As part of the “CY 2021 Medicare Hospital Outpatient Prospective Payment System and Ambulatory Surgical Center Payment System Final Rule” (the “Final Rule”) published on December 2, 2020, the Centers for Medicare and Medicaid Services (“CMS”) finalized policies designed to overhaul the methodology used to calculate the Overall Hospital Quality Star Rating effective 2021.
By way of background, the Overall Hospital Quality Star Rating was developed as a tool for consumers to use to make informed decisions about where they receive care. CMS designed the methodology for determining Star Ratings with the goal of including as many hospitals and as many measuring criteria as possible.
In the interest of increasing efficiency, CMS is now simplifying the methodology used to calculate Star Ratings. Certain features of the calculation, such as the annual refresh, the scope of included measures, standardization of measure scores, and the use of k-means clustering to assign a rating, will remain in place. However, whereas the Quality Star Ratings system previously categorized hospitals into seven process measure groups, the Final Rule combines three of these groups into a new “Timely and Effective Care” group. As a result, beginning in 2021, the Overall Star Ratings will be made up of five groups – Mortality, Safety of Care, Readmissions, Patient Experience, and Timely and Effective Care.
In addition to the foregoing, the Final Rule makes the following additional changes to the Quality Star Ratings system:
- Uses a simple average methodology to calculate measure group scores instead of the current statistical Latent Variable Model;
- Standardizes measure group scores (that is, make varying scores directly comparable by putting them on a common scale);
- Changes the reporting threshold to receive an Overall Star Rating by requiring a hospital to report at least three measures for three measures groups, however, one of the groups must specifically be the Mortality or Safety of Care group; and
- Applies peer grouping methodology by number of measure groups where hospitals are grouped into whether they have three or more measures in three, four, or five measure groups (three measure groups is the minimum to receive a rating).
According to CMS, the above changes will not only simplify the Star Rating methodology by consolidating 7 measure groups into 5 measure groups, they will also simplify the methodology by implementing an explicit approach to calculating measure group scores; improving predictability of the Overall Star Rating system over time by categorizing hospitals based on a simple average of measure scores with equal measure weightings that hospitals can better anticipate; and improving the comparability of the Overall Star Rating system by updating the reporting threshold, and peer grouping.
Lastly, the Final Rule extends the reach of the Overall Star Rating System by expanding the system to apply to critical access hospitals and Veterans Health Administration hospitals. Notably, CMS is not, however, finalizing its proposal to stratify readmission measures under the new methodology for dually-eligible patients because it is still developing a strategy for measuring and addressing care quality for this vulnerable population.
 As defined in the DISPLAYR Blog, k-means cluster analysis is an algorithm that, “groups similar objects into groups called clusters. The endpoint of cluster analysis is a set of clusters, where each cluster is distinct from each other cluster, and the objects within each cluster are broadly similar to each other.” See, https://www.displayr.com/what-is-k-means-cluster-analysis/. As applied here, K-means clustering is used in the Overall Star Rating methodology to create five star categories and assign hospitals to star categories in a way that ensures hospital summary scores are more similar within that star category and more different than summary scores in other star categories.