BPCI is testing the effects of innovative episode-based payment approaches on patient experience of care, outcomes, and cost of care for Medicare fee-for-service beneficiaries. Episode-based payment bundles Medicare payments for services related to a particular clinical condition for a period of time across multiple providers and suppliers. The goal of BPCI is to align payment incentives among providers and suppliers with the health care experience of the Medicare beneficiary who is undergoing a period of treatment for a clinical condition.
Awardees may choose among three levels of reconciliation risk, or Risk Tracks, for each episode. Awardee may opt to bear risk up to the 75th, 95th, or 99th percentile. Awardees bear 100 percent of the risk up to the risk track threshold and 20 percent of payments above the threshold for a given risk track. Risk tracks may be changed quarterly.
Risk track choice is based on three concepts. First, the size of the difference in the truncation points (i.e. the dollar amount of the 75th, 95th or 99th percentiles) determines the additional risk from switching to a higher risk track. The size of the difference in the offer price for a bundle among risk track selections indicates the reward for moving to a higher risk track. Lastly, the expected volume of bundles determines the level of risk for an episode initiator.
Figure 1 depicts the expected value for switching from a lower risk track to a higher risk track. We first calculate the expected value of the revenue loss for any outlier episodes that occur during the initiative. Expected revenue loss is equal to the value of the risk assumed (i.e., the difference in truncation points between the 75th and 95th percentiles) multiplied by the probability of an outlier event (determined by bundle volume). The price difference revenue gain is simply the difference in the offer price of the two risk tracks times the total number of episodes. The sum of these two values provides the expected net impact.
It is important to note that lower bundle volumes increase the riskiness of moving to higher risk tracks. A single outlier event in a small volume bundle will have a disproportionate impact on overall revenue. At Archway we look for statistically significant differences in expected net impact to account for this higher risk. Statistically significant differences indicate that either the volume is large enough to warrant taking on the additional risk or that the difference in expected revenue gain is large enough to make taking the risk a good bet. Figure 2 depicts the result of this analysis.
CMS has created a very complicated system for the BPCI initiative to protect CMS and episode initiators from the disproportionate impact of outlier events. By breaking down this methodology and identifying a simple expected net revenue impact, episode initiators can optimally select risk tracks for their program. While the algorithm CMS has designed is complicated, the risk track choice doesn’t have to be!