Abstract
BACKGROUND:
Long term sick leave (SL) is increasing in Europe, several countries have legislative initiatives to reduce long-term absenteeism.
OBJECTIVE:
We evaluated the impact of a legally defined return-to-work (RTW) trajectory on the RTW of employees on sick leave in Belgium.
METHODS:
This was a retrospective register-based cohort study of employees (n = 1416) who followed an RTW trajectory in 2017. We linked workers’ data from a prevention service with social security data. By multinomial logistic regression, we analysed which characteristics predicted the RTW with the same or another employer.
RESULTS:
One year after their RTW trajectory, 69.2% of the 1416 employees did not RTW; 10.7% returned to work with the same employer and 20.1% with a new employer. Duration of SL was an important predictor for the RTW with both the same employer and another employer. The odds of RTW were lower when the SL duration was > 6 months compared to < 6 months. Marital status, organization-size, and the occupational physician decision had a significant impact on the RTW with the same employer. Age and who initiated the RTW-trajectory were important predictors on the RTW with another employer.
CONCLUSIONS:
Overall, 30.8% of employees returned to work after their RTW trajectory. A one-size-fits-all approach is not recommended. A stepped approach with an early, informal start of the RTW process is advised. When employees or employers fail to initiate the RTW on their own, a legally defined RTW trajectory could be useful. In particular, RTW with another employer seemed a positive effect of the RTW-trajectory.
Background
The number of people on sick leave increases every year. In the European Union (EU-28) in the year 2019, 4.3% of the working-age population (age 20–65 years) is inactive due to sickness or disability [1]. The reasons for the increase in long-term sick leave (SL) are multi-layered. On the one hand, there is the ageing of the population, the increased retirement age, and the increased employment rate of women in the labour market [2]. On the other, there is a shift in the conditions that lead to long-term SL and disability. The most frequent causes of long-term SL and work disability are mental and musculoskeletal disorders [3, 4]. These illnesses result in longer periods of absence and difficulties in a return to work (RTW) [5, 6]. Furthermore, psychosocial and work-related factors, such as social support or work pressure, are linked to (long-term) disability [7, 8].
Long-term SL is problematic for multiple reasons. It has severe negative consequences on an individual level. Not being able to work has a negative impact on overall health and wellbeing because working can contribute to mental and physical health and quality of life [9, 10]. Also, the longer the SL, the lower the chance of a successful RTW [11]. Therefore, early interventions are necessary and crucial for a successful RTW [12]. In addition, the impact at the employer and societal level is substantial [13,14, 13,14]. Although the exact costs of SL and disability are lacking [15], in most EU-28 countries, about 5.9% of the total Gross Domestic Product is used to pay for sick leave and disability [3]. In some countries, the expenditures seem to be decreasing (e.g., Denmark and Spain), but in other countries they are increasing (e.g., Germany and Belgium) [3].
To tackle SL and its negative consequences, there is a body of research on the barriers to and facilitators of RTW [16–20]. Different types of RTW interventions such as early, ergonomic, multidisciplinary, time-contingent and activating interventions have been found effective in reducing the SL and helping people get back to work in specific situations [21, 22]. In general, implementing multidisciplinary, workplace-based interventions is recommended [23, 24]. Nevertheless, the effect of specific RTW programmes on the SL duration and the RTW in general remains unclear [22, 25].
Legislative initiatives are important to create a context to promote RTW [26]. Although different European countries have legislative actions for the RTW, their effect on SL and RTW remain uncertain [27]. In Belgium, in 2017, the government implemented an official, legally defined RTW trajectory to promote and facilitate the RTW [28]. To evaluate the impact of this legislative initiative on RTW, we conducted a retrospective register-based cohort study of employees who followed this official, legally defined RTW trajectory. The aims of our research were twofold: 1) to investigate how many employees RTW with the same or another employer after an official, legally defined RTW trajectory and 2) to study the characteristics that predict the RTW with the same or another employer for employees who follow an official, legally defined RTW trajectory. We hypothesized that this official, legally defined RTW trajectory could contribute positively to the RTW of employees on long-term sick leave.
Methods
Belgian RTW setting
In 2020, more than 470,000 people of working age in Belgium were sick for≥1 year (data retrieved from the National Institute for Health and Disability Insurance) [2]. Different legal RTW options exist for employees on sick leave in the Belgian context. An employee can independently and spontaneously RTW, they can be obliged to RTW by the medical advisor, or they can also consult an occupational physician (OP) who can evaluate the (dis)ability to work. The OP works for a Service for Prevention and Protection at Work (SPPW). An SPPW uses prevention advisors in different domains of prevention (e.g. occupational health, psychosocial aspects, ergonomics) and is responsible for the health and wellbeing of employees [29]. Every employer in Belgium must have an SPPW. An employer can choose to organise an SPPW internally or, more commonly, be affiliated with one of 10 available external SPPWs (ESPPWs).
The consult with the OP can be an informal RTW consultation, but since January 2017, Belgian legislation has put into practice an official, legally defined re-integration trajectory (RIT) to promote and facilitate the RTW of employees on sick leave [28]. The OP of the SPPW is the coordinator of this RIT. An RIT can be initiated, depending on different conditions, by the employee or the treating physician, the employer, or the medical advisor of the mutual insurance organisation. Mutual insurance organisations are public interest institutions that are responsible for granting sick-leave benefits in Belgium. When an RIT is initiated, the OP must decide on the remaining work capacity of the employee. According to the existing legal framework, the OP assigns the employee to one of 5 categories, based on their estimation of the future possibilities to RTW. These possibilities range from being able to RTW with other or modified work to being permanently unable to perform the job with the current employer. In the latter case, when there are no possibilities to RTW with the current employer, the contract of the employee can be ended due to medical force majeure. A detailed explanation of the RIT and its categories can be found in Appendix 1.
Data source
To evaluate the impact of the RIT on the RTW of employees on sick leave, we conducted a retrospective register-based study using patient records data from Group IDEWE, one of the largest ESPPWs in Belgium. These data were linked to social security data that were made available through the Crossroads Bank for Social Security (CBSS). The CBSS is a Belgian federal government organization that enables, among other things, decentralized storage, and secure and organized exchange of data between social security institutions.
Sample
We included all employees who followed an RIT with an OP of the ESPPW ‘Group IDEWE’ in 2017 in the study. Group IDEWE provided variables related to the RIT for each employee: their national registration number, the start date of the RIT, the initiator of the RIT and the outcome of the RIT (OP decision). We excluded employees with double records because the anonymisation process in the later stages would disallow correctly interpreting the linked results of double records.
Variables and measurements
Social security data for the employees according to their national registration number were requested from CBSS. The variables requested were sex, age, having children, marital status, working regime, sector, employer registration number, organization size of the employer and duration of the SL. Those variables were collected by CBSS at different regional and federal government institutions (e.g., the national sickness fund organizations, etc.). After linking the data on a person-identifiable level, the exported data results contained only anonymised data in order to comply with privacy regulations. A total of 1787 unique persons were identified in the sample, of which 371 were lacking information on the study variables. We excluded employees with missing data on one or more of these variables from the study sample.
Analysis
Characteristics of participants are reported as number (%) or mean. To study the predictors of the RTW with the same or another employer, we used multinomial logistic regression. The dependent variable was categorical, with 3 categories: no RTW, RTW with the same employer, and RTW with another employer. The RTW was defined as having returned to work, fully or partially, 4 quarters after the start of the RIT. A quarter is defined as 3 consecutive calendar months. The reference category for the outcome was no RTW. Odds ratios (ORs) and 95% confidence intervals (CIs) were estimated. The variables retrieved from CBSS and Group IDEWE were added to the model as independent factors to investigate their predictive value on the outcome. The model met the necessary assumptions for multinomial logistic regression: each independent variable had one value for each case, the dependent variable was categorical and there were no reasons to assume multicollinearity between them. Using the Variance Inflation Factor (VIF) test, multicollinearity was tested. VIF values greater than 4 indicate multicollinearity. The VIF values in this case were all less than 2. P < 0.05 was considered statistically significant. For statistical analysis R-studio version 4.0.2 was used.
Ethics
The study was approved by the Education-Support Committee (ESC) of Health Care Management and Policy (KU Leuven) (MP016655, 25/10/2020). Because of the retrospective, anonymized nature of the data, no further ethical approval was deemed necessary by the ESC. CBSS also approved our data request. The dataset provided by Group IDEWE was sent to CBSS in an encrypted file. CBSS linked the sample to their data sources and anonymized the dataset. An anonymized sample of this dataset was then delivered to the researchers of KU Leuven. According to the sample provided by CBSS, an analysis was pre-programmed and then performed on the local server of the CBSS for the entire population. Only aggregated results were exported and further used in the analysis to remain compliant with privacy regulations.
Results
The final sample consisted of 1416 RITs (employees with no missing data). More than two thirds of the employees (69.2%, n = 979) did not RTW; 10.7% (n = 152) returned to work with the same employer, and 20.1% (n = 285) with another employer (Table 1).
Characteristics of the employees in the study sample by return-to-work (RTW) outcome
Characteristics of the employees in the study sample by return-to-work (RTW) outcome
RIT: re-integration trajectory; OP: occupational physician; SL: sick leave.
Multinomial logistic regression of the study sample by outcome: RTW with the same employer or another employer (reference: no RTW)
OR: odds ratio; CI: confidence interval; Reference: reference category; SL: sick leave.
The characteristics of the sample according to their outcome group are in Table 1. Many employees were women (66.0%), had children (61.8%) and were living with a partner (69.8%). The mean age was 40 to 44 years old. Almost half of the participants (45.8%) worked in an organization with more than 200 employees. More than half (53.3%) were absent from work for > 6 months.
The results of the multinomial logistic regression are in Table 2. Four factors significantly predicted the RTW with the same employer: marital status, size of the organization, the OP decision and the SL duration. The odds of the RTW with their employer were reduced for employees without a partner (OR = 0.53, 95% CI: [–1.14, –0.15]]), employees who worked for smaller organizations (OR = 0.44 95% CI: [–1.38, –0.26] for < 50 employees; OR = 0.53, 95% CI: [–1.17, –0.08] for 50–200 employees), and employees who had been absent for > 6 months at the start of their RIT (OR = 0.25, 95% CI: [–1.83, –0.97]). When the OP decided that RTW was possible, the odds of the RTW with the same employer were 84-fold increased (95% CI: [3.54, 5.31]). The odds of the RTW with the same employer were also increased when the OP decided that RTW was only possible with a definitive other or modified job (OR = 10.71, 95% CI: [1.45, 3.29]), and when there was no decision possible at the start of the RIT (OR = 48.17, 95% CI: [2.96, 4.79]).
Three factors significantly predicted the RTW with another employer: age, the initiator of the RIT and the SL duration. The odds of the RTW with another employer were reduced for older employees (OR = 0.79, 95% CI: [–0.31, 0.35]), employees whose RIT was initiated by the employer (OR = 0.67, 95% CI: [–0.75, –0.06]), and employees who had been absent for > 6 months at the start of their RIT (OR = 0.34, 95% CI: [–1.38, –0.80]).
The SL duration was the only factor that significantly predicted both RTW with the same and another employer. The proportion of RTW was lower with SL duration > 6 months than < 6 months (Fig. 1).
The proportion of no return to work (RTW), RTW with the same employer, and RTW with another employer according to the sick leave duration.
To investigate the impact of a legislative initiative taken to promote RTW after (long-term) SL, we conducted a retrospective register-based data analysis. We hypothesized that the introduction of this official, legally pre-defined trajectory (RIT) would have a positive impact on the RTW of employees on sick leave. According to our hypothesis, such an official trajectory would lower thresholds for all parties (employees, employers, insurance organisations) to start this trajectory toward the RTW. According to our sample of 1416 employees who followed this trajectory, approximately three in 10 employees (30.8%) were back at work after this intervention. We expected a higher success rate of the RIT, especially as compared with the existing informal, non-pre-defined ways of RTW in Belgium, which have a high percentage of successful RTW (73–83%) [30, 31]. However, the RIT result is still positive as compared with no intervention. The chances of a successful RTW are < 20% after 1 year of SL [11]. However, we found that many employees still could not be covered by this legal initiative. In 2017, only 16,000 people on sick leave followed an official, legally pre-defined trajectory in Belgium [32], which is not even 5% of the total population of employees on long-term sick leave. In other words, there is still an opportunity to reach a large group of employees, which leaves room for improvement.
Nonetheless, the difference between the success of informal and official, legally pre-defined trajectories could be explained by the characteristics of the people who follow these trajectories. Previous research in the Belgian context showed that employees who follow more informal routes of RTW have more favourable characteristics in terms of their chances of RTW, such as shorter periods of SL or a better socio-economic background [31]. Moreover, employees with low socio-economic levels are susceptible to ill health and inactivity in the labour force [33]. Furthermore, our research in the Belgian context has shown that especially vulnerable employees who have been absent for longer or for whom the contact with the employer during their SL is difficult or non-existing, follow an official, legally pre-defined trajectory [31]. A possible explanation for the low odds of an RTW following the RIT could be that the employees who are oriented toward an official RIT instead of an informal trajectory have unfavourable characteristics for a successful RTW.
Moreover, after a RIT, many of our employees returned to work with another employer (20.1% of all employees, or 65% of employees with a successful RTW). With informal trajectories, employees often RTW in the same job with the same employer [30, 31]. A difficult relationship with the employer may explain why employees apply for an official trajectory to be able to start working elsewhere. This suggestion was confirmed in qualitative research, in which OPs stated that many (56%) employees started an official trajectory to leave their employer via medical force majeure [31, 34]. This outcome is not entirely unexpected: in Belgium, such termination of employment offers a number of advantages for both the employee (no waiting period to receive unemployment benefits) and employer (exemption from severance pay). Although the initial goal of these official trajectories is not to provide employees an easy way to change employers, a successful RTW with another employer may still be a positive outcome on a societal and individual level.
The odds of the RTW with the same or another employer depended on different characteristics (Tables 1-2). However, one of the most important predictors for the RTW with both the same and another employer was SL duration. Our findings confirm other international research showing decreased odds of the RTW with longer SL [11, 36]. To guarantee an early start of the RTW process, promoting part-time sick leave or progressive RTW is recommended. In this case, an employee can rely on a part-time salary paid by the employer and a part-time allocation paid by the sick leave fund. Part-time RTW can have a positive effect on recovery and can promote full-time RTW and future work participation [37, 38].
Older age was a hindering factor for RTW with another employer. These results confirm the difficulties that older employees can encounter when looking for a job. Older employees might experience negative stereotypes regarding their attitudes and productivity [39]. Regardless, age did not have an impact on the RTW with the same employer, so we found no proof of age discrimination. A possible explanation could be that the current employer already knows the value of their older employee and is willing to support the RTW. However, unexpectedly, sex did not have a significant impact on RTW. Although not a generalization, SL may be more frequent in women than men but most often restricted to short-term SL [40]. More than half our employees were in long-term SL (>6 months), which might explain why sex did not affect our outcome.
We have no previous data on a successful RTW related to the size and sector of the organizations employees were working for in Belgium. Because of the objective data retrieved from CBSS, we were able to investigate these factors. Odds of the RTW with their current employer were reduced for employees of smaller organizations (<200 employees). Employers of small- and medium-size enterprises may have less capacity or options to offer other or modified work to their disabled workers. Our observation confirms previous international research [41, 42]. However, national Belgian research revealed that a larger organization does not automatically increase the chances of a successful RTW in the context [31]. According to OPs, medical advisors, and general practitioners, the presence of an RTW policy rather than the size of an organization has the largest impact on RTW [31]. In addition, longitudinal research in The Netherlands revealed that employees had better psychological health and less SA when there was an organizational policy with a strong emphasis on health [43]. Therefore, an important suggestion is to encourage employers, especially those of small- and medium-size enterprises, to develop a health and SL policy. Occupational healthcare professionals, such as those working at SPPWs, could help shape these policies and have a positive impact on dealing with SA on an organizational and individual level [44].
The organizational sector had no significant predictive impact on the RTW. However, we had to cluster different types of sectors because of our sample size, which might have caused the lack of significance.
Implications for policy and practice
A one-size-fits-all approach to RTW might not be the best because socio-demographic–, work-, and SL-related characteristics play a role in the RTW process. This finding agrees with the systematic review of Venning and colleagues (2021), who suggested that a stepped approach to RTW might be the most successful [22]. Low- to moderate-intensity interventions are preferred before investing in high-intensity interventions that might only be useful for specific groups [22].
Moreover, SL duration had a great impact on the odds of the RTW, which confirms previous research [11]. The sooner the RTW process is initiated, the greater the chance of a successful RTW. Therefore, measures to promote an early start of the re-integration process are needed.
Hence, we propose a stepped approach for RTW: in the beginning of the SA, employers must keep in touch with their employees on sick leave (step 1) [45]. As soon as the RTW is (partially) possible, an informal RTW process can be initiated, ideally with the support of the OP (step 2). These informal, low- to moderate-intensity interventions have the greatest chance of the RTW with the same employer [30, 31]. However, when personal contact between the employer and employee is difficult or not possible, an official, legal trajectory could be used to initiate the RTW process (step 3). Although more employees RTW with another employer after an RIT, this can still be seen as a successful RTW at an individual and societal level.
Finally, employers, especially those of small- and medium-size enterprises, could benefit from support to promote the RTW. Employers should receive support to hire older employees and to provide other or modified work for employees who might need these modifications to stay at work. Both these actions, as for those promoting an early start of the RTW process, should be formalized in an RTW or health policy within an organization. Occupational health professionals can guide organizations in shaping and executing these policies.
Strengths and limitations
The strength of this study is that we could link data from one of the largest ESPPWs in Belgium to data available at the CBSS. The advantage of the CBSS is that the data are based on objective parameters collected by different regional and federal government institutions. Therefore, data were available for all employees in our sample, which allowed for following them, even with an RTW with another employer. A disadvantage of the CBSS is that there are no data on the injury or illness that initiated the SL, which could be a predictor of the RTW as well. A limitation was the fact that our final sample was reduced due to missing data in the initial study sample.
Another strength of the study is that the method of data analysis allowed us to evaluate the impact of an official, legally pre-defined RTW trajectory on the RTW in practice. Although the available data were limited to 1 year and one ESPPW, our methodology could easily be applied on a larger scale. The pre-programmed script in R Studio can simply be copied and pasted for use with other datasets. Hence, future research should focus on expanding this study and including data from multiple ESPPWs and multiple years to validate and generalize the findings.
Conclusions
Almost one third of our employees (30.9%) in Belgium was back to work 1 year after their official RTW trajectory, most (65%) with another employer. Although this proportion of successful RTW after an official trajectory was lower than expected, this outcome could still be seen as a success. More research on this topic seems necessary to confirm our findings and determine the reasons for the high RTW with another employer. As previous research has shown, Belgian employees who follow an official RTW trajectory might have characteristics that can negatively affect their chances of the RTW, such as a less favourable socio-economic background or longer SL duration [31]. These factors can reduce their chances of success. Nevertheless, despite these possible hindering factors, a large group of employees still succeeds with the RTW after their official trajectory. Nonetheless, we should not implement a one-size-fits-all-approach. A stepped approach is preferred, with low- to moderate-intensity interventions and informal ways for the RTW to start with. When employees have difficulties in the RTW with these informal pathways, high-intensity interventions or official trajectories can be used. However, for both informal and official trajectories, an early start to the RTW process should be promoted.
Ethical approval
The study was approved by the Education-Support Committee (ESC) of Health Care Management and Policy (KU Leuven) (MP016655, 25/10/2020).
Informed consent
Not applicable.
Conflict of interest
Isabelle Boets declares that she has no conflict of interest.
Steven Luyten declares that he has no conflict of interest.
Sofie Vandenbroeck declares that she has no conflict of interest.
Lode Godderis declares that he has no conflict of interest.
Footnotes
Acknowledgments
The authors would like to acknowledge Chris Brijs as their contact person at the Crossroads Bank for Social Security, and Martijn Schouteden for his help with programming the statistical analysis in R Studio.
Funding
The cost of the data-request at CBSS was covered by group IDEWE.
Appendix 1
The re-integration trajectory (RIT) can be requested by four main actors, depending on the following pre-conditions: 1) the employee themself or 2) their treating physician, regardless of the sick leave (SL) duration, 3) the employer but at the earliest of 4 months of continuous SL, or 4) the medical advisor of the sick leave fund, if they believe the employee is eligible for the return to work (RTW).
If a RIT is started, the occupational physician (OP) must decide on the remaining work capacity of the employee, which can be divided into 5 categories (
).
When other or adapted work is possible (case A or C), the employer must make a re-integration plan, in which they decide on the modalities of the other or adapted work. Employers can also write a report to state why they cannot offer other or adapted work. In case of decision D, or when there is a report after decision C, that is, when there are no possibilities to RTW with the current employer, the employer or employee can end the contract of the employee due to medical force majeure. In that case, the employer does not have to pay severance.