The Centers for Medicare and Medicaid Services (“CMS”) announced that it will use its oft maligned “latent variable” methodology to update its “Overall Hospital Quality Star Ratings” in early 2020. Healthcare providers say the star rating methodology is lacking in “accuracy and meaningfulness.” CMS also announced that it expects to publish a proposed rule in 2020 aimed at revising the quality rating methodology by 2021.
Currently, CMS assesses hospital quality with an overall score, reflecting as many as 60 quality measures in the following seven categories: effectiveness of care, efficient use of medical imaging, mortality, patient experience, readmission, safety of care, and timeliness of care. CMS publishes these Overall Hospital Quality Star Ratings on its Hospital Compare website.
Many are critical of the rating methodology which uses a statistical model referred to as the “latent variable model,” and some studies have questioned its accuracy. The latent variable model, which was created by CMS and Yale New Haven Health, gives more weight to certain measures over others. This process “inadvertently penalizes” large hospitals and academic medical centers, due to biases related to socioeconomic status, hospital size and patients requiring “heroic care.”
CMS provided its most recent ratings in February. CMS subsequently announced that it will update these ratings in 2020, notwithstanding the drawbacks of the latent variable methodology. However, updates to that methodology are in the works.
Also in February, CMS issued a request for public comment asking for feedback on how to improve the rating system with the goal of “increasing simplicity, predictability, and comparability among hospitals.” The request identified nine potential changes to the ratings methodology, including replacing the latent variable model with a non-statistical “explicit approach,” as well as “peer grouping” to adjust risk for socioeconomic status, and allowing users to customize star ratings by weighting measures according to their own preferences. CMS received over 800 comments in response to this request. Many of those comments “expressed interest in or strongly supported” discontinuing the use of the latent variable model. CMS is expected to use this feedback as a guide to draft the proposed rule. CMS further announced that it will create a Technical Expert Panel to assist with developing the new ratings methodology.
In response to these announcements, CMS Administrator Seema Verma noted that “[t]ransparency is the cornerstone of the Trump administration’s commitment to patients,” and that CMS is committed to enhancing transparency for patients, thus “empowering patients to make informed healthcare decisions, [and] leading providers to compete on the basis of cost and quality.”