Better Data Would Help Us Get Better Hospital Care
UPMC touts the fact that it’s the only Pittsburgh hospital on the U.S. News and World Report “honor roll” of best hospitals in the country. The West Penn Allegheny Health System (WPAHS) promotes the fact that it’s the only health system designated as a top national performer by Thomson Reuters, and notes that Allegheny General Hospital (AGH) is also recognized by U.S. News and World Report as one of the country’s best hospitals.
Should you use either of these rankings to choose a hospital? Ideally, what you’d like to know is whether a hospital will improve your health rather than worsen it, e.g., how likely you are to die, be injured, or be infected during your hospital stay, or to be readmitted because of complications you experience after discharge. You’d also like to know if the hospital staff will treat you respectfully, respond promptly when you need help, and make sure you’re not in pain.
The U.S. News and World Report “Best Hospitals” list only looks at two outcome measures (patient survival and safety), and gives them relatively low weight. Only 32.5% of a hospital’s score is based on how often patients survive, and a mere 5% is based on patient injury rates. Another 30% is based on “structural measures” like whether the hospital has the latest technology and how many nurses it has. The remaining 32.5% is based on an opinion poll: 200 randomly selected doctors around the country in each specialty are asked to pick the top 5-10 hospitals for “inpatient care for the most complex or difficult conditions [in that specialty].” Patients’ ratings of their experience don’t count at all.
Let’s look at the rankings for cardiac care, since heart problems are the most common reason people are hospitalized. The most recent “Best Hospitals” report shows that AGH performs better than UPMC on patient safety, and that both hospitals performed the same on patient survival. (“UPMC” here means only Presbyterian/Shadyside Hospitals; the other UPMC hospitals are ranked separately and have lower scores.) But UPMC was ranked higher (#20) than AGH (#44) because 5.2% of the cardiologists surveyed listed UPMC as one of the best U.S. hospitals for cardiac care, whereas only 1% put AGH on their list. For U.S. News, that small difference in the opinion poll outweighed the better patient safety scores at AGH. (The #1 hospital was the Cleveland Clinic, which had better patient survival rates than either UPMC or AGH and was chosen as a top hospital by 75% of the physicians surveyed.)
If you’re thinking about a knee or hip replacement, both UPMC Presbyterian/Shadyside and AGH are on the “Best Hospitals” list for orthopedic care. Here, AGH outperformed UPMC on both patient survival and patient safety, but AGH was again ranked lower overall because a smaller percentage of the 200 physicians surveyed nationally picked AGH as a best hospital. The same was true for cancer care, diabetes/endocrinology, gynecology, and other specialties.
The bottom line is that for the most common types of conditions, the limited objective data in the U.S. News rankings give a slight edge to the quality of care at AGH, but also show that neither AGH nor UPMC has outcomes as good as many other hospitals in the country.
Thomson Reuters ranks hospitals based only on objective data, not opinion polls, and uses a broader range of measures of outcomes, care process quality, patient experience, and cost of care. (The patient survival measures used by U.S. News are actually generated by Thomson Reuters.) None of the hospitals in our region made Thomson Reuters’ list of “100 Top Hospitals” in 2011, but AGH made the list of “50 Top Cardiovascular Hospitals,” and WPAHS was ranked in the top 20% on the list of “Top Health Systems.” However, since Thomson Reuters doesn’t make the underlying data public, you can’t easily find out why any hospital did or didn’t make the lists, so they’re not of much value for patients who want to choose a hospital.
A number of quality measures (including many of those used by Thomson Reuters) are publicly available for each of the hospitals in our region on the U.S. Department of Health and Human Services’ Hospital Compare website (www.hospitalcompare.hhs.gov). However, the outcome measures are based only on patients admitted for a heart attack, heart failure, or pneumonia. Although some hospitals in our region have better outcomes than others, most of the differences aren’t big enough to be statistically significant. Once again, though, there are hospitals in other parts of the country with significantly better outcomes than any hospital here.
In sum, the limited data available indicate that you can get good quality care for the most common types of health conditions at both UPMC hospitals and non-UPMC hospitals, but that all hospitals here could deliver much better care than they do today.
We shouldn’t be satisfied with anything less than the best possible healthcare, but today, we don’t have enough data to know whether we’re getting it. People living in places like Iowa, Minnesota, and Wisconsin can access far more data on the quality of hospital care in their communities than we can. They can find out how often patients get blood clots and infections after different kinds of surgery in each of their local hospitals, how often mothers and babies are injured during labor and delivery, and a host of other measures of quality. The reason is their hospitals have made a commitment to measure and publicly report on their performance. And because they publicly report and compare their performance, they are also working hard to improve the quality of care they deliver.
Few people would buy a car without first looking at ratings of the car’s quality and safety. How can we choose hospitals without similar information? And how can hospitals improve if they don’t know how well they’re performing? Pittsburghers should demand that their hospitals publicly report data on the quality of care in a common format, along with the prices of that care. That would enable us to make informed choices about which hospitals to use, and it would encourage our hospitals not just to be good, but to be the best in the nation.
(A version of this post appeared as the "Regional Insights" column in the January 1, 2012 Pittsburgh Post-Gazette.)