A Look at Accountable Care Organizations and the ACO Explorer

How does ACO Quality Performance vary by savings or losses?
Accountable Care Organizations (ACOs) are groups of doctors, hospitals, and other health care providers that coordinate care for Medicare beneficiaries that are being touted as potential solutions for the inefficiency and fragmentation of the U.S. healthcare system.  The Centers for Medicare and Medicaid has several initiatives focused on ACOs, including the Medicare Shared Savings Program, which includes 430 ACOs and over 7 million Medicare beneficiaries.

The ACO MSSP has been subject to both praise and criticism over the past few years, with the CMS citing millions of dollars in shared savings and improved quality of care for Medicare beneficiaries (CMS, 2016a), and critics pointing to ACOs leaving the program due to issues with benchmarks and a lack of incorporation of regional spending factors (McClellan et al., 2015).  These problems have been particularly troublesome for high-performing ACOs, which critics argue have been penalized for being too efficient and thus are subject to less room for improvement.  While the Centers for Medicare and Medicaid (CMS) have recently proposed several changes to the ACO Medicare Shared Savings Program (MSSP), some of which are aimed at improving benchmarking and regional spending issues mentioned above (CMS, 2016b), questions remain about potential for ACOs to improve quality while reducing spending. This blog explores ACO quality performance by whether ACO had losses or gains compared to their benchmark spending.

Using data from the first two years of the ACO MSSP (CMS, 2014; CMS, 2015), we explored differences between ACO quality performance by comparing two groups of ACOs based on their savings or losses: ACOs that experienced greater than a 1% loss from their benchmark; and ACOs that experienced greater than a 1% gain from their benchmark.

Table 1: ACO Quality Performance by Loss/Gain compared to Benchmark

Loss Year 1
Gain Year 1
Significant Difference Year 1
Loss Year 2
Gain Year 2
Significant Difference Year 2
Number (%)  of ACOs
73 (36%)
92 (46%)

112 (34%)
159 (48%)

Spending per beneficiary


Benchmark per beneficiary
Patient Experience


Care Coordination


Preventive Health

Diabetes Composite

CAD Composite


Health Status
Table 1 shows few significant differences between the ACOs in the two groups, though for both years ACOs with losses had higher patient experience scores, better Diabetes scores, and better patient health status, while also having lower preventive health  and CAD (Heart-related) scores. It also important to note that ACOs with losses had significantly higher benchmark spending, while having roughly equivalent spending per beneficiary as ACOs that had experienced gains.
Similar to previous analyses (Muhlestein and Hall, 2014), the results are mixed.  ACOs that experienced losses have better quality scores in some areas than ACOs that experienced gains, but overall spending per beneficiary for the two groups is almost identical.  The results point to the complicated relationship between spending and quality, and the diversity of the ACOs participating in the MSSP (McClellan et al., 2015).  Future research should focus on how ACO spending and quality performance varies geographically, and identifying the “bright spot” ACOs that are achieving high performance while also reducing spending.

For more information about the Medicare Shared Savings Program (MSSP), see https://www.cms.gov/Medicare/Medicare-Fee-for-Service-Payment/sharedsavingsprogram/index.html?redirect=/sharedsavingsprogram

An Online Tool for Exploring Accountable Care Organizations (ACOs) Quality and Spending

In order to compare how ACO performance varies geographically, HealthLandscape developed the ACO Quality Explorer, which is a spatial visualization tool displaying data for the ACOs taking part in the MSSP.  The ACO Explorer allows users to visualize 33 quality measures organized into five domains: patient/caregiver experience, care coordination/patient safety, preventive health, at-risk population diabetes, and at-risk population heart-related measures. Each ACO is given a site score based on the percentage of quality measure benchmarks set by the Centers for Medicare and Medicaid that the ACO has met. ACOs are displayed as points on the map colored green (4 or more domains with all measures at or above thresholds), yellow (2 or 3 domains with all measures at or above thresholds), or red (less than 2 domains with all measures at or above thresholds), allowing users to quickly compare different ACOs across a geographic area.  When users click on a specific site, they can view exploding rollover windows which contain data about the quality measures included in each of the five domains. In addition, we’ve added patient characteristics, spending, and services indicators for the Year 1 ACOs, and we’ll be adding these data to the Year 2 ACOs in the near future. 
Map view showing exploding rollover windows for ACOs

Further, the ACO Explorer includes a Stats tool for performing statistical analyses to explore relationships between ACO quality metrics, patient characteristics, spending, and services, while also examining associations between ACOs and population health indicators for the county in which an ACO is located.  For example, the image below shows ACOs that exceeded their spending benchmark are located in counties with lower overall Medicare spending.

View of Stats Tool showing relationship between ACOs that exceeded spending benchmark and Medicare Spending at County level (in which ACO is located)

To visit HealthLandscape’s ACO Quality Explorer, see http://www.healthlandscape.org/ACOExplorer/map.cfm

Michael Topmiller

Health GIS Research Specialist 

Related Posts