Patient preferences for heart failure medications: a discrete choice experiment
R TurgeonAbstract
Background
Heart failure (HF) reduces quality of life, decreases survival, and increases the risk of hospitalization. Guideline-recommended medications improve these outcomes at the expense of pill burden, side effects, and cost.
Purpose
To assess how patients with HF weigh the trade-offs between medication-related benefits and downsides.
Methods
We administered a discrete-choice experiment to 202 patients with HF from seven HF clinics and three online platforms across Canada. Participants were shown 12 choice-sets, each offering three options: two HF medication regimens and no medication. Each option was described according to its impact on six HF medication attributes (quality of life, survival, all-cause hospitalization, side-effects, out-of-pocket cost, and pill burden). Participants were randomized to the survival attribute presented as either 5-year survival probability (reference) or average life-years gained. We used conditional logistic regression to estimate preference weights, presented as marginal rate of substitution (MRS) using hospitalization as the anchor attribute. Larger MRS values represent stronger preferences.
Results
The mean age was 63 years and 43% were female. Patients expressed stronger preferences for medication-related benefits, namely for survival (maximum MRS 22.9% for a 15% improvement in 5-year survival) and quality of life (maximum MRS 25.2% for a 60% probability of improved quality of life). Patients placed less importance on downsides (maximum MRS 11.2% for side effects; pill burden not significant), with the exception of the highest cost level ($CAD 4,500 annually; MRS 32.1%). Participants exhibited a strong aversion to the option of taking ‘no medication’ (MRS 54.9%). Presenting survival as life-years gained led to a greater preference for survival improvements (maximum MRS 39.5% vs 22.9% when presented as improvement in 5-year survival).
Conclusion
For patients with HF, medication benefits generally outweighed the downsides, except for high out-of-pocket costs. Patient preferences are influenced by the format of survival communication, with stronger preferences when the same benefit is presented as life-years gained rather than as improvement of 5-year survival.Preferences for HF medication features