Advanced Bundled Payment Model Reduced Medicare Episode Payments

The CMS Bundled Payments for Care Improvement Advanced (BPCI Advanced) Model reduced Medicare fee-for-service episode payments for both surgical and medical episodes, according to a new report for CMS prepared by The Lewin Group.

CMS launched the BCPI Advanced Model in October 2018, building upon the original BCPI initiative. The BCPI Advanced Model requires providers to assume more financial risk for clinical episodes. The model aims to test if linking payments for an episode of care can reduce Medicare spending while maintaining or improving care quality.

Under the model, participants are financially accountable for the cost and quality of healthcare services during a clinical episode, which ends 90 days after a discharge or procedure. Episode payments are then compared to the episode initiator’s risk-adjusted target price. Participants receive a reconciliation payment if the episode payment is below the target price. But if episode payments are above the target price, providers must repay Medicare a portion of the financial losses.

In the third evaluation report for the BCPI Advanced Model, CMS shared results from Model Years 1 and 2, which spanned from October 2018 to December 2019. The evaluation reflected data from 13 hospital-initiated clinical episodes and 18 physician group practice (PGP)-initiated clinical episodes.

For all clinical episodes, the BPCI Advanced Model reduced average episode payments by $743 per episode, or a 2.7 percent decline from the baseline mean, the report found.

The reduction was higher for surgical episodes, which saw a 4.5 percent reduction from the baseline mean, amounting to $1,353 per episode. The model reduced payments by 2.2 percent for medical episodes, or $564 per episode. Payment changes were similar among hospitals and PGPs.

Changes in post-acute care utilization drove the episode payment reductions, the report noted.

Hospitals and PGPs saw a decrease in institutional post-acute care use by discharging a smaller share of episodes to post-acute settings. Hospital episode initiators cut down the number of days in skilled nursing facilities (SNFs) for medical and surgical clinical episodes. Decreases in SNF days were concentrated in surgical clinical episodes for PGPs.

In addition, post-acute care utilization may have dwindled in hospitals because hospitals increased home health use, the report stated.

To determine care quality, CMS evaluated the BPCI Advanced Model’s impact on unplanned readmission rates and mortality rates in the 90-day post-discharge period. The model reduced readmissions for surgical episodes by 4.1 percent but did not change readmission rates for medical episodes. Similarly, the model did not significantly impact mortality for surgical or medical episodes.

Although the BPCI Advanced Model reduced episode payments, Medicare experienced a net loss of $65.7 million after accounting for reconciliation payments. This equals 0.4 percent of what Medicare payments would have been in Model Years 1 and 2 without the payment model. Researchers said that the loss was not statistically different from zero. However, the target prices included in the model were designed to achieve a 3 percent net savings.

The model produced net savings of $204.4 million for surgical clinical episodes, or 3.6 percent of what payments would have been without the model. The model decreased spending by 3.9 percent for hospitals and 3.5 percent for PGPs. These savings were offset by the losses experienced for medical episodes.

The model resulted in a $275 million loss for medical episodes, or 2.2 percent of what Medicare would have spent without it. The loss increased to 2.8 percent for hospitals and fell to 1.3 percent for PGPs.

Based on the results, CMS determined that target prices were likely too high for medical episodes but were more accurate for surgical episodes, the report stated.

CMS implemented changes to the model for Model Year 4, which started on January 1, 2021. The adjustments aimed to improve the model’s target pricing and prevent future significant financial losses. For Model Year 4, the agency altered target price calculations using an adjusted retrospective element, required participants to choose clinical episode groups, and addressed clinical episode overlap.

CMS said it would analyze these design changes in future evaluation reports.