What matters in treating non‐oncological rare diseases?—Eliciting experts' preferences in Poland with PAPRIKA

Abstract

Value in drugs do not always lies simply in them curing diseases: they may also be used to prolong life, improve its quality, and relieve patients' families. How much these elements matter should weigh on reimbursement decisions. Rare diseases often lack strong clinical evidence; hence, overlooking other elements that society values may lead to suboptimal decisions. We examine the preferences for treatments of non-oncological rare diseases among 16 experts in Poland with PAPRIKA-a choice-based elicitation method. Among seven attributes, clinical effect was found to be the most important, followed by returning the patient's family to normal functioning. Whether a treatment is used in children or any alternative treatment is available were the least important attributes. The respondents have a non-linear, diminishing preference for clinical effects, and an almost linear (negative) preference for cost. We develop a new method of sensitivity analysis that accounts for the non-uniqueness of part-worth utilities and the possibility of respondents answering inattentively to some questions, and we find our results to be robust. When looking at heterogeneity, three types of respondents deviating from mean-preferences emerge, in terms of what particular factors they place emphasis on: (a) cost and effectiveness, (b) to whom technology is offered (prognosis and availability of alternative), (c) the impact on family, safety, and caring for children. The part-worth utilities identified can be used to support multiple criteria decision-making in Poland.

Multiple criteria decision making in health and medicine

“…In Jakubczyk et al (2022), the authors examine the preferences for treatments of non‐oncological rare diseases among sixteen experts in Poland with PAPRIKA—a choice‐based elicitation method. Among seven attributes, clinical effect is found to be the most important, followed by returning the patient's family to normal functioning.…”

Identify Lao farmers' goals and their ranking using best–worst scaling experiment and scale‐adjusted latent class models

“…Partial‐profile DCE is often used in that case. The adaptive PAPRIKA method (Hansen & Ombler, 2008) has been widely used (for recent examples, see Jakubczyk et al, 2021, Payini et al, 2022). One of the reasons for its popularity is that it produces results even with small samples.…”