Designing a Performance-Based Risk Sharing Arrangement for Aducanumab

OBJECTIVES: A recent coverage decision by the Center for Medicare and Medicaid Services in the US approved the drug, Aducanumab, for Coverage with Evidence Development (CED). CED means CMS will pay for the drug in the context of a clinical study to produce additional evidence on Aducanumab’s effectiveness at slowing Alzheimer’s disease progression. We consider a different approach that could have been taken by CMS, a performance-based risk sharing arrangement (PBRSA), that would allow for broader coverage of the drug but with a unit pricing structure tied to the effectiveness eventually observed in the clinical study and possible additional real-world evidence.

METHODS: We adapted a model of Aducanumab’s cost-effectiveness versus standard of care and combined with estimates of incidence and costs relevant to the payer, CMS. The model estimates the 5-year budget impact and total QALYs accrued with existing standard care and Aducanumab. Using the standard of care budget impact as a baseline for spending by CMS, we designed two dynamic pricing structures for Aducanumab that would change depending on what hazard ratio for progression was ultimately observed: 1) fixing total spending on this population at the status quo spending(no coverage for Aducanumab) and 2) each additional QALY accrued would be valued at $150,000/QALY.

RESULTS: In our base model, the 5-year budget impact of treating 1.4million patients is $305.7billion, and the total QALYs expected to be earned by this group in 5 years is 3,572,324. Under Scenario 2, a hazard ratio ranging from 0.60-1 would yield value-based prices of $7,784 to -$1,046per year of treatment.

CONCLUSIONS: Our results highlight that alternate pricing arrangements like a PBRSA could increase utilization compared to CED at potentially similar budget impact for CMS. Our hypothetical pricing structure ignores legitimate complexity about initiating one of these contracts, including negotiating under uncertainty, adjudication, study design and inadequate time horizon.