Survey questionnaire

The survey sought to characterize a sample of Canadians’ priorities and perspectives associated with decision-making for DRDs. An online survey, available in English and French, was administered in January and February 2016 (Additional file 1).

The priorities and scenarios in the survey were developed over 2 years of collaboration with researchers and decision makers involved in provincial drug funding decisions and Canadian federal recommendations, and with the Canadian Institutes for Health Research (CIHR) New Emerging Team for Rare Diseases [12]. Supporting respondents’ ability to identify their priorities and receive descriptions of how the priorities would be promoted or mitigated by the scenarios provided some decision-support that is a rough parallel with the structure of a deliberation. This approach moved from the identification of participants’ hopes and concerns to assessing how these were affected by different policy choices. The survey design by MetroQuest encouraged respondents to think about priorities in relation to DRD funding decisions, and provided estimates of how the each respondent’s selected priorities would be affected by each of the funding scenarios. This structure aimed to guide respondents through a basic understanding of the policy challenge, identify their priorities, and relate those priorities to policy decisions in a survey intended to take less than 10 min. The final version of the survey was reviewed and revised by the CIHR New Emerging Team [12].

Survey respondents were asked to rank priorities, and then use the priorities to indicate their degree of support for the funding scenarios. This method may not achieve the degree of reflection and reason-giving typical of a deliberative engagement, but it provides an opportunity for analysis to assess the priorities in relation to the scenarios [9, 13].

Respondents were able to view an estimate of how each funding scenario agreed or disagreed with their selected priorities. Providing estimates of how each scenario promoted or compromised the priority selection was a decision support that helped survey respondents to more efficiently consider the scenarios based on their selected priorities. It did not, however, preclude them from rating the funding scenarios using a 5-point Likert scale independently of the projected effect on their selected priorities. Prior to submitting the survey, respondents had the opportunity to revise any of their responses in previous screens and to submit comments in an open-ended question with regards to DRD funding.

Survey sample

A third-party organization, Asking Canadians, was responsible for recruiting the survey sample of approximately 1600 individuals from across Canada. Respondents were solicited from AskingCanadians’ panel. The company had several quality assurance checkpoints to ensure the integrity of the survey respondents. For our study, once AskingCanadians’ panel members were recruited, they were asked to confirm their desire to participate in the survey before being redirected to the MetroQuest survey. A survey was classified as complete if at least three priorities were ranked, and all four scenarios were rated. Demographic data was also collected for each respondent which included age, annual household income, geographic region, sex, and education level.

Survey data analysis

Descriptive statistics

Surveys that ranked at least three priorities and rated all scenarios were labelled as complete, but 24 completed surveys were missing the scenario ratings. This figure represents 1.47% of responses, so the impact on the analyses was minimal.

Among the four scenarios presented, over 75% of respondents agreed with the scenario that described a funding model that grants equal access to treatment to all Canadians, independent of where they live, followed by 73.0% who agreed that a DRD should be funded if deemed to be cost-effective. Moreover, half of respondents support DRD funding to improve patient care for rare diseases, while over 35% agreed that provincial funding should be allocated to programs that target reduced sickness and death in underserved populations instead of funding DRDs.

Logistic regression analyses

In our initial analyses, we evaluated the relationship between agreement with each funding scenario and respondent demographics including geographical location. See Additional file 2 for frequency of priority ranking and ratings of scenarios by province. We did not, however, find any significant associations between demographic characteristics and respondents’ agreement with any specific scenario.

The selection of Cost Containment as one of the top priorities was associated with the level of agreement on three of the four funding scenarios. Specifically, those who selected Cost Containment in their top priorities were less likely to agree with paying for all drugs for rare diseases and Canada-wide equality of drug access versus those who did not select this priority. Respondents, who ranked Cost Containment in their top priorities, were also 67% more likely to agree with the prioritization of other programs over programs for rare diseases. Those who selected Longer Life were more likely to agree with both paying for drugs for rare diseases as a matter of principle, and the necessity of Canada-wide equality for access to treatment. Selecting Effective Health Care in the top priorities was positively associated with both prioritizing other programs over programs for rare diseases, and only funding a drug treatment if justified. As well, respondents, who selected National Access as one of the top priorities, were four times more likely to agree with the scenario to provide equal access to drugs across Canada.

Content analysis on open-ended question on funding for expensive drugs

Respondents had the opportunity to submit their comments on DRD funding. Among the 1631 respondents who completed the survey, 410 responded to the open-ended question (25.14%).

Several themes that emerged from the comments submitted on funding for expensive drugs were aligned with the priorities listed in the survey, such as national access to DRDs and cost containment. Other themes captured in the comments were directly related to the patient’s well-being and welfare. They included the financial burden imposed on patients and consideration for their age and autonomy. For example, several respondents wrote that families experience a heavy financial burden when drugs are not funded, and individuals should not have to choose between paying for drugs for their health versus paying for their food. Respondents also commented on drug coverage decisions based on cost-effectiveness versus decisions driven by compassion for the patient, and they questioned how the value of life was measured. The impact of media coverage on the individual cases and ensuring the accountability of industry, who charge high costs for DRDs, were highlighted as concerns by numerous respondents. For instance, one respondent wrote that decisions should be based on real data and not on emotions or “over-the-top” media-coverage. Themes identified also focused on the balance between funding for DRDs compared with funding drugs for common diseases that affect a larger population, coverage through private insurance versus public sources, and increased frequency of bulk purchases to obtain large volume discounts. One respondent commented that a national drug program was needed, and another respondent suggested a Canada-wide purchase scheme to reduce the cost of drugs. Additional themes that surfaced from the respondents’ comments included placing more emphasis on preventive care and the availability of non-pharmaceutical therapies and generic drugs to treat patients with rare diseases. Additional file 3 presents sample comments provided by the survey respondents for each theme.

Limitations

The interpretation and application of our study may be influenced by several limitations. The survey presented a description of each priority and scenario, but it is uncertain how the respondents interpreted the descriptions and if they required additional information for clarification purposes. Moreover, they did not have much of an opportunity to consider expert and stakeholder input, and our methodology did not use time-trade-off required by decision makers when prioritizing or allocating finite resources. The survey responses suggested that a majority would support national consistency of funding decisions based on evidence of comparative cost-effectiveness, and 50% agree to fund DRDs. Although the survey responses highlighted a diversity of perspectives, it is not feasible to identify all Canadian perspectives based on the number of limited number of respondents [9, 11]. In the current sample, a greater proportion of our survey respondents had a college diploma or university degree or earned at least $100,000 compared with the Canadian population [17, 18]. Although the survey sampling targets were not based on the actual Canadian population density, targeted demographics by geographic area, as well as age and sex, were specified and were within 5% of the targets. This approach helped ensure that a variety of perspectives were represented for comparative purposes.

Directions for future research

The survey responses will help to inform the planning and implementing a deliberative public engagement workshop on decision-making processes for DRDs. The theory of deliberative democracy argues that the mini-forums created for the deliberative engagement present advice that is the best estimate of how citizens would advise knowledge users if citizens were informed, civic-minded and deliberative [19]. It is a common belief that citizens are not generally well informed, nor are they supported to consider the perspectives of others and the common good. That said, the online survey provides an estimate of advice from non-deliberating informed citizens. The responses can be considered by the deliberation participants, and by knowledge users as they decide whether and how to incorporate advice from these differently constituted forums. One objective of this study was to use demographic associations, including geographical location, with different the survey responses to guide the sample selection for the deliberative engagement. Initial analyses did not identify any statistically significant associations that can be used for this purpose. However, themes that emerged from the content analysis can help to identify information that participants might require to support deliberation, which can be provided prior to the workshop and during the deliberations. For example, several comments were associated with the promotion of preventive care and use of generic drugs as a treatment option. A large proportion of rare diseases have genetic origins, and DRDs usually have extended patent protection compared with therapies for more common indications. The current survey uses the Likert scale to gauge the respondents’ top priorities and level of agreement with the funding scenarios associated with DRD funding. It remains uncertain if the results of this survey align with the perspectives of the decision-making bodies in Canada, especially in jurisdictions with programs for DRDs. An analysis on the Canadian public’s perspectives compared with the values and principles of these programs is merited. Studies on the public’s perspectives in other countries and how they contrast with those among Canadians are also warranted. Future research can apply the trade-off technique to measure the Canadian public’s perspectives in which respondents must consider and compare all key attributes associated with each priority and funding scenario. Their responses can then be compared with those derived from this survey.