Commercial risk will be a critical catalyst of progress – it’s complicated, but is it possible? We think so.



How to improve your risk performance

“What makes an ACO successful—especially MSSP E track?” That’s the question a Director of Population Health recently asked me.

Previous ACO research suggests success factors include being physician-led, duration of beneficiary attribution, site of care shifts from IP to OP and simply taking risk. Other research concludes the size of the total medical expense budget and beneficiary demographic attributes, particularly regarding disability, play an important role in ACO success. A recent study on ACO financial performance, Promise v. Practice: The Actual Financial Performance of Accountable Care Organizations, concludes that on balance the Medicare ACO program at best was financially neutral. Based on the aggregate data it would be unwise to disagree. At the same time some ACOs appear to perform very well financially in terms of not only generating savings for the ACO but also in lower beneficiary expenses.

With the existing research in mind, we conducted our own analysis to answer the question. Based on MSSP ACO performance data, what made an ACO more likely to be successful?

Our main takeaways:

  • Take risk: Risk taking ACOs earned significantly more savings than non-risk taking ACOs. Reading between the lines here, organizations more formally committed to risk more formally committed to demand destruction and savings generation.
  • Manage visit types through site-of-care shifts: This again is where full risk arrangements may reveal even greater success. Changing utilization patterns, assuming quality remains the same or improves, is a good thing.
  • Leverage the primary care enterprise: Increasing the number of primary care encounters appears to directly translate into better management of unplanned admissions and drives better performance on several important quality measures

Here is what we did

We took a cautious approach when answering the question using the MSSP ACO public use files. Within the MSSP program, the “glide path” for various levels of risk in the Basic ACO tracks still has two tracks (A and B) with no downside risk.  Three tracks (C, D and E) have progressively increasing levels of downside risk.  There is also an Enhanced track with both a greater reward potential and more downside risk.  Given this delineation in risk-levels, we ran a cohort analysis based on the amount of risk the ACO took.  The two cohorts consisted of the non-risk bearing ACOs in tracks A and B (n=190) and risk bearing ACOs in tracks C, D, and E plus Enhanced (n=164). 

The performance data for all MSSP ACOS in CY2019 has two Public Use Files (PUF), one from July and the updated PUF from January 2020.  The July PUF contains data not published in the January 2020 PUF, particularly reporting on some quality measures (see greater detail in the results section).  We used the January 2020 PUF for the majority of the analysis. We used ANOVA to detect mean differences between the cohorts.

While our analysis is specific to Medicare ACOs in a single performance year (2019), the three indicators of ACO success we identified are consistent with the findings of the previous research cited above.  This lends some validity to the idea of consistency over time.  The findings discussed below strongly suggest these same factors play an even more important role in full risk programs, like global budgets.

We tested the two ACO cohorts to ensure their comparability and found they exhibited only minor differences in their beneficiary populations.  In terms of demographic differences, the non-risk taking ACOs had fewer Hispanic patients than risk taking ACOs (p=.028) but none of the other demographic differences concerning age, gender and race were significant.  Some differences also existed in HCC risk scores by beneficiary status.


Variable Non-Risk Taking Risk Taking p-Value

























Three big findings from our analysis
  • 1. Take risk
  • 2. Manage visit types through site-of-care shifts
  • 3. Leverage the primary care enterprise

Parting thoughts

Commitment. Our three big findings answer my colleague’s question about what makes an MSSP track E ACO successful – at least in the 2019 performance year. In our analysis, committing to downside risk and committing to provide care at lower cost settings associated to greater savings.  It would be reasonable to ask, since MSSP ACOs typically do not start by taking risk, whether it is taking risk that contributes to savings or whether it is savings in other models that contributes to deciding to take risk.  These results can’t answer the question of which came first.  At the same time the act of committing to risk-based payment is the cornerstone of the other success factors we found.

Some caution is necessary. Looking at these same success indicators through a different analysis with a multivariate lens only explains half of the variation between risk taking and non-risk taking ACOs when it comes to shared savings.  This suggests other factors are at play, whether they are ownership models, number of hospitals, HCC risk score differences, or other measurements.  And those as yet unknown chapters are in no small part why the VBP book is still in development.

Have a Question?


Ask our experts a question on any topic in health care by visiting our member portal, AskAdvisory.