Introduction
Return-to-work (RTW) is conceptually framed as a complex process involving the individual and the environment (Nielsen et al.
2018; Young et al.
2005). As sick leave due to common mental disorders (CMDs) impacts the individual, the workplace, and society at large (OECD
2018) knowledge that can inform strategies for a sustainable RTW is important for the individual, employers, health care professionals, and policymakers. Earlier theoretical models (Demerouti et al.
2001; Karasek and Theorell
1990) and research (Aronsson et al.
2017; de Vries et al.
2018; Harvey et al.
2017; Duchaine et al.
2020) have identified psychosocial demands at work, such as high psychological job demands, low job control, and low social job support, as associated with the development of CMDs and resulting sick leave. Similarly, work-home interference, i.e. work-to-home interference (WHI) and home-to-work interference (HWI) have been associated with CMDs and sick leave (Blom et al.
2014; Svedberg et al.
2018). To date, less is known about the impact of psychosocial work factors, work-home interference, and the time to RTW (cf. de Vries et al.
2018; Joosen et al.
2021; Nybergh et al.
2020). The employer has a central role to accommodate the employee in RTW after sick leave due to CMD (Gensby et al
2019; Corbiere et al.
2020; Joosen et al.
2021; Nybergh et al.
2020). Understanding associations between RTW, psychosocial work factors, and work-home interference has the potential to guide organisational measures for RTW.
To date, a few studies have found associations between high psychological job demands (Ekberg et al.
2015; Flach et al.
2012; Netterstrøm et al.
2015) and low job control (Flach et al.
2012; Netterstrøm et al.
2015) and RTW after sick leave due to CMDs. However, in a recent scoping review, de Vries et al. (
2018) found the evidence for many factors influencing RTW, including psychological job demands, to be insufficient. Similarly, the evidence for the association between low social job support in the workplace and RTW is inconclusive (Flach et al.
2012; Netterstrøm et al.
2015; Sikora et al.
2022). A good balance in everyday life is perceived as essential for a ‘successful’ RTW, while interference between demands at work and in private life has been identified as a barrier (Hees et al.
2012; Joosen et al.
2021; Nybergh et al.
2020). Because of the gendered structure of demands in private life, the negative effect of domestic strain on sick leave and RTW is greater among women (Holmlund et al.
2022; Nybergh et al.
2020; Östlund et al.
2004). To date, there is limited evidence of an association between WHI and HWI and time to RTW. Only recently, Sikora et al. (
2022) found that employees receiving in-patient treatment for their CMDs and reporting work-home interference as associated with a longer time to first and full RTW.
When investigating RTW it is important to consider the administrative setting (Krause et al.
2001a; Ståhl et al.
2018). The associations between work factors, work-home interference, and RTW can differ between countries or between phases of the RTW process depending on cultural and legal systems (Krause et al.
2001a,
b). In Sweden, the risks associated with work factors and work-home interference for both genders are interesting because labour market participation is nearly equal for men and women. Moreover, factors associated with time to RTW are interesting because of policies implemented in Sweden to provide efficient RTW. Examples of this are employers’ responsibility to design an RTW plan within 30 days of sick leave (if the employee is expected to be on sick leave for at least 60 days) and time intervals for eligibility for sick leave. In Sweden, after > 6 months of sick leave, an assessment of eligibility for any job in the entire labour market is carried out, whereas at 0– ≤ 6 months an assessment is carried out of the person’s ability to take on work at their present workplace (exceptions are made in cases of serious illness). A better understanding of work factors and work-home interference and time to RTW can lead to improved RTW practices. Moreover, a phase-specific analysis may show time-dependent influences of factors in specific settings (Krause et al.
2001a,
b).
This study aimed to (1) examine the time to first full RTW, and (2) investigate whether psychosocial work factors and work-home interference are associated with time to first full RTW after sick leave due to CMDs. A post hoc analysis investigated whether the associations differ between employees returning to work between 0 and ≤ 6- and between > 6 and 12 months.
Discussion
This study showed that 80.9% of the participants achieved a first full RTW during the 12-month follow-up, with the majority (59.3%) within the first six months. High psychological and emotional job demands, high WHI, and low social job support at baseline were independently associated with a longer time to first full RTW. Significant differences between time intervals (0– ≤ 6- and > 6–12 months) were found in a post hoc analysis of the associations between low job control and high emotional job demands, and time to first full RTW. Low job control was associated with a shorter time to first full RTW between 0 and ≤ 6 months and with a longer time to first full RTW between > 6 and 12 months.
A direct comparison across studies of time to RTW after CMDs is difficult due to differences in administrative settings and design. Our results are in line with a recent Swedish RCT conducted in occupational health services among employees absent from work due to CMDs. Here, 82% had returned to their ordinary working hours during a 12-month follow-up (Keus van de Poll et al.
2020). A German study of patients receiving in-patient care for CMDs found that 93.7% returned to work after 18 months of follow-up (Sikora et al.
2022). In a Danish study of employees on sick leave with CMDs enrolled from PCCs, using an outcome of full-time work at 12 months, 67% returned to full-time work after 1 year (Netterstrøm et al.
2015). The Danish results correlate with our last point estimate. In our study, as in the study by Keus van de Poll et al. (
2020) and Sikora et al. (
2022), the outcome was a first event of working ordinary working hours for an uninterrupted period of at least four weeks/28 days. The different results might indicate frequent sick leave episodes after the first event of RTW, which is a known problem for employees with CMDs (Arends et al.
2014). The results point to the relevance of longitudinal studies for evaluating factors associated with RTW. Moreover, the results indicate a need for at-work strategies in RTW interventions for the maintenance of work functioning after the first RTW (Arends et al.
2022).
The association between high psychological job demands and a longer time to first full RTW in our study has been demonstrated in previous studies (Flach et al.
2012; Netterstrøm et al.
2015). The results also align with qualitative research reporting high psychological job demands and inadequate accommodation as barriers to RTW after sick leave due to CMDs (Joosen et al.
2021; Nybergh et al.
2020). Moreover, we found independent associations between high emotional job demands and a longer time to first full RTW, earlier shown in qualitative research (Nybergh et al.
2020). As in our study, Netterstrøm et al. (
2015) found low social support at work to be associated with RTW after sick leave due to CMDs. However, in a recent study Sikora et al. (
2022) found no such association. Job resources, including high social support and high control, are typically described as a buffer for high psychological job demands on CMDs (Bakker and Demerouti
2017; Karasek and Theorell
1990). However, it has also been suggested that experiencing high psychological job demands over time probably affects employees regardless of resources, and that it is, therefore, important to reduce psychological job demands per se (Bakker and Demerouti
2017; Fagerlind Ståhl et al.
2018; Karasek and Theorell
1990). Moreover, in our study high WHI was independently associated with a longer time to first full RTW. Because WHI is associated with work-related outcomes (Amstad et al.
2011), reducing high psychological job demands may reduce the risk of negative spill-over of job demands into people’s private lives, and benefit the RTW process.
The post hoc analysis showed differences between time intervals for the associations between low job control and high emotional job demands and time to first full RTW. Investigating phase-specific factors is relevant because of the developmental nature of the RTW process (Krause et al.
2001b; Young et al.
2005) and the anticipated impact of RTW policies (Krause et al.
2001a,
b) such as time intervals on eligibility for sick leave. Due to the small subgroup (
n = 35) between > 6 and 12 months, the results of the phase-specific analysis must be interpreted with caution. Although, the differences suggest that future studies should be carried out to determine whether specific factors merit particular attention at specific points in the RTW process.
A strength of the study was the longitudinal design and the follow-up by repeated text messages. Although self-reported sick leave outcomes entail a risk of bias, text messaging has demonstrated high compliance (Axén et al.
2020), and the response rate was 90.2%. Another strength was the use of measures validated in a Scandinavian setting (Berthelsen et al.
2020; Sanne et al.
2005; Wännström et al.
2009). Using single items from the QPS-Nordic is, however, less robust and work-home interference was only investigated with two single items. The recruitment of participants from different PCCs in one Swedish region can help us to extrapolate the results to other Swedish settings.
One of the study’s limitations is the possibility of response and recall bias because of the self-reported data collected retrospectively. Baseline data were collected when the employee was on sick leave. Moreover, the eligibility criteria ‘having to accept employer involvement’, might imply that included participants had a positive perception of their employer and decreased the variance in, and affected the outcome of, fair leadership. Moreover, we do not know exactly when the first full RTW occurred. Because text messages were sent every fourth week, precision was limited. This uncertainty was incorporated into the estimation to reduce the risk of bias. Additionally, missing values preceding the first positive answer were accounted for by widening the censoring interval. Missing values followed by a negative answer were interpreted as a negative answer, ensuring that we did not underestimate the time to the first full RTW. Given the sample size, only the main factors associated with RTW after sick leave due to CMDs were introduced in the adjusted model, the possibility of other confounders needs to be considered in the interpretation of the results. Moreover, because a large majority of the sample were female employees born in Sweden, sub-group analyses based on gender or ethnicity were not possible.
The importance of organisational measures to improve RTW is previously highlighted (cf. Gensby et al.
2019; Joosen et al.
2021; Nielsen et al.
2018). This study implies that workplace accommodations for the employee in rehabilitation after sick leave due to CMDs may well focus on reducing high psychological and emotional job demands, for example by reducing workload, work pace and conflicting/or emotional job demands. Reducing high psychological job demands may also be important to reduce WHI and foster a balanced everyday life for employees (Amstad et al.
2011). In addition, the RTW support offered by primary care and occupational health services should include assessments of the individual’s work and private life to design a sustainable rehabilitation strategy. At the policy level, it is important to acknowledge the adverse effects of high psychological and emotional job demands and high WHI on time to RTW. The estimate for emotional demands in this study is close to one, however, the result corresponds with research showing employees’ experience of emotional demands as a hindrance in RTW (Nybergh et al.
2020) indicating the clinical importance of the results. Moreover, in countries with nearly equal labour-market participation and a gender-segregated labour market, it is important to recognize the possible risks of inequalities for women earlier shown in private life (Eydal et al.
2015) and work domains, such as among care workers (Aronsson et al.
2021; Lidwall et al.
2018). We recommend that future research investigate the clinical importance of emotional demands and the mechanisms of work-home interference in more detail, including the potential differences between social groups. More research is needed to investigate program activities in RTW interventions aiming to reduce high psychological and emotional job demands and high WHI, to further our understanding of the effectiveness of such activities.
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.