The Centers for Medicare and Medicaid Services (CMS) is moving forward with plans to launch a new 5-star rating system on its Medicare Hospital Compare web site. The new rating system gives a single star-rating from 1 (worst) to 5 (best). According to CMS:
“The Star Rating, which was developed through a public and transparent process, takes 64 existing quality measures already reported on the Hospital Compare website and summarizes them into a unified rating of one to five stars”.
Seems reasonable, right? Well several industry groups don’t think so. The star-rating’s publication was delayed in April 2016 after many in congress and several industry groups sent a list of concerns to CMS. They asked that CMS:
- Delay the release of the star ratings to allow additional time for CMS to understand the full impact of the current methodology and determine if any category of hospitals are disproportionately disadvantaged
- Share the full star ratings data with hospitals prior to their public release to allow them to review and replicate CMS’s calculations.
- Remove the PSI-90 composite measure and other quality measures from the methodology that are biased against certain hospitals and re-run the results to understand if any category of hospitals are disproportionately disadvantaged by the measures.
- Adjust the star ratings for sociodemographic factors, as has been proposed for the Medicare Advantage star rating system.
The CMS data brief released on July 21, 2016 gives a summary of the star ratings across the nation’s 4,599 hospitals (see infographic below) and concluded that:
“All types of hospitals have both high performing and low performing hospitals. In other words, hospitals of all types are capable of performing well on star ratings and also have opportunities for improvement.”
CMS also shared the methodology used to generate the ratings and had positive feedback on the model from Consumer Reports and LeapFrog. Despite that, the announcement to move forward was met with a more thorough complaint from the industry groups citing additional concerns after reviewing the aforementioned methodology.
My observations:
- The ratings are based on data that’s already reported on the Hospital Compare site, so the objections are all about the methods CMS used to turn those 64 data points into a single composite measure.
- There’s concern that some hospitals will be treated “unfairly” due to sociodemographic factors around the hospital. If you’re comparing two hospitals that would be in competition then they’re likely to have similar sociodemographic factors. I agree that if you compare outcomes at a small hospital in an affluent suburb to those at large, inner-city teaching hospital then you’ve got apples and oranges. I just don’t see where the harm is in that. Those hospitals are not in “competition” and if the worst thing that happens is we acknowledge that outcomes are worse in the inner-city then maybe we can more effectively apply our resources to solve that problem.
- Dr. Janis Orlowski, an executive at the Association of American Medical Colleges, saidthe fact that so many prestigious hospitals fare poorly in the star ratings is a signal that Medicare’s methods are flawed. Maybe, but let’s not jump to conclusions. What if the new rankings are correct and the old ones weren’t? Many of
the rankings that are out there now are arbitrary and the ranking process is closed. When hospitals get a good rating they throw up billboards and wrap busses without questioning the methods. When they get a bad rating they cry foul. - Small hospitals performed best: The preview numbers show 102 hospitals getting a 5-star rating. 50% of those are small hospitals (1-99 beds). There are 2,244 small hospitals and they had the highest mean rating (3.29). More than hospitals with 100-199 beds (2.96) or 200+ beds (2.81). What’s even more interesting is that there were 821 small hospitals that weren’t rated due to lack of data. Think about that. 51 of 2,244 small hospitals got 5-stars. So 3.6% of the rated small hospitals got 5-stars. If that % held on the hospitals that weren’t rated we could pick up another 29 5-star small hospitals. If you do the same projection across all hospitals, you would see 63% of the 5-star hospitals being little guys. Given that the large hospitals control the industry groups, I wonder if these numbers have anything to do with the objections.
CMS should release the star ratings as planned along with the data that was used to calculate them. That way, the hospitals can validate the numbers and dispute them if necessary so that the 2017 ratings are more accurate. Sure, some unflattering ratings may make the news, but that impact will last about 15 minutes and the openness will accelerate our progress towards a viable rating system. There are plenty of ratings out there already and patients just aren’t using them yet. So put the rating system out there, fight it out in public and get it to a place that works for everyone.
#hcbiz 23 Discussion Details
12:00 PM EST – Tweetchat
We’ll setup this week’s show with a tweetchat from 12-12:30 PM EST. We’ll ask 3 questions in 30 minutes:
Q1: Is it a good idea to roll-up 64 measures into a single composite star rating for overall hospital quality (as opposed to reporting them separate only)? What are the pros and cons?
Q2: Do patients pay attention to or even care about quality ratings? Do they impact patient behavior in any meaningful way?
Q3: How are hospitals affected by star ratings? What should they do about them?
Follow the #hcbiz hashtag on Twitter or use an app like tchat.io to join the conversation.