Medicare Shared Savings Program (MSSP) accountable care organizations (ACOs) receive quarterly Expenditure and Utilization (EU) reports that provide a snapshot of mid-year performance. Many ACOs use these reports as a barometer for their chances of generating shared savings in the performance year. ACOs also receive a Final Reconciliation Report (FRR) after the end of the year, which is the definitive summary of performance.
Even in normal, non-pandemic times, the expenditures on the EU reports do not directly align with the annual FRR due to a number of differences in the approach used to develop each report.
This blog provides an overview of the differences in methodology between the reports, and how the COVID-19 national health emergency may further exacerbate the inaccuracy of the EU reports.
Premier recommends that ACOs not use the EU reports or CMS calculators for projecting performance without making adjustments for potential error.
As noted above, there are differences in CMS’ methodologies for calculating expenditures in the MSSP EU and the FRR upon which settlements are determined. The methodological differences for calculating expenditures may distort the 2020 EU results due to COVID-19 - which could lead providers to overstate the amount of savings they will receive or understate the losses they will owe.
The concern is related to CMS’ use of completion factors when producing the EU report compared to the time allowed for full claims runout on the FRR.
- At the end of the performance year CMS allows three months of claims runout (time for claims to be billed, adjudicated and paid) before calculating expenses on the FRR
- CMS then increases ACOs’ expenditures by 1 percent to estimate, based upon CMS’ experience for the nation as a whole, the value of claims that have been incurred but not reported (IBNR). The increase in claims is called the standard completion factor.
- To develop the quarterly EU reports in a timely manner, CMS allows approximately seven days of runout and applies a standard completion factor to the aggregate claims reported to date.
- The completion factors used can be found in the parameters tab of the EU and the table below. Different factors apply, depending on the quarter and type of attribution the ACO receives (retrospective or prospective).
- CMS develops retrospective EUs using a rolling 12 months of claims and prospective on year-to-date claims, resulting in different completion factors. The table below highlights how completion factors are higher for “younger” claims, an important concept for 2020.
The standard completion factor applied to the quarterly EU report is based on national historical experience. Based on the parameters tab of the EU, the completion factors applied in 2020 are the same that were used in 2019 – meaning they fail to account for fundamental changes in volume, visits and operations during the pandemic.
This standard aggregate completion factor may not adequately reflect the ACO’s actual performance, due to the impact of COVID-19, by way of:
- Unstable month-to-month utilization volume, resulting from varying degrees of service deferral and postponement of non-emergent services throughout 2020. Many providers experienced reduced utilization in the beginning and middle of the year, with recovery to higher utilization levels later in the year. CMS applies higher completion factors earlier in the year and lower quarterly completion factors toward the end of the year, which could skew outcomes for providers who saw an uptick in volume and delayed care at the end of 2020.
- Changes to the historical mix of services provided, including the increased use of telehealth, decreases to emergency department visits, and variation in SNF admissions compared to years prior.
- Delayed claims submissions, whether due to reassigned staff, a delay in receiving updated COVID codes, payment changes (i.e., CMS increased payments for COVID-19 admissions by 20 percent) or other factors.
- Regional variations in the impact of COVID; some regions experienced a significant number of COVID-19 cases, while other regions saw minimal cases.
- MIPS payment adjustments, which may be less for 2020 than 2019 payments due to reduced provider utilization/payments.
Beginning with the Q2 2020 EU, CMS provides expenditure and utilization information including and excluding COVID-19 episodes – but does not appear to have adjusted the completion factors to account for the impact of COVID-19, which could have a meaningful impact on performance projections.
The following fourth quarter E&U completion factor examples show that these completion factors may need to be increased, if they are going to be applied in aggregate, because the later months in the year (when the bulk of the “completing” occurs) are higher cost than the middle months. In that case, the true aggregate of the monthly completion factors will be more heavily weighted toward those later months and rise.
Example 1: Value of services being provided and the claims adjudicated are relatively equal each quarter of the year, similar to a non-COVID-19 year.
- At the end of the year, one would expect effectively 100 percent of the claims from the first and second quarter to be adjudicated (with a full six months of runout), requiring no additional completion adjustment.
- The third quarter (with more than three months of runout) would need a slight completion factor and the fourth quarter would need almost the entirety of the completion.
- Since the volume of total services is equivalent across quarters, the build-up of completion can be performed quarterly or in aggregate.
Example 2: Value of services being provided and the claims adjudicated dip mid-year and the later quarters have higher values, similar to 2020.
- As in Example 1, at year end, one would expect the first and second quarters to be adjudicated (with six months runout), requiring no additional completion; the third quarter (with three months runout) would require minimal completion factor; and fourth quarter would need almost the entirety of completion.
- Note, the by-quarter completion factors do not change from Example 1; however, since the volume of services varies quarter to quarter in Example 2, the total completion estimate is different if calculated using the historical completion factors (Example 1) or aggregate factors.
- Simply applying the historical aggregate completion factor results in the completion being understated (~1.7 percent) compared to applying factors quarterly
- Using a claims triangle for this example, the completion factor is approximately 8.9 percent, which indicates the standard EU completion factor of 7.2 percent may underestimate the expenditures for 2020. Therefore, this ACO should consider increasing the completion factor for a more accurate representation of expenses
If our understanding is correct, relying on the completed Quarterly Expenditure and Utilization (QEXPUs) figures may lead to a meaningful understatement or overstatement of an ACO’s performance year expenditures, without proper adjustments.
We do not have explicit confirmation from CMS as to how the agency will handle the completion factor for 2020, but Premier’s understanding is CMS applies completion in the QEXPUs reports on an aggregate basis. An ACO’s aggregate, uncompleted, per-beneficiary-per-year figure is calculated and simply multiplied by the claims completion factor stated in the parameters tab.
Note, this is a separate, but overlapping issue of uncertainty with the OACT results.
Given the variables at play, every ACO’s approach to solve for variation in these reports will be different. Sophisticated ACOs may be able to estimate a change in completion factors for the ACO; however, the impact on the local or national reference population (from which the annual trend factors are calculated) is likely unknowable.
Key strategies should include:
- Working with an actuary to create a claim and claim line feed-powered IBNR analysis and claims triangle. These steps will help organizations calculate an ACO-specific completion factor and create a realistic range of their 2020 performance.
- Offering a range of performance metrics to present to senior leaders and the Board. While the reports generated by CMS may be one helpful indicator of performance, consider pulling in other metrics and records to demonstrate overarching results.
Reach out to our ACO experts today to explore how we can help you achieve accurate MSSP results and drive greater progress.
Interested in more? Download our white paper on how to build successful two-sided risk models.