CMS released star ratings for thousands of hospitals around the U.S. in July, receiving mostly negative feedback. This feedback has come from everyone from industry stakeholders to care providers to members of Congress who believe the ratings aren’t fair. I’m supportive of the push for transparency, but I have to say I agree with them in regards to fairness.
I understand why CMS felt this was necessary. As the consumerization of healthcare continues to gain momentum, it makes perfect sense that people would want to know which hospital will give them the best care simply by looking at its rating, as they would when selecting a movie or restaurant. However, there is a big problem with treating your healthcare in that same way: boiling down quality to a star rating will not give you an accurate view of a hospital or health system, and there are a number of reasons why not.
You can’t measure performance if your data isn’t accurate
Having been a chief medical officer at five different healthcare organizations, I understand the pressures hospitals face to not only provide quality care but also achieve top rankings. I also know that the CMS star ratings, and others like them, are based on billing data. That’s a problem. Probably the biggest issue with these ratings is the fact that it assumes all of the hospital’s billing data is correct and that everyone is documenting the same way.
In fact, this data is often incorrect and incomplete due to a disconnect between clinical and billing speak. Physicians tend to document clinically, and while that may be crystal clear to another physician, this does not translate accurately to billing data. There can be miscommunications between which codes impact Diagnosis-Related Groups through ICD-10 and which do not. Physicians often times do not document with a billing-focused mindset, so things can be interpreted by coders in a different way than they were intended by the physician.
Every time I’ve implemented a Clinical Documentation Improvement (CDI) program, I’ve seen significant, positive change in quality metrics, such as perceived mortality. But the amazing part is that I didn’t do anything to change the care that was given – just the way it was documented. This goes to show that these ratings will, of course, be skewed if the data behind them is not accurate and complete.
Comparing apples to oranges
Averaging out quality across different areas is not a true reflection of care. What I mean is there are 64 measures of all different types. They paint different pictures. Certain hospitals could be exceptional at mortality and average in-patient experience, or vice versa. The rating system takes these two separate things, for example, lumps all the information together, and then averages it. The problem is that doing so negates individual things that are vitally important to the overall care a hospital provides, and that likely are very important to consumers when deciding where to go for care. Consumers may see a high rating, and assume that provider will deliver the best care, but they are being misled because they cannot know if the rating is high due to exceptional care or exceptional patient experience.
We all have different wants and expectations—for example, one patient may want a hospital that is exceptional at a certain type of surgery, but doesn’t mind if it is considered “average” for patient experience. Another patient may be looking more for an exceptional experience, and is comfortable with more of an average rating on the procedure they intend to receive. Today’s ratings don’t provide this level of detail and leave consumers thinking they have the answers without a way to distinguish what kind of care they will receive at one hospital versus another.
Let’s do it right the first time
Make no mistake—this is a complicated situation. It is understandable that the government and the industry want to simplify things for consumers. We should start by capturing patient information correctly at the front end of documentation so physician notes paint an accurate picture of a patient’s needs including patient severity. For coders, having this clarity from the start decreases the number of variables they will have to worry about later and produces more accurate billing. Then the whole process improves. The earlier the information is correct and complete, the better it is for care and quality measures in the long run, and then comparing hospitals will be based on a level playing field.
The answer is not to leave consumers on their own. But star ratings and others like it have oversimplified things to the point where it’s not accurate. Thankfully, we live in a world today with different technologies that can easily and fully capture a patient’s story. If we can get the healthcare system aligned to understand how important this is, then eventually these star ratings could become more reflective of the truth.