The four participating regions differed in: in for example collaboration starting point and stakeholder composition and progressed differently through the roadmap steps. Descriptions of starting point characteristics, collaboration needs, and agreements on funding are provided per region in Table 1. The developed interventions are explained in more detail in Box 1. Moreover, details of the roadmap progress per region and contextual aspects can be found in Additional file 2.
After analysis of the semi-structured interviews and field notes, four themes with ten subthemes on factors influencing the roadmap progress were identified in the data.
Motives and interests
Several motives and interests for participating in the experiment were mentioned by participants related to their willingness to participate and their process in following the roadmap sessions, including both personal motives and organizational interests.
Personal motives were mentioned by participants regardless of their functioning level and type of organization. A common driver to participate in the experiment reported to be: improving the quality of care and life for clients. Societal gains were also mentioned to be an important factor.
‘ All in interest of the client’ – Manager SSI.
Moreover, participants mentioned that client goals were more important than financial gains. In addition to this, participants mentioned that they were in search of proof for their gut feeling that their personal investment in and support of clients in vocational rehabilitation would lead to improved client well-being.
‘But I want to emphasize again: first I was also about quantities and the financial aspects. But now, first we look at what we want to achieve [for the client], then we look how, financially. This is also important for the story.’ – Director of a job agency.
In terms of organizational interests, reducing the number of clients receiving social benefits were mentioned by participants from municipalities and the SSI. Mental health care professionals discussed wanting to serve more clients. Financial incentives, such as decreasing costs and obtaining funding, were only mentioned by project leaders and participants from municipalities and health care insurances.
‘Still, we are a money-driven organization so we hope to achieve structural financing through this project / research.’ – Municipality manager.
Other organizational interests that were mentioned included: improving expertise, broadening organizational scope, and having an interest in short-term and long-term financial and client outcomes. Organizational interests were also related to societal gains, e.g., reducing the number of people in the benefits was understood to reduce crime. On the other hand, mostly field professionals mentioned that they had the feeling that their personal and client goals did not always meet organizational goals, for example when improving the societal participation of clients does not always lead to a decrease of the use of benefits but can improve client quality of life.
‘There are societal benefits also in it. Sometimes you can help somebody. From nothing to something, like helping in a nursing home for two hours a week. I also think that’s a profit. There are no savings on the benefit payments, so sometimes you receive internal criticism. But you want help people move forward’. – Professional SSI.
Collaboration was a key element in this experiment. Important factors that were found to impact collaboration included the conditions and values under which people chose to participate, their reasons for participating, and the structure and organization of the entities of which they were a part.
Conditions and values
Participants mentioned that knowing each other, aiming for the same goal, understanding each other’s perspectives and having trust in each other were all characteristics of a well-functioning collaboration.
‘There was already a reasonable base to talk with each other; people know each other, are aiming the same goal and have sympathy for each other’s perspectives’. – Mental health care professional.
‘Well yes, when I need somebody, I call that one. You know, so that makes…., because you know each other and understand each other perspectives, …it easier to call each other.’—Mental health care professional.
Setting up a collaboration takes both time and commitment. In two regions a prior history of a collaboration was seen as a distinct advantage. During sessions, participants showed more mutual knowledge and shorter lines of communication if their relationship pre-dated the experiment. Participants pointed out key elements of collaboration which were: functioning on the same level, structured organization and specific key persons that initiate and consolidate collaboration.. However, a lack of continuity makes collaboration more difficult, as illustrated by the quote below.
‘But yes, then [project leader x] left, there was no owner, no driving force anymore and then it becomes difficult.’ – Mental health care professional.
Reasons for participating in collaboration
Reasons for collaboration between stakeholders raised during sessions included: sharing expertise and knowledge and, improving organizational alignment and cliental support. Reasons mentioned for not wanting to collaborate were disinterested in looking beyond their own organizational interests and, having the feeling that the collaboration had no added value because it was less efficient or not useful. Participants also showed different perceptions on client goals and perspectives making collaboration more complex.
‘Not relevant to us. We only look at our municipality, not at the SSI’ – Policy maker municipality.
Other reasons that kept participants from collaborating were not wanting to work in favour of other organizations’ benefits and the feeling that they were in competition of the same client goals and outcomes, e.g. in searching for jobs and employers.
“But still, we are fishing in the same pond, for vacancies and employers” – Social security professional.
Collaboration structure and organization
Due to legislation, stakeholders from different sectors have different client goals (like curing vs. helping them obtaining employment) and, different tasks (like paying benefits or caregiving). Moreover, stakeholders differ in size and geographical demarcations between client groups that stakeholders serve were observed, making collaboration difficult. Goals, tasks, and working areas overlapped partly between stakeholders but never completely. These structures are historically grown, based on legislation, and ingrained in the way professionals work. This complicated determining a common target group, among other things.
“[Stakeholder X] is not involved [in this collaboration] because the postal area of clients they work with differs. In [city X] there’s a good overlap of the postal areas they work with so there they are involved in the collaboration.” – Regional project leader.
Financial and collaboration structures between stakeholders changed over time due to regional policy changes; for example, when the services of one of the stakeholders was not purchased by another stakeholder anymore. Also, several similar collaboration plans initiatives were initiated during and interfered with the experiment (e.g., a nationwide parallel funding initiative aiming to increase the number of supported employment interventions). For participants these changes were seen both to be a complication (making the collaboration vaguer and more insecure) and a chance (good timing for introducing new initiatives or collaborators).
‘Yes, something changed in this city. This [day care*] provider is not covered by the collaboration structure from the municipality anymore. So from that point on, [day care] from this [day care] provider could only be used for the first three months of care and then the professional need to look for another day care provider’ – Social care professional.
* adults support service, see also explanation Box 1.
Politics, finances and legislation
Politics, finances and legislation played a role on different levels through the experiment, including political trade-off and financial political interaction. Moreover legislation was hindering participants on different levels and organization of funding was found as a theme.
Political trade-off and financial political interactions
Political trade-off played a role in collaboration between stakeholders and were evident during the experiment. Participants showed some reservations regarding the financial aspects: sentiments around unfair finance and task division played a role in this and hindered trust among stakeholders, while this was the actual reason for starting the experiment. At first, distrust kept participants from investing money, especially when the application of the intervention or the profit lay with another stakeholder.
’I do all the work and mental health care gets funded. That doesn’t feel good.’ – Professional SSI.
Support from regional political leaders was mentioned by participants and project leaders as being very important in the continuation of their collaboration, especially in terms of obtaining funding. But it was difficult to determine at what point these political leaders should be involved: interests and political hassle were pointed out as important factors in this.
‘Everything is with reservation/caveat of the opinion of the Alderman for the municipality/ municipality city councilman’. – Mental health care professional.
“Something is going on within a related collaboration pilot. It has something to do with the decision making by the alderman for the municipality, it’s complicated… there is a lot of talking and ‘moulding’ going on, to plan the needed meetings. I have some concerns about what’s needed and the relevance for the continuation of this experiment’. – Regional project leader.
Moreover, participants mentioned political interests for the outcome of the experiment: political leaders need a supportive narrative to defend decision making and their political and financial responsibility. A business case and showing the proof of (cost-)effectiveness were mentioned as supportive tools for both obtaining support from decision makers, as these provide insights and substantiation.
‘Our local alderman is well informed, is interested in this project but needs a proper narrative for the city council to say: “we should continue this”.’ – Policy maker municipality.
Some participants even expressed that business case outcomes are essential for convincing decision makers like managers, directors, and political leaders. On the other hand it was also mentioned that if a political leader wants to achieve something money is no issue, especially when political leaders show an affinity with the target group.
‘How we get money for this is still a question… this is difficult. But yes, when our alderman wants something, either left or right, then it will be achieved.’ – Policy maker municipality.
Participants felt dependent on both local and national legislation and existing funding systems. This is institutionalized in (organizational) system thinking and IT features. In practical terms, professionals were not allowed to access certain administrative or IT systems. This made data needed for the business case calculations inaccessible. Sharing information (like on the use of benefits and care) between stakeholders was also prohibited and participants did not feel they had capacity to take on resolving this barrier. Finally, involved professionals showed frustration at not being allowed to use certain vocational rehabilitation interventions with their clients because they had not been purchased by their organization or they do not receive a reimbursement for it. Participants expressed they wanted to resolve this. These barriers made determining a common target group and setting up a business case time consuming and complex.
‘I see a lot of obstacles within the organization, a little sneak peek: “This is an tightly organized organization. This is allowed and this is not’. – professional SSI.
Legislation changed during the experiment, which made agreements between stakeholders prone to instability. The legislative changes sometimes counteracted existing collaborative goals or resulted in new parallel goals being drawn up (e.g., an introduced subsidy scheme was only made available for one type of vocational rehabilitation intervention (IPS)), which was not in line with some collaboration agreements stakeholders were planning to make.
Organization of funding
Participants mentioned legislation prohibited them from finding financial means to invest in advance—which was needed to make the intended shared savings agreements. For example, the legal ground of financial resources was hindering. Other financial barriers mentioned included lack of financial resources to invest and the bureaucratic process for obtaining funding.
‘Yes, you know, they want me to also pay for it but I can’t pay because I’m out of financial resources and I don’t have lawful grounds to pay’.—Health care insurance professional.
Still, existing barriers were bypassed by two of the four regions. These regions did made agreements on funding for the investment of the developed pilots (Box 1). Participants in these regions made no shared savings agreements but suited agreements to the regional situation: they did agree on fees for activities by a subsidy scheme, based on stakeholder intention to invest in the target group.
Complexity of a new experiment
The concept of ‘shared savings agreements’ was rather new for the involved professionals of both sectors and this novelty has influenced the process according two different aspects: the conceptual and the organizational aspect.
Conceptual aspect: goal and concept of the experiment
At the beginning, participants mentioned that the overall goal of the experiment was not clear to them. During sessions they discussed finances, collaboration, and client outcome goals. Moreover, participants discussed about which aspect was more important and whether these can co-exist, before reaching consensus. For participants, the concept of ‘breaking down barriers’ to integrate care among two sectors was experienced as conflicting because they had the feeling that in fact some boundaries needed to be set to concretize collaboration agreements.
‘On one side I think, let’s just sit down and decide but on the other side we need to have some boundaries.’- Mental health care professional.
The idea of investing first before receiving savings was clear for participants but the arrangement of this felt difficult, especially with regard to providing money on advance. Likewise, in the development of the business case, participants experienced complicating factors, such as deciding which costs were supposed to be relevant and the difficulty of predicting costs, outcomes and benefits for a newly developed collaborative intervention. To overcome this, the business case had to be discussed in several meetings with professionals, policy makers, and directors to reach consensus. Making scenarios, supported by business case calculations, was experienced helpful in reaching consensus.
The experiment consisted of different steps on different topics, like determining the target group and realizing funding, which made the organization of the experiment complex for participants. Moreover, it was unclear which stakeholder representative was needed during which step, so ‘missing’ or ‘wrong’ stakeholder representatives were seen during multiple sessions. This led to a repetition of comparable sessions, which took time. Subsequently, new participants were needed for the next step, so the experiment had to be explained multiple times.
Participants mentioned that they had no clear picture of the required preconditions, which hindered decisive action (i.e., regarding inclusion and exclusion criteria for the target group, the continuation of the experiment, and maintaining of the savings). The set-up of the roadmap and the facilitating support by the project leaders and researchers during the sessions by was experienced helpful in going through the roadmap, making the experiment and the needed activities more clear.
‘Where are we now in the project? I do understand the relevance. There’s a project plan but I feel like I don’t really understand how we get here. Now we are looking back at the execution of supported employment and how it’s going, I don’t really understand. First, we had the health care insurance company involved as well and also other parties. Where are those parties?’ – Mental health care manager.
The volume of the experiment was felt to be too big, participants wanted to start with a smaller pilot in terms of numbers or for a shorter amount of time. Starting the collaboration on a pilot level with a small number of clients was experienced successful as a way to implement the intervention.
’Hundred clients is ambitious. Why not start with ten, learn from it and if results are positive expand from there?’—Mental health care manager.
Finally, participants indicated that decision makers can be supported by smaller decision point, intermediate and at the end. Mostly, these decisions points were introduced between steps 2a, 2b and 3 (e.g., verifying the target group, the goal of collaboration and the business case outcomes).