Download as PDF

NAACOS Summary of the ACO Chapter in the June 2018 MedPAC Report 

Background and NAACOS’s reaction: The Medicare Payment Advisory Commission (MedPAC) is an independent congressional agency that advises Congress on issues affecting the Medicare program. MedPAC recommendations are typically included in one of two annual reports, released in March and June, and the report released on June 15 contains recommendations on ACOs. The full June MedPAC report is available here, and most important for ACOs is Chapter 8 , Medicare Accountable Care Organization Models: Recent Performance and Long-term Issues. 

NAACOS is pleased with MedPAC’s supportive comments about the Medicare ACO program and the acknowledgement that ACOs are contributing savings to Medicare and improving quality. We specifically appreciate MedPAC’s recognition that ACOs save more money for Medicare than what is reflected in basic evaluations of performance compared to CMS manufactured benchmarks. The Commission correctly notes a number of key ACO policy issues that need to be addressed. NAACOS calls on Congress and the Administration to fix ACO policies to ensure the long-term viability of Medicare ACOs.


Introduction and Evaluation of ACOs based on CMS Benchmarks

The June MedPAC chapter on ACOs discusses performance and long-term issues confronting policy makers and ACOs. MedPAC acknowledges the pitfalls with only looking at ACO performance relative to CMS benchmarks and that benchmarks are not the best measure of what spending would have been in the absence of the ACO and thus may not be a good measure of true program savings. MedPAC further acknowledges that ACOs may have saved Medicare from 1 to 2 percent more than indicated by their performance relative to benchmarks, and MedPAC notes that two-sided ACO models appear to save more than one-sided ACO models. In general, ACO models at two-sided risk “seem to be the models that best meet the Commission’s principles because they encourage clinicians to be responsible for the quality and cost of care for a defined population of Medicare beneficiaries.” MedPAC notes that MSSP ACOs generally perform well on quality metrics and that the MSSP ACOs on average had strong patient experience results and high-performing readmission results from 2012 to 2016. They also note that Next Generation ACOs have performed well on quality and on cost, based on the results from the first year of that program. 

ACO Evaluation According to Other Researchers

MedPAC discusses howbenchmarks and counterfactuals differ, which is a point that NAACOS has repeatedly emphasized.  Benchmarks are designed to fulfill policy goals, such as to encourage clinicians to participate in ACOs or to increase equity across the country. Therefore, “savings” relative to benchmarks will not be the best estimate of program savings relative to what would have occurred in the absence of the ACO, i.e., the counterfactual. MedPAC notes that a counterfactual may be the better estimate of whether ACOs are saving the Medicare program money, a point which NAACOS strongly supports. However, savings relative to benchmarks is how ACOs earn shared savings and determine whether they want to stay in the program; thus CMS-computed savings remain important.

MedPAC reviews a number of evaluations that use a counterfactual (i.e., what spending on the beneficiaries in the ACO would have been in the absence of the ACO) to estimate savings. The Commission explains how most studies in literature compare changes in ACO spending with changes in spending for a control group. It cites research by academics at Harvard, including Michael McWilliams, showing net MSSP savings in 2014 of $287 million or 0.7 percent of spending for ACO beneficiaries and other positive results for Pioneer ACOs. They also cite a study from L&M Policy Research showing $280 million, or 3.7 percent of spending, in savings for the first year of the Pioneer ACO Program. Another analysis by Colla et al in 2016 examined combined performance of MSSP and Pioneer ACOs in 2012 and 2013 and showed net savings of $592 million, or 1.1 percent of the benchmark in 2013. Finally, MedPAC cites CMS’s Office of the Actuary (OACT) evaluation of potentially expanding the Pioneer ACO Model that showed doing so would reduce spending. Overall MedPAC concludes that given the CMS benchmarking analysis, studies in the literature and work by OACT, it appears the ACO programs have generated positive savings estimated to be up to 2 percent, and that ACOs also have a positive effect on quality. MedPAC notes, “While these savings may appear modest, they are more than most care coordination demonstrations have achieved, including the most recent Comprehensive Primary Care Initiative.” In addition to direct savings, the literature shows indirect savings of two kinds: spillover and reduced Medicare Advantage (MA) benchmarks. The spillover effects benefit patients that receive improved care but are not assigned to the ACO. In terms of the reduced MA benchmarks, this is a result of a county’s fee for service (FFS) spending being reduced, which ultimately lowers MA benchmarks. The literature shows that a driver of ACO savings comes from decreasing post-acute care utilization. 

New Tools for ACOs

MedPAC highlights some new tools that will become available to ACOs to better manage care, such as those from the Bipartisan Budget Act of 2018 (BBA), which are detailed in this NAACOS summary. Specifically, these tools include the ACO Beneficiary Incentive Program, expanded use of telehealth, more opportunities to use prospective assignment, and enhanced voluntary alignment. 

Long-term Issue for ACOs

MedPAC discusses several issues confronting Medicare ACOs that will need to be resolved for the program to be successful in reaching its goals. Specifically, MedPAC considers the following questions: 

  • Are hospitals a viable participant in ACOs? MedPAC notes that hospitals may have conflicting interests that make it hard to be in an ACO (i.e., reduced FFS volume from fewer admissions can harm a hospital’s bottom line but help generate shared savings). However, the Commission discusses how hospital ACOs may have more resources to help ACOs assume risk and that hospitals should continue to be part of ACOs.
  • Should asymmetric models be continued? Asymmetric models include those with more favorable shared savings opportunities compared to shared loss obligations, such as Track 1+. Asymmetric models increase opportunities for participation but have the potential to cost the Medicare program more money than symmetric models. MedPAC will continue to monitor Track 1+ and evaluate whether it should be extended and whether asymmetric models should be used as a tool to attract ACOs into two-sided models.
  • How should benchmarks be initially set and rebased? MedPAC supports incorporating regional expenditures. They discuss the challenges of beneficiaries moving in and out of the ACO and how the patient churn has been more significant than anticipated. Policy makers and MedPAC need to know whether benchmarks create a useful incentive at the individual ACO level and savings at the national level. MedPAC notes that benchmarks methodologies need to continue to be refined and that while benchmarks will always incorporate policy goals, they will not – and are not intended to – represent the best counterfactual to ACO participation.
  • Should the 5 percent Advanced Alternative Payment Model (APM) bonus be distributed differently to encourage Advanced APM participation? MedPAC notes that there is uncertainty if clinicians in Advanced APM ACOs will receive bonuses because they also have to meet certain payment/patient thresholds (i.e., Qualifying APM Participant or QP thresholds). If there was more certainty around getting this bonus, it would be a stronger incentive to join a two-sided ACO. MedPAC reiterates its previous recommendation that the QP threshold be eliminated and that the Advanced APM bonus should only apply to the clinician’s revenue derived through the Advanced APM.  MedPAC notes it is unclear if this revised approach would result in more or less spending.
  • What will be the relationship between specialists and two-sided ACOs? MedPAC shares its analysis using the 2016 MSSP Pubic Use File (PUF), which shows that about 60 percent of ACO participating physicians are specialists. Specialists are included in many ACOs and there may also be a growth in specialty focused ACOs. MedPAC discusses the benefits to specialists for participating in ACOs (referrals, data, Advanced APM bonuses, help with quality reporting, etc.). MedPAC will continue to monitor the relationship between specialists and ACOs.
  • Are ACOs only a transition step to MA? MedPAC discusses how some in the industry question whether ACOs should evolve into MA plans or if there are benefits for ACOs remaining ACOs. Although MA plans have more tools to control service use, MedPAC discusses how they have higher administrative costs, approximately $1,300 per beneficiary per year (PBPY). In contrast, ACO administrative costs are roughly $200 PBPY. Which model will generate greater savings depends on whether the MA plan’s reduction in spending on medical services offsets its higher administrative cost relative to an ACO’s spending and costs. MedPAC acknowledges that one advantage of MA is the ability to keep patients in-network. MedPAC identifies that the challenge going forward is to set MA and ACO benchmarks in such a way that the models can compete and the most efficient model can gain market share in each individual market.

MedPAC’s Conclusion

ACOs show some positive results, which may be understated, and they face a number of challenges for the long-term. Some have easy solutions (e.g., fixing the Advanced APM bonus) while others require more careful consideration from policy makers. They must decide whether a preference should be given to one model (MA, FFS, ACO) over another and whether that preference should be temporary. ACOs in Medicare are popular with providers, but depending on certain policy decisions, it is unclear whether they will remain so going forward.