Development of an Open-Source Web Application for Probabilistic Value-Based Contracting Performance Estimation

OBJECTIVES: Value-based contracting (VBC) can mitigate the financial challenges due to uncertain clinical performance of novel therapies like cell and gene therapies (CGTs). By tying reimbursement to actual patient outcomes, VBCs reduce payer risks of overpaying and so may enable greater patient access. However, estimating VBC financial performance given specific metrics, performance thresholds and performance rebate levels is beyond the capabilities of current deterministic contracting tools. We demonstrate a probabilistic VBC impact Assessment Model’s (VBCAM) ability to estimate the net reimbursement envelope (mean and variance) based upon existing evidence and each party’s beliefs regarding a therapeutic’s likely real-world performance – and so facilitate VBC contract negotiations.

METHODS: Upon consulting industry leaders in CGT and VBC to better understand barriers to VBC and using a Bayesian framework, we designed a VBCAM to calculate the financial impacts (expected rebates and their variance) given a VBC design (metric and up to three rebate levels based on metric thresholds), efficacy evidence (mean effect and variance) from clinical development, and stakeholder beliefs regarding real-world effectiveness in the treated population (mean effect and variance). We developed an open source VBCAM as a Shiny web application tailored to the needs of payers and manufacturers in CGT. Initial user testing provides a preliminary assessment of the VBCAM quality, performance and fit for purpose.

RESULTS: The VBCAM appropriately calculates the VBC financial payouts if the therapy performs according to its Randomized Control Trial (RCT) evidence or per stakeholder beliefs regarding real-world effectiveness in the treated population. Initial feedback suggests the VBCAM will help stakeholders decide if a VBC may be attractive to them and acceptable to their counterparty.

CONCLUSIONS: A priori probabilistic modelling of VBCs is both feasible and likely helpful in aiding stakeholders to design, to understand the performance of, and to negotiate VBCs.