Pilot Resources for CMMI's bundled Payments
Considerations and Implications
Financial Considerations
For providers, the bundled payment initiative creates upside and downside risk. Contrary to the proposed financial arrangements for accountable care organizations (ACOs), there is no limit to the upside risk except for the natural cost of providing the episode. As such, if a team of providers can produce knee replacements for $20,000 on average, with a “bid price” of $24,000 per episode, the team could earn $4,000 per episode.
The downside risk can be limited by procuring re-insurance at a per-episode limit. This
is no different than re-insurance for transplants or any other episode, but does carry a
premium cost.
Since providers can select the final DRG or DRGs to focus on after they have performed
a data analysis, they may be able to limit the downside risk simply by selecting episodes
that currently have wide variation and present opportunities for cost reduction.
Organizational Issues
Non-integrated providers will have to select an Applicant to apply. To facilitate that process, we have created a simple list of options to consider. Ultimately, the core of the mechanism will rely on the agreement to split the upside and downside risks, and to select the entity that will have to provide CMS with evidence that they can pay the downside if it occurs.
Agreement on these business terms is almost a litmus test on whether or not the providers will be able to clinically integrate in order to optimally manage the patients under a bundled payment. If they can’t agree on how to split the upside and downside, then how will they ever agree on sharing responsibility for the management of a patient.
We have created a case study to illustrate how non-integrated providers can come together and team around a bundled payment.
Risk-sharing
Risk sharing under capitation models includes two types of risks: insurance risk and technical risk. Insurance risk is the risk that an episode will occur. Technical risk is the risk that technical mistakes will be made during the services provided for an episode. It is also the risk incurred in selecting the types of services included in the episode. In the Bundled Payment Pilot, only technical risk will be assumed and should be almost entirely within the control of the providers.
The opportunity for cost reductions within an episode, and therefore the opportunity for gain sharing between providers, will vary depending on the episode. For example, the current ACE demonstration has shown that there is a significant opportunity for hospitals to work with surgeons to reduce the costs of implants, which can lead to significant margin improvements per episode. Our work has shown that some episodes have high rates of potentially avoidable complications. Reducing those can also lead to significantly improved margins per episode. For example, if PCI episodes, on average, have a 30% rate of avoidable complications, reducing those by half would yield a savings per episode of 15% of current average price. After taking out the 3% discount to CMS, that would still yield a net savings of 12% per episode for the providers to share.
As such, the focus of any risk sharing arrangements should be not on the weight of the current costs per provider, but rather on the zone of compression and where that compression will come from. The % share of upside and downside risk should be based on where the zone of compression sits.
“Two-fer” Issues
Applicants can greatly benefit from working with private sector health plans to (a) create an upside “two-fer” and (b) facilitate initial discussions between non-integrated providers. In our implementation sites, participating health plans are either currently working on bundled payment or are in process. This will clearly benefit the Applicants because their evaluation will be more favorable than for those not working with private sector or Medicaid plans. Plans can host meetings between non-integrated providers to discuss terms of an episode of care bundled payment. In particular, they need to address how risk will be shared and the specific percentages of the distribution of upside dollars (or the collection of downside penalties). The agreements that can be reached between these non-integrated providers and the plan could then be ported out to apply for the CMS Bundled Payment Pilot. While the providers will still need to figure out who will be the Applicant, at least the business terms will have been settled.
Finally, the extent to which CMS has provided flexibility in the definition of the episodes, private sector plans have a unique opportunity to use definitions that would be identical to those proposed by the Applicant to CMS, this ensuring that the signal is the same, and quite strong.
There’s no question that the tasks involved in applying for the CMS Pilot will seem daunting to many, but the benefits far outweigh any potential anxiety. First, this is a relatively low-risk opportunity for hospitals to engage in much long-awaited discussions with community specialists on how to better manage patients and the input costs to procedures. Second, the discount demanded by CMS is really nothing more than an inflation rollback. It’s certainly not a deep discount. Consider, for example, the results of an analysis done on a 2.5 million Medicare fee-for-service database of beneficiary Part A and B claims. We focused on total knee replacements because of their likelihood to be picked for this pilot. The total average cost for an episode with a 30 day lookback and a 180 day look forward is $22,611. Of that, as illustrated in the figure below, a little over a quarter ($5,785) can be attributed to potentially avoidable complications – costs from related readmissions post discharge, and additional professional services incurred during the stay as a result of inpatient complications.
The costs in the figure below include the DRG, the professional services during the stay and any post-acute costs throughout the 180 day period post discharge. An episode with a time window greater than 90 days would qualify for a 2% discount as opposed to a 3% discount for a 90 day or shorter time window. That amounts to $452.22 per episode. As such, the Applicant team could provide CMS with a “bid price” of $22,158. Reducing the current PAC costs by 10% would cause the Applicant team to get a small bonus on each episode: $578 - $452 = $126. Reducing PAC costs by any additional amount creates an opportunity for a significant bonus.
And of course there is also the added potential for decreasing the current input costs of the TKR, namely the implant. Results of participants in the ACE demonstration have shown that hospitals saved hundreds of dollars per episode as a result of better control over implant sourcing.
These results can be replicated by hospitals and physicians working collaboratively across the country and, contrarily to other payment efforts, does not punish the pilot participants, but rather rewards them for doing the best job possible in caring for the patient.
