Prof. Joseph Doyle: Cutting health care costs may have negative impact on patients

MIT Sloan Assoc. Prof. Joseph Doyle

No matter how the U.S. Supreme Court rules, one of the hopes underlying health care reform – that the steep cost curve of medical care can be reduced without jeopardizing patient outcomes – will remain. But a new study by me and some fellow researchers challenges that basic view.

We studied a “natural experiment,” based on the random assignment of ambulance companies dispatched to treat patients and the resulting random assignment of those patients to different types of hospitals.  Some patients ended up in higher-cost facilities such as teaching hospitals, while others received their care in lower-cost hospitals. The result? We find that when we can compare similar patients who are treated at different hospitals, higher-cost hospitals had better outcomes than those who ended up in the lower-cost ones. With other factors being equal, patients in higher-spending hospitals had one-year mortality rates as much as 30 percent lower than those who received their care in the lower -cost facilities. We cannot necessarily cut spending and expect quality to stay the same

Our findings, which used Medicare data from 2002 to 2008 and has just been published by the National Bureau of Economic Research, challenge current health policy literature, which generally concludes that while teaching and other high-cost hospitals differ in many ways from non-teaching hospitals, patients do about as well at both.  Instead, our results show that patients do better in different situations. More research is needed into what is causing this reduction in mortality at higher-spending facilities. We need a much better understanding of what care is effective at saving lives.

Joseph Doyle is the Alfred Henry (1929) and Jean Morrison Hayes Career Development Professor; Associate Professor of Applied Economics

See the full NBER Working Paper

Read more in the Washington Post WONKBLOG

Post a Comment

Required fields are marked *

*
*