Solidarity and cost management: Swiss citizens' reasons for priorities regarding health insurance coverage

Abstract Context Approaches to priority‐setting for scarce resources have shifted to public deliberation as trade‐offs become more difficult. We report results of a qualitative analysis of public deliberation in Switzerland, a country with high health‐care costs, an individual health insurance mandate and a strong tradition of direct democracy with frequent votes related to health care. Methods We adapted the Choosing Healthplans All Together (CHAT) tool, an exercise developed to transform complex health‐care allocation decisions into easily understandable choices, for use in Switzerland. We conducted focus groups in twelve Swiss cities, recruiting from a range of socio‐economic backgrounds in the three language regions. Findings Participants developed strategic arguments based on the importance of basic coverage for all, and of cost‐benefit evaluation. They also expressed arguments relying on a principle of solidarity, in particular the importance of protection for vulnerable groups, and on the importance of medical care. They struggled with the place of personal responsibility in coverage decisions. In commenting on the exercise, participants found the degree of consensus despite differing opinions surprising and valuable. Conclusion The Swiss population is particularly attentive to the costs of health care and means of reducing these costs. Swiss citizens are capable of making trade‐offs and setting priorities for complex health issues.


| 859
SCHINDLER Et aL. rather than addressing the trade-offs required by priority-setting. 3 Empirical research exploring the views of the public and their ability to deliberate, reach consensus, and provide utilizable direction on priority-setting in health care, can offer valuable information about the feasibility of such approaches. Previous studies of public attitudes in the United States have shown that deliberation on priorities in health care leads citizens to increase priority to the uninsured, 5 prioritize socio-economic determinants of health 6 and have led to increased participant willingness to abide by group decisions even when they had made different choices themselves. 7 We conducted a study of public deliberation for health-care priority-setting in Switzerland, a country with high health-care costs, universal coverage through an individual health insurance mandate and a strong tradition of direct democracy with frequent votes related to health care. 8 The Swiss health-care system is not a National Health

| Adapting the deliberation tool
We adapted the Choosing Healthplans All Together (CHAT)* exercise that has been previously developed for similar exercises in the United States, New Zealand and India. [9][10][11] This is a simulation exercise based on a serious game enabling participants to put together a health insurance package and thus express priorities and trade-offs. Initial development of this exercise has been described elsewhere. 12 Adapting the CHAT tool for use in Switzerland involved four steps: (i) identification of the most relevant questions, (ii) modifying the exercise materials for Switzerland, (iii) developing scenarios fitting the Swiss health-care system and (iv) translation into local languages. To identify relevant questions, we held preparatory discussions with Swiss "key informants" expert people on health financial implications: physicians, politicians and patient representatives involved in issues regarding the healthcare system. Based on these discussions, we included options designed to assess attitudes regarding aspects of health-care financing such as the level of co-pay or premium subsidies. To design trade-offs based on realistic scenarios, we worked with Milliman, an international actuarial company experienced with adaptations of the CHAT project for different US states, to create insurance benefit options that would be compatible with the Swiss health-care system and relevant there.
We then created scenarios-or health events-to help participants think about and appreciate the practical consequences of their benefit choices. Finally, we translated the material into German, French and Italian. Translations were back-translated and checked by individuals familiar with these languages.

| The Swiss health-care system
Switzerland's health system reflects the federal structure of the country. It is based on an individual mandate for insurance covering a federally defined basic package. Basic insurance is provided by dozens of private health insurance funds, with the 10 largest insurers covering over 80% of policyholders. Coverage for services included in insurance packages is 90% of the cost above the deductible in Swiss francs (300-2500 CHF), with co-pays capped at 700 CHF per year for adults, and 350 CHF for children. Provision of care is organized by the 26 cantons. The confederation thus guarantees a health system where everyone must be affiliated and covered to the level of basic health insurance. Premiums vary with the canton of residence, but cannot be risk-adjusted in other ways.

| Participants
Participants were recruited throughout Switzerland through a market research agency (Yxplora, Zürich). To illustrate the diversity of organizational and cultural aspects within our health-care system, volunteering participants were selected based on five stratification criteria: rural or urban, gender, age, socio-economic level and language. We were not able to selectively recruit healthy and sick persons given the private nature of that information. To recruit as broadly as possible geographically, we conducted focus groups in four French-speaking cities (Geneva, Lausanne, Bienne and Sion), six German-speaking ones (Bern, Basel, Zurich, St-Gallen, Chur and Luzern) and two Italianspeaking ones (Lugano and Bellinzona), recruiting each time to include participants from both rural and urban areas.

| Deliberative exercise
The CHAT exercise is designed to allow participants to prioritize the type and level of health insurance benefits they prefer. It brings together small groups of people for approximately 3 hours and confronts people through a simulation exercise with the problem of prioritizing benefits to be covered by basic health insurance. Participants are expected to consider the choices they would make for their own sake and to deliberate about the best coverage for the entire population.
To do this, participants use a pie-shaped board on which the various benefit options (See Figure 1) are arrayed to make their choices. The board is shown in Figure 1 and the benefit options outlined in Box 1.
We deliberately designed the board so that no single level would represent the Swiss status quo in every area. Participants are given 50 stickers representing units of currency to use in the selection of their benefit packages. Each sticker represents 1/50th of the average annual cost of health coverage for one person. Participants were guided in all rounds to first choose benefits at the basic level before selecting *Choosing Healthplans All Together: www.chat-health.org higher coverage levels. A CHAT manual written in simple language and describing the benefits for each coverage level, and the number of markers required to cover them, was also given to participants.
Each group participated in four rounds of decision making.
The first round required each person to choose a health plan that matched their own needs. In the second round, participants deliberated in groups of three or four to decide a collective health benefit plan for their canton. In the third round, all participants had to establish group consensus and choose an insurance plan together.
In this third round, the moderator facilitated a group discussion by asking every person in turn to select a category of coverage (dental,

| Data preparation and analysis
In this study, we report the results of qualitative analysis performed to understand the deliberative process and shed light on reasons put forward by participants to extend or deny coverage for different domains of health care in Switzerland. Group discussions were recorded, transcribed verbatim and translated into English (American Language Services, California, www.alsglobal.net) so that the research team could have a common language for analysing data.
Coding was based on content analysis, an approach considered appropriate "when existing theory or research literature on a phenomenon is limited" as was the case here. 13 The first step involved examining the data, breaking them down and making comparisons and conceptualizations, which were then labelled with a set of codes.
We then grouped codes into categories and distinguished groups of arguments used against or in favour of coverage for different domains of health care. For example, the following quotation: "Appendicitis is episodic care, a poisoning is episodic care, you can't wait for a month to do (sic) your appendicitis, waiting times can become long, and you need episodic care," was initially coded as Cannot wait for an appendicitis simply to highlight the theme. Later, it became clear that a group of similar codes could be integrated into the category Benefit for early intervention and later in the more overarching theme of Importance of medical care. As categories emerged, we established a structured code list.

Box 1 Domain descriptions
Optional categories 1. Severe injury or illness care: Care for sudden, bad injury or illness. Examples-sudden liver failure from food poisoning; massive injuries from an accident; a very premature and sick newborn.
2. Complicated Chronic Illness: Care of serious long illnesses like diabetes, heart failure, rheumatoid arthritis. These illnesses are complex and need lots of medical care to keep patients functioning as much as possible.
3. Dental: For care by dentists to prevent and treat dental problems. (Surgery of the jaw after injury, for example, is not here but under severe injury).
4. Vision: Testing and correcting for problems with eyesight that can be corrected with glasses or contact lens. Does not include other eye care. Laser treatment of the retina for diabetics would be covered by complex chronic illness.
5. End-of-life care: For patients with a terminal illness who are likely to die in a few months.
6. Episodic care: Treatment such as office visits, tests and drugs for short-term problems, such as a sore knee, constipation, cough, heart burn or skin rash, but also short-term urgent problems like appendicitis.
7. Chronic illness care: Routine checkups and care of chronic conditions that are new and not complicated.
8. Sexual and reproductive care: for care of birth control, pregnancy, sexual function and fertility.
9. Mental and behavioural care: For detecting and treating mental illness. May also cover behavioural health problems such as drug and alcohol abuse.
10. Quality of Life: For problems that are not badly disabling but affect quality of life, such as injuries affecting athletic performance. These problems affect a person's ability to act, look or feel well.
11. Prevention: To help prevent many diseases or illnesses. To identify medical problems as early as possible. There are no co-pays for preventive services.
12. Rehabilitation: To restore or improve ability to do daily activities. This includes walking, speaking, bathing, eating and critical work functions. Often needed if a person has a stroke, a joint replaced or a limb removed.
13. Long-term care: To pay for the care of a person who can no longer function independently that is provided at home or an institutional setting where a part of the quote is left out, this is indicated by an ellipsis.
For each quote, we indicate the focus group city, as well as the health domain and level of coverage discussed at that time.
Participant characteristics are shown in Table 1. Despite being covered in a universal coverage health system, 20% of participants reported forgoing medical care for reasons of cost in the past twelve months.

| Reasons for and against coverage
Views put forward by participants in deciding which services to cover include strategic arguments and arguments for financial pro- Analysis of the arguments, either in favour or against coverage, revealed the following overarching categories for inclusion of benefits (Table 2): strategy for coverage, financial arguments, identified groups, importance of medical care and responsibility argument. Additional insurance argument (21) Complete financial coverage (14) Alternative insurance argument (12) Another benefit is more important (9) Point argument (8) Lower level is justifiable (3) Redundancy (2) Financial reasons (170) Cost-benefit argument (70) Importance of financial protection (32) Adverse effect of health costs (18) Protection against individual costs (17) Prevention for financial reason (13) Incomplete financial coverage (12) Argument about cost (8) Protection of identified groups (91) Importance of protection for identified groups (64) Attention to the elderly (13) Concern for family members (10) Difference between two groups (4) Importance of medical care (91) Benefit of early intervention (41) Prevention for health reason (23) Concern for treatment (10) Endorsement of triage by good doctors (9) General doctor argument (7) Appreciation of patient-centred care (1) The place of responsibility (61) Responsibility for illness (extent or limitations) (20) Individual responsibility argument (15) Medical responsibility argument (14) Argument of individual choice (12) Argument by example (52) Personal experience argument (32) More severe disease deserves higher priority (15) The problem is not a health problem (5) Collective argument (46) Collective benefit argument (35) Importance of protection for everyone (11) Considering disease factors ( (14) Insurance coverage as a stop-gap (6) Medicine can foster social inclusion (2) Criticism of medicine and the health system (13) Against merchandizing medicine (6) Criticism of consumerism (4) Criticism argument (2) Against medicalization (1) Comparison argument (5) Comparison between two countries (5) TA B L E 2 (Continued) maximum amount, 1,000 francs, wouldn't be enough for anything. If people didn't go to a dentist for 10 years, and then they finally do go, then they have so many holes and so much calculus that those 1,000 francs aren't enough.

| Importance of identified groups
Participants argued on behalf of several groups of individuals as vulnerable and worthy of priority including families, seniors, children, diabetics, pregnant woman, mentally ill persons, teenager, young couples and workers. Codes in this category included importance of protection for identified groups, attention to the elderly, concern for family members and difference between two groups. Here is an example for families: Understood their own position better (13) Valued giving their own opinion, setting priorities and having influence (11) The discussion, argumentation and consensus (10) The discussion between generations (2) Going back to an individual plan at the end (2) People cared about the health system (2) People agreed on how expensive health care is (1) People were reasonable (1) The degree of consensus (1) Choices were difficult (1) Coverage is unequal (1) Everything is precious (1) Surprised you most in today's session?

Because it says that it would include even financial support for family members, and that is in my opinion
The diversity of opinions (28) It was constructive and interesting (10) The degree of consensus within the group (9) Choices were difficult (6) How little they and others understood before (6) Others were emotional or selfish (5) It was a game (4) The importance or unimportance of various domains to others (4) The examples presented by others and their importance (3) Essential cost-saving mechanisms were not discussed during the exercise (3) Becoming aware of their role (2) Difference in costs between levels (2) That coverage was given to alcoholics and addicts (2) Becoming aware of implications within the health system (2) How much certain things cost (2) Others participated and changed their minds (2) Some remained opposed to vaccination (1) That they had learned something (1) That some could not afford care (1) The lack of data from health insurance (1) People want efficiency (1) to have this kind of support included in there. The importance of having a family doctor can be an argument for forgoing coverage of direct access to specialists. It is important for episodic care but also to direct people when they need to go to a specialist and for long-term monitoring of people.

| Struggling with the place of responsibility
The "Responsibility argument" category included arguments regarding the causal role of individual behaviour and medical care. The question of individual responsibility for health problems recurred frequently as a source of controversy in discussions. Some people thought that we are not responsible for their health problems and the need to ensure coverage:

| Participant perceptions of the CHAT exercise
In their short open-ended answers for the post-exercise questionnaire, participants reported that the game made it easy to understand information (94%) and to make decisions (70%) and that they were satisfied with the group choice (91%) ( Table 3). They also reported what they found most valuable and most surprising about the exercise.
The aspects reported as most valuable by participants were that they had learned something, heard the opinions of different people and understood their own position better. They valued giving their own opinion, setting priorities in the exercise and having influence.
Participants reported being most surprised by the diversity of opinions, by how constructive and interesting the game was, and by the degree of consensus reached. General remarks included requests for the results, and one comment that this was better than voting.  We translated our material into English to have a common language for the coding phase, and this may have modified nuances and influenced results. This is unlikely to have affected the main structural components presented here, as they occurred repeatedly with different phraseology in all three study languages. Moreover, as the research team included members fluent in all the study languages, checks were conducted against the original version when clarification or confirmation was required. As is usual with qualitative methodology, any generalizations to different countries and health systems must be cautious. We used a highly structured focus group methodology and this may have influenced some of the responses.

| D ISCUSS I ON
This was necessary, on the other hand, to explore specific trade-offs as this required that participants focus on these choices rather than providing us with only more general views of how priorities ought to be set in the health system.
Our findings regarding the issues raised by participants-as well as their own perception of what was valuable during discussion-confirm that this exercise fosters deliberation and enables ordinary citizens to balance individual priorities and collective responsibilities. 11,21 Rather than limiting public input to either outlining general principles or commenting on specific interventions, 3 this process enabled the emergence of general principles for priority-setting based on a trade-off exercise by citizens. This exercise revealed several levels of reflection among the participants. They made priority-setting decisions, but also addressed more general moral issues associated with solidarity and responsibility. Our participants did not consider personal responsibility to be a straightforward matter of personal fault, justifying lesser coverage. Rather, they considered individual responsibility to be a component of solidarity: healthy behaviour was part of individuals' share in contributing to a health-care system responsive to all. From this angle, responsibility for health does not justify punishment if a non-responsible action is performed, but is rather something to internalize to maintain solidarity for all. Cost-effectiveness was integrated with solidarity as well. Efficient management of the health system was considered one of the means to ensure basic coverage for all and special protections for vulnerable groups.
The Swiss-CHAT exercise created conditions for reflection and deliberation, enabling participants to think through trade-offs more complex than the "yes/no" questions presented during public referenda. Findings supporting a similar conclusion were also reported following citizen deliberation in the United States and India. 21 Although the Swiss population is particularly accustomed to participatory democracy, this exercise has been used elsewhere with similar effectiveness in promoting deliberation and group engagement. Our results suggest that the capabilities of the general public to take part in setting priorities on complex issues should not be underestimated.
These issues are likely to become yet more pressing with developments in personalized medicine and increasing costs of biologics and cancer drugs. 22 At a time when European health-care systems are poised to face further increases in health-care costs in a context of important economic challenges, public engagement placing tradeoffs on the table may be more feasible than one might think.