Risk-based contracting in the time of coronavirus
- The focus now is on making sure we have enough capacity to provide care for those with coronavirus who require hospitalization.
- However, one issue in the aftermath of the coronavirus outbreak that must be addressed is how bundled payment models and risk-based population models will be adjusted to reflect circumstances beyond participants’ control.
- For population-based models, some of the increased hospitalizations will likely be offset by decreased elective procedures but will not offset all the increased spending related to COVID-19 utilization.
The focus now is on making sure we have enough capacity (as discussed in my March 16 blog) to provide care for those with coronavirus who require hospitalization, by:
- Implementing social distancing strategies
- Leveraging every strategy at our disposal to expand acute capacity
- Making sure we have sufficient staff to meet volumes and that those staff have the required PPE necessary for protection
Knock-on effects to address
But there are plenty of potential knock-on effects from the coronavirus that also must be addressed. For example, if and when we see a spike in testing, ED visits and inpatient admissions, what does it mean for organizations participating in:
- Risk-based population models like the Medicare Shared Savings Program or Next Gen?
- Pneumonia bundles, if a patient admitted for coronavirus is more likely to trigger an outlier payment due?
- Bundled payment models in general if the coronavirus results in readmissions or ED visits during the episode window?
For population-based models, some of the increased hospitalizations will likely be offset as a result of decreased elective procedures. But my guess is it won’t offset all the increased spending related to COVID-19 utilization. So participants will incur losses (if in two-sided models) as a result of force majeure.
On the flip side, assuming this is a one-time occurrence, we’re likely to have a spike in utilization that is built into the benchmarks and target prices for future performance periods. So organizations newly entering the program will receive an unintended windfall.
Once we get through the initial anticipated surge in utilization and the coronavirus outbreak is behind us, the target price/benchmark issue is one of many ancillary policy issues beyond paying for surge capacity that CMS and providers will need to work through. And it’s not just CMS that providers will need to work with. These conversations will likely extend to both commercial and Medicaid models.
For population based models and non-pneumonia bundles, CMS, CMMI and commercial payers should consider removing admissions and ED visits with a pneumonia diagnosis, coronavirus diagnosis or CPT/HCPCS code for the coronavirus test (87635/U0001, or U0002) associated with the claim, starting with admissions/dates of service six weeks prior to the beginning of the national emergency declaration through the end of the declaration. To ensure an apples-to-apples comparison, claims during the same time period should be removed from any years used to create the benchmark/target price.
Accurately adjusting pneumonia bundles to ensure that the performance period does not penalize participants who provided care under extenuating circumstances may be more challenging. If CMMI cannot develop a methodology accepted by participants, it may need to simply hold providers harmless for episodes that begin six weeks prior to the national emergency declaration through the end of the declaration.