CMS Reveals Methodology for its Hospital Compare Website

The Centers for Medicare & Medicaid Services (CMS) has revealed the methodology it uses to calculate and display star ratings for its Hospital Compare website.

The overarching goal of the Overall Hospital Star Ratings is to improve the usability and interpretability of information posted on Hospital Compare, a website designed for consumers to use in order to make healthcare decisions. Late last year, CMS announced that that it had updated its quality performance data on its Physician Compare and Hospital Compare websites.

Now, CMS has said that it developed this methodology with the input of a broad array of stakeholders to summarize results of all measures currently posted on Hospital Compare. The Overall Hospital Quality Star Ratings provide consumers with a simple overall rating generated by combining multiple dimensions of quality into a single summary score.

According to CMS’ methodology report, Yale New Haven Health Services Corporation – Center for Outcomes Research & Evaluation (CORE) developed the methodology for the Overall Hospital Quality Star Ratings summarizing the quality information conveyed by many of the measures publicly reported on Hospital Compare. This development work was conducted over two years and included substantial stakeholder input.

• The methodology to calculate the Star Ratings is comprised of a five-step process, according to CMS’ report:
• Selection and standardization of measures for inclusion in Star Ratings
• Assignment of measures to groups
• Calculation of latent variable model group scores
• Calculation of hospital summary scores as a weighted average of group scores
• Application of clustering algorithm to categorize summary scores into Star Ratings

As such, the methodology allows CMS to:

• Generate a single, aggregate representation of available hospital quality information;
• Account for the heterogeneity of the existing measures available (process, outcome, etc.);
• Account for different hospitals reporting different numbers and types of measures;
• Accommodate changes in the included measures over time (retirement of existing measures or addition of new measures); and,
• Utilize an evidence-based approach reflecting both modern statistical methods, and expert insights that previously has been applied to healthcare quality measurement.