Results
In total, 25 CD projects from four different regions were analysed (two from Region A, five from Region B, three from Region C, and fifteen from Region D). The CD projects sometimes consisted of multiple activities and different types of dialogues and sometimes just a single activity. One CD project in Region C, for example, included three group interviews, four deep interviews and four targeted visits to school classes. In Region D, it was more common that a CD project involved a single activity. These differences can potentially be due to the regions’ different ways of reporting on their CDs. In Region D, the documentation generally consisted of short summaries from a single activity presented on the web site. In the other regions, it was more common that a CD project was summarised in a comprehensive report. The CD projects were conducted between the years 2013 and 2019. Most dialogues were held between the years 2015–2016 (15 out of 25 CD projects).
The results from the study are presented in accordance with the four categories in the analytical framework (see Additional file
1 for results from the document study).
Representation
Which citizens participate?
It was rather common not to target any specific group and to select members of the public within the region (in eight out of 25 projects). Those who sought participation from the public in some cases stated that they aimed to reach a cross section of the public. For example, in a CD in Region B, age and geographic spread in the region was considered (3. Region B). Some CDs focused specifically on age, both by reaching out to younger persons and their parents and by targeting older people. Other groups that were targeted were people of low socioeconomic status or immigrants. Sometimes these groups were referred to as “hard to reach groups.”
How are participants recruited?
Many methods were used to recruit participants. If the aim was to reach the public, a broader recruiting campaign was often used. When having a specified targeted group, recruitment was more directed. For example, in Region C when trying to recruit young persons, they contacted youth coordinators in the municipalities to get outreach help (9. Region C). The interviews confirm that many methods were used when recruiting participants, for example by advertising, information letters, population registers, contact persons, and social media. One public manager stated that there is no person that is hard to reach if using the rights methods, for example through libraries, student unions, open houses for seniors, or internet cafés. A couple of the interviewed, however, described that it was difficult to recruit participants to the CDs. One of them thought that it was difficult because it is not as self-evident in what way the citizen is affected compared to a patient dialogue where the participants can influence and improve their own care. It was expressed like this:
“Spontaneously, I would say that it is harder to recruit people to a citizen dialogue compared to a patient dialogue. Because in a patient dialogue you have a purpose to participate as you can improve your own future care /…/ Then you want to be part and improve care to a higher extent. As a member of the public, it might be harder to see your part in it and why you should be involved. (IP3)”
Conditions of the citizen dialogue process
Do politicians/public managers participate?
In the documentation, it was mentioned that politicians were present during twelve of the CD projects. In one case, no politicians or public managers participated, but there was a consult firm who had been assigned to conduct the dialogue (10. Region C). All interviewed public managers mentioned that politicians were usually part of the CDs and that public managers acted as a supporting function. However, the role of politicians and public managers and their relationship differed depending on project and region. For example, two of the interviewed public managers said that they mainly work to support the CDs by taking notes and writing reports, and that the politicians are the ones having the dialogue with the members of the public.
“It is actually the politicians who do the work, while the public managers do the background work of taking notes and writing reports. The politicians conduct the interviews and conversations, and are involved in analysing the results, after that the public managers write the report based on what they concluded. (IP3)”
One respondent instead stated that politicians did not participate when meeting with organisations or service providers, but often when meeting with individual members of the public. Another interviewed public manager said that the politicians are the ones initiating the dialogues but that the public managers often design and conduct the dialogue activity and inform the politicians about it afterwards.
What type of dialogue activity is conducted?
The most common type of CD activity was discussion groups or focus groups (12 out of 25). During the group discussions, different methods were used to activate the participants and to get their views. Other discussion types were workshops, meetings, speed interviews, and group interviews. Another usual type of CD was to conduct a directed visit to a strategically chosen site. Examples were markets, festivals, school classes, or civil society organisations. It was also common that CDs were carried out as interviews instead of group discussions, in which the participants did not have the chance to discuss topics with other participants. Some of the CD projects were carried out as surveys or digital surveys, but these activities were not included in the study.
In what part of the decision process is the dialogue conducted?
Most of the CDs were held when initiatives and ideas were raised early in the decision process (in 19 out of 25 projects). For example, in a CD project in Region B, the aim was to gather the citizens’ views on how self-care and health promotion could be improved (3. Region B). Three projects were conducted when inquiries are made before a decision, for example, a case in a CD project conducted in Region A where it was stated that the participants’ thoughts on healthcare needs were to be used as input for the decision on the new strategy for future healthcare services in the region (1. Region A). Three projects were conducted when a policy was evaluated. For example, in Region D when a directed visit was made to a family centre to explore what the visitors thought about the centre (13. Region D). No project was conducted in the process of drafting a decision or when a decision was made. Four CD projects were not described as being part of the decision process at all.
At what level of involvement is the dialogue conducted?
Six of the CD projects were categorised as consultation. One example is a CD in Region D where region representatives met with participants, but where the participants did not get the chance to discuss issues with other participants (23. Region D). The most common level of participation was dialogue (11 out of 25 projects). One example is a CD project in Region A where 24 discussion groups were conducted to use the citizens’ thoughts as input in decisions about the region’s future healthcare strategy (1. Region A). Only one of the CD projects was categorised as influence, a CD project in Region C, which was about future healthcare within the region (9. Region C). In this CD, the participants were involved in a longer process with multiple meetings with the purpose to present the solutions from the dialogues to the decision-makers on the healthcare committee. None of the CD projects reached the co-decision step, as none of the CDs involved the participants in making decisions or let them take responsibility for the implementation of policy. Eight of the CD projects could not be categorised due to lack of information.
What is the purpose of the dialogue?
The CD projects often had several aims. The most common purpose was to gather citizens’ opinions on a specific or general matter (in 18 out of 25 projects), such as collecting thoughts from members of the public on their healthcare needs (1. Region A). It was also rather common that the purpose was to use the views of the citizens as input in decision-making (ten out of 25 projects). For example, in a CD project in Region B, the purpose was to create a decision-basis for a future strategy for primary care services (5. Region B). In some of the CDs, the decision-makers were not only interested in the views of the citizens but also aimed to inform them about a specific matter (four out of 25 projects), such as the future challenges for healthcare (9. Region C). Some CD projects stated that the aim was to make citizens aware of the politicians’ role and their policies (three out of 25 projects). For example, in a dialogue in Region D, it was stated that one of its aims was to inform the citizens about the role and responsibilities of the healthcare committee (11. Region D). In a few projects, the aim was partly described as creating democratic legitimacy by involving citizens in the decision-process (in four out of 25 projects). Five of the CD projects could not be categorised due to lack of information.
In the interviews two main purposes were mentioned. Firstly, that the CDs aim to improve the basis for decision-making and collect views on needs, as well as suggestions for improvements to be used by politicians in improving healthcare. The second aim was described as a democratic value of creating legitimacy in the decision-making process by involving citizens and giving them an opportunity to be part and to be able to exercise influence.
Content
What subjects are discussed?
The subject in seven out of 25 CDs was healthcare services in general, as in Region A when discussing the future healthcare strategy (1. Region A). The rest of the CD projects focused on specific aspects of healthcare such as equal care, child healthcare, mental health, e-health, primary care, illness among young, and cooperation between care levels. For example, in a CD project in Region C, the focus was citizens’ views on e-health and communication with the healthcare services (10. Region C).
In ten of the CD projects, it was stated in the documentation that information about the CD or the topic was given in advance. For example, in a CD project in Region B, participants were given information about the concept of e-health before discussing their thoughts on the topic (7. Region B). In the interviews, however, it was indicated that background information was generally given about the CD project and the topic to be discussed.
What opinion do the participants express and is policy proposals discussed?
In 18 of the 25 CD projects, there were discussions about policy proposals. These proposals were often broad, as in a CD in Region C where participants suggested ways to involve younger persons in society by discussing everything from how to get a job to how to stop young people from taking drugs (9. Region C). This can be compared with a dialogue in Region B where specific suggestions to improve primary care were presented, for example ways to make it easier to book an appointment through the web page, to increase the availability of specialised doctors, and to make it easier to make complaints (5. Region B).
Outcomes
How will decision-makers proceed with the results?
In eleven of the 25 CD projects, it was stated how the results from the dialogue would be used in future decision-making. Many of the CD projects specified that the results would be used as input in political decisions. None of the CD projects however stated how the results would be used more specifically. In some CD documentation, it was stated that the results would be presented to the healthcare committee. The conductors of the dialogues sometimes gave their own advice and recommendations in the written reports about what decisions should be made based on the results from the dialogues. For example, a CD project in Region C included 13 specified recommendations to the region on how to involve younger people in society (9. Region C).
In the interviews, all public managers claimed that the results from CDs are used in decision-making. However, the routines differed between the regions. Two of the region representatives described that the results often end up in reports that are presented and discussed in the healthcare committee. One public manager highlighted that the region should improve the documentation and, in a more structured way, include experiences and conclusions from the CDs in the regions’ policy documents, needs assessments, and documents specifying priorities and goals. One of the region representatives, however, described a structured routine where the CDs are part of the decision-making process and the recommendations from the CD reports are supposed to be used when writing assignments to the providers in the rulebook for primary care, agreements, and contracts with care providers. However, this was described as difficult because CDs are more general than patient dialogues:
“The recommendations become more general, to society, for example that schools should have information about tooth brushing (IP3).”
What is the outcome of the dialogues?
As the documents only summarised the results of the CDs, they could not be used for analysing the outcome or impact of the dialogues in policymaking. In the interviews, the public managers’ perceptions of impact were somewhat ambiguous. They simultaneously stated that the CDs led to impact in the decision-making process and gave examples of such impact, and said they were unsure of the impact. One example of impact was when a CD report became part of the decision-basis for a regional development strategy. Another example was when a CD at a family centre showed that it was hard to reach some parents, which led to the healthcare committee starting an investigation around this. An example of disappointing outcomes was also given by a public manager whom described a CD project on equal care that did not lead to as many intent policy formulations as wished for, but only some general ones, which were not only the results of the CD, but were issues that the region already worked with, such as equal treatment, patient centred care, and a right to an interpreter in healthcare if you are born abroad. As mentioned, the representatives also described that they really did not know how much impact the CDs had, as it was not documented and not traceable. For example, one of them described how, after a CD, they always send the summary reports to the politicians, but that it is not obvious how the results are used in decision-making. It was expressed like this:
“The reports [CD results] are always sent to the politicians and are processed by their support functions, so it always undergoes some sort of political process, but what they do with it and how it impacts decisions is still arbitrary (IP1).”
Discussion
The aim of the study was to evaluate the functioning of citizen dialogues (CDs) in Swedish healthcare in terms of representation, process, content, and outcomes. These four aspects of public involvement activities are discussed below and related to the literature on citizen participation in systems based on representative democracy.
Regarding
representation, the results show that participants in the CD projects were recruited from the public in about every third dialogue (self-selection). It was also common with CDs directed towards young or old individuals, but also immigrants or unemployed (selective recruitment). Thus, there was a mix between involving citizens more broadly, which implies a higher risk of skewness (e.g., towards a higher level of education [
14]), and involving a more targeted group of citizens. As pointed out by Slutsky et al. [
4], there is often an ambiguity as to whom public participants represent and if they participate in the role of patient, citizen, or consumer. In some CDs, the targeted groups were selected because politiciansy are likely to consume a specific type of care and in other ones because they are part of a group that is “hard to reach” and have a weaker voice in the democratic dialogue. Thus, some of the CDs can be seen as an attempt to counteract political inequality by inviting groups that are less represented among elected politicians [
14], while the dialogues targeting the general public might be a way to increase legitimacy by reducing the distance between policy-makers and citizens and making policies more grounded [
14]. Also, however, the “hard to reach” participants were partly seen as consumers sharing information about their care needs, implying that these CDs also aim to improve the quality of particular services, i.e., enhance the effectiveness by more closely matching the values, needs, and preferences of certain citizen groups [
14].
Regarding the
process, the results show that politicians often were involved in the CDs (supported by public managers taking notes and producing reports), which signals that they are invested in the process, which is a precondition for the CD to be successful, as they are the ones making the final policy decisions [
35]. However, earlier research has also shown that public managers’ attitudes towards citizen participation is important for the outcome, not least because they constitute the link to the politicians and are a primary source of information [
29]. Based on the interviews with the public managers, they all found CDs as a valuable tool for public participation in the decision-making process. Similar to studies on CDs in the Swedish municipalities, where politicians stated that they found CDs more important in the beginning of the decision-making process compared to the end of it [
50], the CDs in Swedish healthcare were most often held in the beginning of the decision-making process, and the purpose of the CD projects were generally to use the results from the dialogues as input in decision-making and to gain knowledge (i.e., consultation and dialogue). This gives the participants an opportunity to affect policy early on, which is usually described as a success factor of PPI [
29,
51]. However, too early involvement—sometimes even unrelated to a specific policy process—could make citizen input too vague or broad, in particular if members of the public are asked for their opinions on a broad topic, such as how to improve public health. In fact, the
content of most CDs was healthcare in general, although some focused on specific issues such as primary care, mental health, e-health, and communication. Too broad of topics, without scenarios to discuss or prioritise between (i.e., preference expression rather than preference development) [
14], risk leading to watered down or unspecific suggestions or policy input that is of no real use for the decision-makers. One such example from our study was the recommendation that schools provide information about tooth brushing. This may emphasise drawbacks of participation, such as creating more bureaucracy and slowing down decision-making [
27]. One strategy to reach more specific policy suggestions is to develop preferences among the participants. This can be carried out by providing background information and educational material and then discussing merits and trade-offs between different solutions [
14]. In a majority of the CD projects, it was not stated if any background information was given and, even if the public managers stated that they informed the participants to some degree about the project, it does not seem to be used to develop preferences.
Results regarding the
outcome show that many CD projects did not specify
howthe results from CDs would be used in decision-making. Even if the interviewed region representatives were sure that results from CDs were used in decision-making, and could give some examples of impact, they described a problem in tracing the outcomes from dialogues. The regions had different levels of formalisation for inclusion of CDs within the decision-process but most expressed a need to become better at documentation and follow-up on the outcomes. Missing evidence of impact and poor documentation of effects is in accordance with previous studies of public involvement [
9,
10]. Lack of evidence of impact can be seen as a problem, but Conklin et al. (2015) state that too much focus on outcomes risks missing out on the normative argument for public involvement as something good in itself as part of the democratic process (improving justice and legitimacy [
10,
14]). In line with others, they therefore argue that public involvement should not be solely evaluated based on the impact of the activity, but also on the quality of the procedure. Similarly, Lowndes et al. [
27] argue that what is most important to know when evaluating participation is whether the policy-makers have given the results from participatory activities due weight, which we cannot answer in this study.
Further, Thurston et al. [
52] argue that evaluating public involvement by only looking at the impact on decision-making does not capture the effect that a public involvement activity might have. This argument is based on the view that the policymaking process is not linear or rational and is influenced by multiple factors. By viewing policymaking as a political sphere, where public involvement can affect the political space in different contexts, the effect of public involvement becomes more complex. Public involvement can affect the problem formulation as well as identification of solutions, which can lead to political change. Even if no decision is made based on a CD or similar, public involvement can still influence the problem and policy agenda. A counter argument to this view is that impact and documentation of impact is an important part of democratic procedure. Tahvilzadeh states that if there are no reports of impact or input from CD projects on decision-making, they risk becoming trivial [
16]. Then the process risks leading to less democratic legitimacy, as the participants feel disappointed and excluded instead of included. This is an apparent concern with the CDs in the Swedish regions, and triviality (i.e., limited scope and powers of participatory activities) has been identified as one of three challenges to creating successful participatory governance; the others being absence of systematic leadership and the lack of popular or elite consensus on the place of direct participation [
18]. One indication of the latter is that it was perceived as more problematic to recruit participants to CDs compared to patient dialogues, likely because the focus of the CDs is broader and not necessarily on services used by the participants or located in their nearby area. A possible solution could be to change the level of dialogue to the local or community level, for example by recruiting participants from a community when establishing a new health care centre. This might however be problematic in Sweden as healthcare is politically governed at the regional level and not locally in the municipalities.
Limitations
There are some limitations to this study. First, we had to rely on the regions’ own CD documentation (available at their websites or after communication with public managers), which varied in form and detailedness, both within and between the regions. This means that some information was lacking and that not all aspects of the CDs could be examined. It is possible that some CD projects were overlooked due to lack of documentation or guidance on where to find the documentation. Furthermore, another issue relates to the discrepancy in the number of retrieved CD projects from the regions, ranging from two to 15 CD projects. To address this discrepancy, we had to consider the dominance of the region with 15 projects when presenting the results.
Second, only one interview was conducted within each region. However, the interviews were a complement to the main data source (the CD documentation material), and in each region we interviewed the public manager with the most insight into the CD process. As the interviews were a complement, they gave additional information about the region’s strategy, organisation, and outcomes of the CDs, and helped validate the findings from the document study. Another limitation is that we chose to only interview public managers and no politicians or public participants, which are likely to have other perspectives [
48]. We, however, argue that the public managers who support the CD projects and document them are the most suitable to give information about the aspects we were interested in. Their crucial position has been described by Lund et al., [
28] who argue that all input from CDs must be translated into a written document, and this translation gives public managers significant influence over what input reaches the policy process. Their central position could, however, entail a will to convey a picture of CDs in their own region as successful. This was, however, not our impression from the interviews. A potential final limitation is that the collected documents were retrieved from CDs conducted between the years of 2013–2019, while the interviews were conducted in 2021. This time difference may pose a challenge in accurately recalling the details of CDs conducted several years prior. However, our interpretation is that the respondents provided credible accounts of how the CDs were conducted, which aligned with the documentation from that period.
Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.