig. 3
figure 3

Estimated marginal means on the Rosenberg Self-Esteem Scale at all measuring points, grouped by vocational status post intervention (N = 144)

Full size image

Results from the Qualitative Approach

The employment specialists reported several benefits of vocational rehabilitation for the participants. The participants showed an increased belief in themselves as well as improvement in work-related skills and well-being. These improvements were attributed to vocational achievements and positive interpersonal interactions.

Broadened Skillsets

The improvement in skills reported for some participants was regarded as a positive result in itself, as well as being a process where developments in one area contributed to advances in other areas. The reported improvements included improvements in cognitive functioning (for both the CR- and CBT-group), increased ability to cope with stress and to interact with other people. The following is an example from one of the reports:

The vocational training has had a very good effect on the ability to pay attention, interaction, and the social aspect. The participant is much less sensitive to external events (which can take attention away from the tasks at work), experiences increased capability and is much more comfortable in interactions with other people.

Such changes in competencies were believed to increase the participants’ chance of succeeding with later work-related efforts and apparently made them more comfortable with certain tasks and/or situations in their current vocational activity. Moreover, they seemed to increase the participants’ well-being. For example, it was reported that improved cognitive functioning could contribute to positive changes in social life by making the participants more competent and comfortable participating in social situations.

Positive Interpersonal Experiences

Employment specialists considered positive interpersonal experiences to be part of a beneficial process that supported the favourable changes during vocational rehabilitation. Such experiences were associated with increased self-esteem, improved QoL and well-being in several reports. Importantly, positive results were often attributed to a combination of multiple processes. As such, experiencing achievement and positive interpersonal interactions appeared to be particularly potent when combined. One employment specialist wrote that the participant experienced capability and a sense of community at work. He expressed that employment was very important for him and gave weekly feedback about these experiences.

The following excerpt from another report also illustrates the significance of experiencing both competence and social support: “She receives feedback/acknowledgement from other people. She described an experience of achievement, learning new things and being valued. This influenced her dignity and perception of herself.”

These kinds of changes were described for various reports despite the fact that several participants changed vocational activity up to multiple times during the intervention period.

Increased Well-Being

Another aspect that was considered a potential benefit of vocational rehabilitation was improved well-being. This was described as an increase in optimism or hope, more expression of pleasant emotions and bettered social lives. The following is an example: “Gradually, [he] has expressed some more optimism, and belief that he has the possibility to create a better future through work and an identity as a “valuable citizen” among other things.” One employment specialist wrote the following after speaking with a participants’ therapist in the mental health services: “…they also say that he has developed a lot during the time in the project. He is much happier and more content than they have seen him before. He shows greater capability and self-confidence than before.”

Increased Self-Esteem

In terms of the participants’ view of themselves, some appeared to have a more positive view of their abilities after the intervention. The following excerpt illustrates this: “Participant […] made very nice progress in the project. She had good resources to begin with but has built her self-confidence and belief in herself through the project.” Importantly, such results were also reported for participants who had experienced difficulties during the vocational rehabilitation. In this regard, one employment specialist wrote: “Despite many interrupted internships during the project, the participant has a belief that she will be able to manage some tasks in competitive employment (albeit in a reduced work-time percentage)”.

Experiences with achievement was considered to contribute to advantageous results. In regard to this, one employment specialist wrote: “The project gave him the opportunity to enter competitive employment. Through experiencing that he could cope with working life, it gave him self-confidence.” Another report reads: “He describes that doing and experiencing that he is capable has given him confidence in his ability. …exposing himself, stepping out of his comfort zone and experiencing that he survived doing so, has been the largest success factor.”

As in the previous example, an aspect that was mentioned in several instances was that the participants challenged themselves. As such, their experiences with being capable may have been related to situations they initially found daunting.

Discussion and Conclusion

The first aim of the current study was to assess whether vocational activity influences self-esteem in individuals with SSDs. We found that hours spent in vocational activity had a positive effect on self-esteem. This finding has two important implications. Firstly, in line with our hypothesis, it suggests that being engaged in vocational activity may positively influence self-esteem for people with SSDs. Self-esteem appears to be associated with various aspects of mental health—e.g. psychotic symptoms (Romm et al., 2011; Smith et al., 2006)—and has been suggested as target of intervention for reduce risk of suicide among people with SSDs (Jian et al. 2022). Hence, other aspects of mental health may also profit from improved self-esteem. Mental health services should therefore support individuals’ goals regarding employment. Secondly, a potential positive impact of vocational activity on self-esteem may depend on the degree of exposure. Thus, it is not simply having some form of vocational activity that is important, but also the amount of time spent in said activity. This is an important consideration to have in mind when searching for vocational activity.

Regarding our hypothesis that type of vocational activity would have differential effect on self-esteem, the picture is less clear. The lack of statistical significance in the results is probably due to methodological issues and insufficient statistical power as the group sizes were variable and the smallest group (competitive employment) only had six participants. There was a trend in the data suggesting ordinary employment may be particularly beneficial, but more research is needed to firmly establish this. Regardless, based on the results there were no indications that ordinary employment would be less beneficial than other forms of vocational activity for self-esteem. This is in line with previous research that has not found employment to be harmful for the mental health of this group when compared to unemployment (Luciano et al., 2014).

The second aim of the study was to explore potential advantages of vocational activity for people with SSDs as described by employment specialists. The qualitative results indicate that taking part in vocational activities hold multiple benefits, and that these potentiate each other. Learning new skillsets and experiencing that one is capable builds self-confidence and self-esteem, as do positive social interactions. This is in keeping with earlier research showing that employment may provide opportunities for positive interpersonal experiences, improve well-being, and contribute to an increase in skills (Dunstan et al., 2017), and that vocational activity may improve QoL (Marwaha & Johnson, 2004).

Taken together, the findings from this mixed methods approach suggest vocational activity may have beneficial effects for people with SSDs that may contribute to improve QoL, including increased social and vocational skills, well-being, and self-esteem. With regard to competitive versus other types of employment it has been suggested that receiving wages may be related to a greater increase in QoL as compared to not receiving wages (Bryson et al., 2002), though conflicting findings have been reported (e.g. Luciano et al., 2014; Ådnanes et al., 2018). This discrepancy may partly be due to differences in outcome measures or problems with measuring these consepts due to their subjective nature (Luciano et al., 2014).

It should be noted that other factors may also have contributed to the findings in this study. For instance, the comprehensive support provided by the employment specialists may, in itself, have had beneficial effects for the participants. Findings suggest that clients appreciate the close support from the employment specialists (e.g. Chen & Lal, 2020; Fyhn et al. 2020). A previous study on IPS in Denmark also found that some of the participants experienced the IPS method as contributing to bettered self-esteem (Gammelgaard et al. 2017). Hence, the reported results may reflect the combined effect of vocational rehabilitation and the experience of receiving support. In addition, site specific effects – such as different cultures and attitudes in the different counties and individual differences between employment specialists – may have influenced the reported results. Although there is limited research on the specific effects of employment specialists’ characteristics, their skills, knowledge, and competencies have been found to be associated with their clients’ outcomes (Bakkeli et al., 2020).

Strengths

The mixed-methods approach provides a broader perspective on the effects of vocational activity, incorporating both reports from the employment specialists as well as self-reported self-esteem data. The use of standardized measures such as RSS enables comparison between studies, while the qualitative approach allows for a detailed investigation of the participants’ development. The employment specialists are also an important voice in understanding and optimizing vocational support for this population.

The therapeutic interventions (CBT and CR) provided in JUMP may have made the participants better equipped to handle the demands of vocational activity and to uphold vocational activity over time.

Limitations and Directions for Future Research

Though studies on the RSS have found it to be a valid assessment tool for measurement of global self-esteem, it has been suggested that its sensitivity to change should be more thoroughly investigated. Some sensitivity to change has been demonstrated, but it is uncertain how large change is needed for it to be detected by the questionnaire (Eklund et al., 2018; Schmitt & Allik, 2005). Thus it is possible that the RSS was not sensitive enough to detect a potential change. Although using the mean RSS scores is the most common use of the RSS, it has been proposed that categorizing RSS scores into levels of self-esteen may provide a more accurate and sensitive description of self-esteem (García et al. 2019). In addition, vocational activities may encapsulate a specific domain of self-esteem, which we were unable to disentangle with the RSS. As such, future studies may benefit from utilizing other methods such as including domain-specific measures of self-esteem.

Participants may have gone through several changes in vocational status throughout the study. Changes where participants gained vocational activity reflect that the interventions were successful. Nevertheless, they have statistical implications in terms of what types of analyses may be carried out. In addition, the participants with a certain vocational activity (e.g. competitive employment) may not have had this long enough for effects to occur before assessments were made. Insufficient statistical power due to small group sizes may also have disguised potential effects. Future studies may benefit from larger sample sizes, as well as closer monitoring of potential fluctuations in self-esteem and vocational activity.

One person conducted the qualitative assessments and that individual will naturally influence the interpretations. There may also be a bias in the material as the person that had provided the support for the participant served as informants. Consequently, adverse events may have been underreported. The template for the reports did not specifically ask what the employment specialists considered benefits of vocational activity. To counteract this limitation, interpretations were only made from explicit statements where changes were attributed to vocational activity or rehabilitation. Future research may profit from more direct questioning about the consequences of gaining vocational activity and the possible significance of receiving wages.