- Payments declined for approximately three-quarters of the clinical episode combinations in the Medicare Bundled Payments for Care Improvement (BPCI) model without impacting care quality, CMS recently reported.
Of the 67 BPCI model, participant, and clinical episode combinations analyzed in the evaluation of the bundled payment model’s fifth performance year, CMS found that Medicare payments fell for 50 combinations and the reduction was statistically significant for 27 of the combinations.
“This is particularly notable because of the consistent pattern of declines across both Models, different participant types, and a variety of clinical episode types (surgical procedures and acute and chronic medical conditions),” the report stated.
Hospitals and physician group practices in the BPCI realized savings for their clinical episodes by lowering the use of institutional post-acute care services. Specifically, CMS noted that skilled nursing facility payments fell for clinical episodes in both Models 2 and 3 of the BPCI.
Clinical episodes under Model 2 of the bundled payments initiative began with a hospital admission and extended for up to 90 days.
Under this BPCI track, CMS rewarded the hospital or physician group practice through reconciliation payments for decreasing Medicare payments compared to a target price. But if episode payments were higher than the target price, then participants had to repay a portion of the financial loss to CMS.
Model 2 accounted for almost 90 percent of the 796,000 episodes initiated in the BPCI’s first 13 quarters.
During the 13 quarters, CMS observed a statistically significant decline in total standardized allowed payment amounts for the inpatient stay plus 90 days post-discharge for 12 of the 32 hospital-initiated episodes and seven of the 21 physician group practice-initiated episodes.
Among Model 2 episodes, Medicare payment savings specifically stemmed from decreases in skilled nursing facility and other institutional post-acute care payments. A lower proportion of patients who received post-acute care were discharged to an institutional post-acute care provider and the number of skilled nursing facility days dropped for 30 hospital-initiated episodes (17 statistically significant).
CMS also pointed out that home health agency payments increased for 12 hospital-initiated episodes.
However, home health agency payments increased for one-half of the physician group practice-initiated episodes.
For Model 3, Medicare payment savings also primarily came from declines in skilled nursing facility payments.
Model 3 episodes started with post-acute care initiation following a hospital admission and lasted for up to 90 days. Like Model 2, CMS reconciled payments against a target price and participants had to repay a portion of any financial loss to the federal agency.
The BPCI model was more popular than Model 2 during the evaluation period. However, episode volume was lower in Model 3, CMS reported.
But like Model 2 episodes, Model 3 episode payments dropped as participants reduced skilled nursing facility payments and increased home health agency payments for a majority of the episodes, with increases being significant for four episodes.
However, higher home health agency payments did not balance out the reduced skilled nursing facility payments, CMS reported.
CMS did not report on Model 4 results because participation was so low. In Model 4, episodes began with a hospital admission and continued for 30 days. CMS also provided prospective payments and participants kept the difference if actual episode costs were below the payment.
Despite Medicare payment reductions, CMS found that BPCI did not generate net Medicare savings.
Reconciliation payments offset the reductions in Medicare payments, resulting in a net loss of $202.1 million, or $268 per episode, for Model 2 and $85.2 million, or $921 per episode, for Model 3.
“Reconciliation payments were greater than anticipated because CMS eliminated downside risk during part of the intervention,” the report stated. The federal agency waived the downside risk component of the bundled payments model in the first five quarters for all participants over concerns about the accuracy of target price.
CMS also later removed the downside risk aspect of physician group practice episodes because of episode attribution errors.
“If downside risk had not been eliminated, Medicare would have achieved a net savings of $144.3 million under Model 2 (not statistically significant) and a net loss of $55.2 million (p<0.10) under Model 3 through the end of 2016. ><0.10) under Model 3 through the end of 2016,” CMS reported.
CMS may not have reduced produced net savings with the BPCI initiative, which ended on Sept. 30, 2018. But the federal agency remains optimistic that the new iteration of the bundled payments model will.
“The BPCI Initiative had encouraging results. CMS designed the BPCI Advanced Model taking into account evaluation results and lessons learned from other Innovation Center models, industry experience with bundled payment, and stakeholder input from healthcare providers at acute care hospitals, physician group practices, and other providers and suppliers,” the federal agency recently stated in an announcement.
The announcement also revealed that 1,299 healthcare organizations have signed up to be part of the new five-year model, which launched Oct. 1, 2018.
The BPCI Advanced Model differs from the original bundled payments model by offering additional clinical episodes, including outpatient episodes, and giving participants preliminary target prices prior to the start of each model year.
The new iteration of the BPCI also qualifies as an Advanced Alternative Payment Model under MACRA, meaning participants are exempt from the Merit-Based Incentive Payment System (MIPS) and can qualify for the greatest value-based incentive payment offered by the Quality Payment Program.