A discrete choice experiment investigating preferences for funding drugs used to treat orphan diseases: an exploratory study

Abstract

Policy debate about funding criteria for drugs used to treat rare, orphan diseases is gaining prominence. This study presents evidence from a discrete choice experiment using a convenience sample of university students to investigate individual preferences regarding public funding for drugs used to treat rare diseases and common diseases. This pilot study finds that: other things equal, the respondents do not prefer to have the government spend more for drugs used to treat rare diseases; that respondents are not willing to pay more per life year gained for a rare disease than a common disease; and that respondents weigh relevant attributes of the coverage decisions (e.g., costs, disease severity, treatment effectiveness) similarly for both rare and common diseases. The results confirm the importance of severity and treatment effectiveness in preferences for public funding. Though the first study of its kind, the results send a cautionary message regarding the special treatment of orphan drugs in coverage decision making.

Prioritizing Rare Diseases: Psychological Effects Influencing Medical Decision Making

“…Our general findings are in line with previous studies from Norway and Canada [6][7][8]28] that also found that preferences for common disease patients over rare disease patients were more commonly displayed than vice versa. Moreover, discrete choice experiments in the UK found no evidence in support for claims that governments should grant special funding status for orphan drugs [29,30]. Thus, there is now converging evidence from different countries that rarity alone is not an attribute people in general think should influence reimbursement decisions in health care.…”

Challenges in Measuring the Societal Value of Orphan Drugs: Insights from a Canadian Stated Preference Survey

“…Nevertheless, a consistent finding in this literature is that when the only differentiating feature of a drug coverage decision is the rarity of the disease, Western publics do not appear to support prioritization of treatments for rare diseases over treatments for common diseases [7][8][9][10]. However, this literature also suggests that most members of the public are (1) not familiar with and do not have preexisting preferences for the prioritization of orphan-drug funding; and (2) reluctant to engage with scenarios in which the funding of treatments for rare diseases must result in the reduction of care for those suffering from common diseases.…”

“…However, since these are inherently value-laden decisions that have distributive consequences, there is growing acceptance of the democratic deficit created by delegating these types of decisions exclusively to expert bodies [4,5] and of the consequent need to integrate societal preferences into HTA and reimbursement policies as a means of increasing the democratic legitimacy of policies in this area [6]. Researchers have devoted particular attention to determining whether citizens value the rarity of a disease, independent of other factors, when forming resource allocation preferences [7][8][9][10]. If this were the case, it would be an important justification for giving 'special' status to orphan drugs in HTA [11].…”

Non-market resource allocation and the public’s interpretation of need: an empirical investigation in the context of health care

“…Such heterogeneity implies that individual preferences cannot be all depicted by a single underlying parameter vector and that behavioral patterns not explained by observed characteristics must be captured in via the estimation approach. Latent-class models are increasingly employed in the discrete-choice experimental literature (Greene and Hensher, 2003;Mentzakis et al, 2011) because, in addition to accounting for such unobserved individual heterogeneity, they allow for partial relaxation of the assumption of independence of irrelevant alternatives (IIA) and accommodate the panel structure of data. In the latent-class model, individuals are implicitly sorted into a set of C classes, where each class represents a different data-generating process (i.e., in our setting, a distinct pattern of weighing the three need-related attributes when judging need).…”