A central finding was the limited extent to which the subsidies were used, especially the more extensive subsidy for coordinated interventions. Although the contents of the subsidized services were generally perceived as positive, most OH service providers reported no changes in work routines. Further, the subsidies were largely unknown by employers. OH service providers’ explanations for not using the subsidies point to a perceived lack of advantage and too high costs for applying for subsidized services, compared with regular practice. This is further illustrated by the extensive use of the only subsidy which did not require any actual service deliveries (the basic subsidy) while those where this was demanded were used less than expected. OH service providers also reported on a mismatch between the design of the subsidies and the priorities of the OH service providers, who often prioritized preventive interventions which was not within the scope of the subsidies.
Why were the subsidies for coordinated interventions not used?
It is relevant to ask whether the minimal use of the coordinated interventions (the subsidy most strictly focused on RTW) were due to the interventions themselves, to the dissemination of the subsidies, or to characteristics of the users (OH service providers and employers). In answering this, it is relevant to ask whether the potential users
understood, were able to and
wanted to implement the interventions [
25]. These questions refer to different domains identified by implementation literature.
Research on implementation has developed within different disciplines, of which the most elaborated traditions can be found in studies of evidence-based practice in healthcare settings (often labelled “implementation science”), and in studies of public policy implementation. Within the former tradition, a comprehensive conceptual framework has suggested five domains of determinants for implementation outcomes: intervention characteristics, outer setting, inner setting, characteristics of the individuals involved, and the process of implementation [
21]. These domains contain a broad variety of determinants targeting individual, organizational and societal factors, and correspond well to other studies on implementation in healthcare (cf. [
26‐
28]). In the literature on policy implementation, a range of variables has been identified for analysing implementation, and one comprehensive review presented the following seven variables: policy characteristics, policy formation, vertical public administration, responses of implementation agencies, horizontal inter-organizational relationships, responses from those affected by the policy, and environment or policy context [
22]. In many respects, the two traditions have presented similar determinants for implementation studies, although with different focus. The present study analyses policy-driven interventions in an occupational health setting. Thus, the following discussion will combine concepts used in the two fields, based on which are better fitted to explain the different parts of the implementation process (where policy implementation literature is generally more sensitive to organizational and political dimensions, for instance).
The
intervention characteristics involve the relative advantage of the intervention related to other methods, the adaptability of the intervention to local conditions, whether it is possible to try out the intervention before using it on a full scale (trialability), complexity and cost [
21]. The use of the subsidies was influenced by their design: the complexity involved in applying for and administering subsidies was repeatedly mentioned as impeding the implementation. OH service providers generally perceived coordinated interventions as too complex to administer within the given time limit of 45 days, although the contents of the actual interventions were perceived as useful. In this respect, the advantage of applying for subsidies compared with regular practice was perceived as low. The costs involved were also attributed to the heavy administration. Adaptability was perceived differently: some managed to integrate the coordinated interventions into their regular routines, while others did not. Some OH service providers mentioned that they tried using the interventions on a smaller scale, and then planned to increase the use of them, indicating that the trialability of the subsidies was good.
Policy implementation literature has discussed
policy characteristics in terms of their ambiguity (whether or not they are based on knowledge) and conflict (whether or not different stakeholders agree) [
29]. The evidence base for these subsidies may be perceived as relying on the acknowledgement of workplace-based interventions and early RTW [
2,
5]. The conflict between stakeholders could refer to different ideas regarding the responsibilities for rehabilitation, in that the employers, the OH service providers, regular healthcare and the SSIA are all involved with different roles. The subsidies aimed to promote workplace-oriented interventions by engaging OH service providers in RTW processes, which thus was a policy ambition that may have been perceived differently by different stakeholders, due to differing perceptions on the responsibility for carrying out such interventions. The material of this study indicates that both employers and OH service providers generally agree on employers’ responsibility for workplace interventions, although the role for OH services in facilitating or participating in such interventions is not equally clear.
The
outer setting involves policies, incentives and the overall political context in which the implementation takes place [
21], which roughly corresponds to the
policy context in the literature on policy implementation [
22]. The results indicate that it was mostly large employers with well-established connections to OH service providers who used the subsidies. The problems of involving smaller employers in OH service interventions are reasonably an effect of the structural conditions for OH in the Swedish system, in that it is difficult for employers with scarce resources to afford OH services [
13]. It could also be argued that the changing structure of the labour market (towards more insecure employment contracts and increased use of temporary agency workers, cf. [
30]) makes OH service interventions more difficult. Studies have also pointed to the increased vulnerability of those with temporary or precarious employment contracts [
31,
32]. It could be assumed that the interest for employers to offer OH services would decline as the use of temporary agency workers increases. Within the Swedish system, the responsibility for rehabilitation of sick-listed temporary agency workers lies with the temporary work agency, where the decision whether to use OH services in rehabilitation (as well as the cost) is placed on the temporary work agency. The extent of OH service use in such settings is yet to be studied.
In a system where employers’ use of OH services is optional and provided on a free market, it is difficult for governments to promote the use of specific services. If it is not possible for the authorities to implement OH service interventions through legislated employer responsibilities or mandatory OH service consultations, promotional activities will need to be based on various forms of incentives. Financial incentives may take different forms. In many workers’ compensation systems, employers finance the system by paying experience-rated premiums, which has been shown to increase employers’ claims and cost management activities [
33]. Other studies have shown that subsidized wages for disabled people (if sufficiently generous) may be effective for making employers more positive to employ disabled people, although risking to create a segregated form of employment for people with disabilities [
34]. The financial incentives in this study, however, were mostly targeted towards OH service providers, with the implicit intention that employers would be offered cheaper OH services. The study illustrates how such intentions were not realized, due to the problems in influencing the relationships between actors on a market in which governmental regulation is low.
Governments may introduce either permissive or enforcing regulations, where the former will be followed primarily if they are perceived as corresponding to the actors’ own goals, interests and values, and it is common that actors fail to follow rules since they simply do not know about them [
35]. Financial subsidies is an example of a set of permissive rules, where the use was determined by whether the OH service providers and the employers knew about them, and whether they were perceived as applicable and useful (cf. [
25]). As reported in the results, both the knowledge of the subsidies and the perceived usefulness can be questioned, which serves as one explanation for the low use. A more enforcing regulation could possibly have implied that the regulation would have been better known and applied more broadly. However, introducing such regulations would mean that the system for OH services would need to change fundamentally by introducing obligations for employers, which would be a more drastic political move than introducing a subsidy system.
The
inner setting involves whether an organization has a capacity for change, how work is organized, and the networks between organizations [
21]. Policy implementation here discusses
implementation agencies’ responses and
inter-organizational relationships[
22]. The OH service providers approached the subsidies differently: some adapted their organizations more than others in order to integrate the subsidized interventions into their routines. The relationships between OH service providers and employers also differed: some OH service providers seemed to inform employers and discuss the use of the subsidies, while others seemed to prefer to leave the employers out of the process. The results suggest that those OH service providers who managed to integrate the interventions into their regular routines were satisfied with the content of the interventions, although they were still critical towards the administration required.
Characteristics of individuals involves aspects such as knowledge and beliefs regarding the intervention [
21], which varied between OH service providers, and between professionals. Employers also varied in their knowledge of the subsidies: most did not know about them, while others were engaged through tight connections with OH service professionals, especially where there was an in-house OH service provider. Generally, the implementation process is facilitated if interested and committed people (described as “champions” in the literature [
21]) are involved. This could be observed in interviews with OH service representatives. Individual responses to the subsidies thus differed with regard to how they understood the policy, and how they responded to it [
22].
The
process of implementation focuses on the activities carried out to implement the intervention, such as planning, engaging stakeholders, informing users, and evaluating results [
21]. Policy implementation literature also focuses on how policies are
formed, and how this is communicated
vertically, to the street level [
22], and on the impact of professionals executing the policies, who by some are considered as being policy-makers in their own right [
36]. In this study, the results point at a rather single-handed dissemination strategy, where OH service providers were informed via the SSIA (the authority responsible for administering the subsidies) and the Swedish OH service association, while it seemed to be left to the OH service providers to inform employers. The results suggest that employers were rarely informed or engaged. Applying for subsidies and carrying out interventions was seen primarily as a concern for OH service providers, although some OH service providers had understood the subsidies as also aiming to subsidize the costs for the employer (either by splitting the subsidies between OH service providers and employers, or by giving rebates on the interventions). This reflects an ambiguity in how the subsidies were communicated from the responsible authority, where different OH service providers understood this differently. Cost reduction for employers, it seems, rarely happened.
The extensive critique of the design of the subsidies suggests that the final formulation of the subsidies did not manage to convey the needs of the users, despite the representation from the Swedish OH service association in the policy formation process.