PhD Candidate University of British Columbia, Canada
Background: Heart failure (HF) affects approximately 800,000 Canadians, impairing their quality of life, lowering their chances of survival, and increasing their risk of hospitalization. HF medications have the potential to substantially improve these outcomes but at the expense of increased pill burden, risk of side effects, and cost. We aimed to identify the relative preferences of patients with HF when navigating these potentially difficult trade-offs.
METHODS AND RESULTS: We conducted a discrete choice experiment in patients with HF from 7 HF clinics and 3 online platforms across Canada. Participants were presented with 12 hypothetical choice-sets, each involving a decision between 3 options: two HF medication regimens and no HF medication. These choices were described according to 6 key HF medication attributes (quality of life [QoL], survival, hospitalization, side effects, annual cost, and pill burden). We used conditional logistic regression to estimate aggregate preferences and latent class analysis to identify subgroups with distinct preferences. We translated preference coefficients to marginal rates of substitution using willingness to pay (WTP) per year for the overall cohort and relative importance for the overall cohort and latent classes.
Among the 202 participants, the mean age was 63 years, and 43% were female. Within the overall cohort, participants’ WTP per year was $244 per 1% increase in 5-year survival, $123 per 1% decrease in 5-year hospitalization risk, $74 per 1% increase in probability of QoL improvement, and $105 for a 1% reduction in the risk of bothersome side-effects.
Latent class analysis revealed 3 classes. Class 1 demonstrated a similar preference pattern to the overall group, except for cost being of lower relative importance. Class 2 was primarily concerned with cost. Class 3 was primarily concerned with cost and pill burden minimization. The main predictor of class 1 membership was higher income, whereas the inverse was true for class 2. Class 3 membership was associated with lower age and lower baseline HF medication use.
Conclusion: In the context of HF medication choices, cost and clinical benefits are typically the most important considerations. The latent class analysis revealed substantial heterogeneity in patient preferences, which highlights the importance of shared decision-making for HF medication choices.
