- To improve its hospital star rating system, CMS should divide hospitals into peer groups based on patterns in how they report on the quality measures the agency uses to rate them, a study published in the September issue of Health Affairs suggests.
- By dividing hospitals into three peer groups based on the total number, volume and variety of quality measures reported, the study's authors show how groups with more extensive reporting requirements tended to earn lower ratings than those that reported on fewer measures.
- Based on those findings, the study's authors concluded that CMS' current star rating system — which lumps hospitals together in one group to compare them — "may mislead patients, providers and payers about where to seek high-quality care."
Provider advocacy groups have attacked CMS' Hospital Compare star rating system since it debuted in 2005, saying the agency's approach to measuring quality is overly simplistic and the presentation of data to consumers is difficult to interpret.
An overview of a handful of rankings systems published last month in the New England Journals of Medicine gave Hospital Compare a C grade, trailing U.S. News and World Report but performing better than Healthgrades and Leapfrog, which both slammed the report.
CMS collects the data from its Hospital Quality Initiative programs on a variety of metrics, including patient experience, readmission and mortality rates and patient safety.
CMS uses the data to rank more than 4,000 Medicare-certified, Veteran's Health Administration and military hospitals and publishes the rankings on its Hospital Compare website.
Providers have argued for years that CMS' approach of comparing hospitals against each other in a single pool unfairly penalizes large, urban teaching hospitals that care for sicker patients than their counterparts in suburban and rural areas.
And, in fact, the metrics CMS uses to make its comparisons are not uniform across all hospitals.
That's because hospitals must have a minimum number of patient cases for a given measure to report on it, according to the study in Health Affairs. This means hospitals with a larger volume and variety of patient cases report on more measures than hospitals with smaller volumes and less variety.
Recognizing these issues, CMS is convening a technical panel composed of industry experts to recommend improvements to the methodology it uses to calculate the ratings. The agency plans to release a proposed rule in 2020, which it hopes to finalize before issuing star ratings in 2021.
Early next year, CMS plans to update the rankings using its current methodology, despite provider opposition.
In July, hospitals offered their own proposal for how to improve one component of the star ratings system with a plan to overhaul the CMS patient experience survey. They suggested administering the questionnaire electronically and adjusting what is asked to better reflect patient values.
In the study published in Health Affairs, researchers analyzed data from the 2017 star rating data set and other sources. The study included 3,692 hospitals.
They found that hospitals in the peer group with the most extensive reporting requirements were more likely earn the lowest and highest star ratings compared with those in a middle group with moderate reporting requirements. Hospitals with the least extensive reporting requirements tended to earn higher ratings than hospitals in the middle group.