I was pleased to see the Chicago Tribune carry an op-ed piece yesterday by my friend and colleague Michael Millenson. The gist of the piece was that information about hospital quality is readily available online and we should use that information when choosing a hospital. Michael is right — there is no shortage of places to turn to get information about hospital quality. But I think he waxes too enthusiastic.
For one thing, it is not clear whether the widespread availability of quality information is a boon or a problem. Consider Medicare’s Hospital Compare website. Look up quality information for pneumonia and you are overwhelmed with nearly 20 different measures on four different web pages. I couldn’t possibly make sense of all this information even if I used sophisticated computer software; how could the average person sort through it all? One quality measure seems to stand out – mortality. But I wonder if this should be a major concern for pneumonia patients. Are we talking about 5 percent mortality rates, or 0.05 percent? I don’t know and Medicare won’t tell me.
HealthGrades.com is much simpler – it just reports mortality. The widely respected Leapfrog Group reports mortality for pneumonia and also reports another 8 general hospital quality measures, some of which are derived from even more measures.
When reading these report cards I find that my local hospital in Highland Park scores very well on mortality in the HealthGrades and Leapfrog reports but I can’t find it anywhere at the Medicare website. And I wonder if the low mortality rate is due to the hospital or due to the demographics of the patients. Michael Millenson pointed out that these report cards are risk adjusted, but he failed to mention that the available risk are pretty lousy – mostly controls for age, sex, and a few comorbidities. (Much better risk adjustment is possible but requires data not available to Medicare, HealthGrades, or Leapfrog.) Hospitals that get poor quality scores often report that their patients are sicker than the risk adjusters give them credit for. They might be right. Hospitals that get good scores never claim that their patients are healthier. Maybe they are hiding something.
We tend to believe that more information is always better. So everyone in healthcare is gathering and reporting more information, and folks like Michael Millenson are exhorting patients to use it. They may be right about this, but I wonder. Information changes decisions; poor information distorts decisions in unwelcome ways. There is lots of theory and evidence to suggest that healthcare report cards are currently constructed do little good and sometimes do a lot of harm.
In the past I have blogged about what it will take to get credible report cards. Integrated Health IT and more comprehensive outcome measures would be very helpful. The Patient Reported Outcome Measurement Information Set (PROMIS) being developed at Northwestern University would be amazingly helpful.
Today is opening day for my beloved White Sox so let me use a baseball analogy. Minor league baseball is cute and occasionally worth watching. The best you can say is that it prepares some players for the real thing – the major leagues. Today’s health care report cards are cute and occasionally worth using. Let’s hope they are preparing us for the real thing. You can root all you want for the Toledo Mud Hens but I won’t. Give me the major leagues or give me nothing.