Why pay for mistakes?
THE ANNOUNCEMENT that Medicare will no longer pay hospitals for "conditions that could reasonably have been prevented" is a loud and, many would say, long-overdue wake-up call for American hospitals. Officials are reflecting the rising indignation of the public at the high rate of harm they experience when hospitalized. People have long been baffled by -- and unforgiving of -- the so-called "never" events: sponges left in patients, surgery on the wrong site, mismatched transfusions, etc. But they also don't think they should fall or get infections or pressure ulcers when they go into a hospital, and they think hospitals have been far too blasé about their responsibilities to prevent harm.
Since our Institute of Medicine committee issued the report "To Err is Human" eight years ago, patient safety leaders have been calling on hospitals to get serious about safety, to make a commitment to eliminating preventable injuries, and to implement known safe practices that will prevent them. Some have. Most have not. It was as if the safety folks were speaking a foreign language. Now the hospitals are being spoken to by people with authority and in a language they understand: the language of money.
Hospitals have had the opportunity for some time to help their patients and save money by implementing a number of proven safe practices -- bar coding of medications, computerized ordering, prevention of blood stream and ventilator-associated infections, to name a few -- and most have ducked it. The "business case" for safety has been well-established. Now it is the payers, not the hospitals, who will save the money -- and, if we're lucky, the public -- in the unlikely event that savings get passed on in reduced premiums.
Why now? Why not before this? Three reasons are apparent.
First, the data continue to come in showing high rates of injury -- particularly of infections acquired at hospitals. There is some evidence of improvement; the Institute for Healthcare Improvement conducted a successful "100,000 Lives" campaign to prevent patient deaths. But things are not getting better fast enough.
Second, evidence is increasing that many injuries are preventable if hospitals implement proven practices. Case in point: the dramatic reductions in hospital-acquired bloodstream infections and pneumonia in Michigan intensive care units.
Finally, the public is increasingly frustrated at hospitals' failure to move aggressively to make safety the priority. Thanks to government-sponsored research, the work of the IHI, and many local initiatives, a huge armamentarium of safe practices is available. Why aren't they being implemented?
Hospitals object that they will have to collect additional data, such as to prove that patients had an infection at admission. While true in some cases, they should have been checking high-risk patients anyway. And it will not be necessary to check all low-risk patients. Hospitals will now almost certainly be more interested in screening everyone for antibiotic-resistant staphylococcal infection, however. The (lost) cost of treatment for one patient would pay for hundreds of tests. But other precautions, such as attendants for patients at risk for falls, can be expensive. The new rules will provide a strong incentive for development of more cost-effective methods.
Another concern is that even with the best of care it is not possible to prevent 100 percent of these "preventable" adverse events. But that is what people said about catheter-associated blood-stream infections -- before hospitals began to show that by assiduously following an explicit protocol they could eliminate them. Preventing falls has proven more difficult, but, again, some hospitals have had impressive success. Nonetheless, even the best hospitals will have an occasional misstep. There is no reasonable way to make exceptions to the new policy, so hospitals will have to live with it. It will, however, be important to ensure that the list of noncompensible events is fair -- that there are evidence-based practices available that do, in fact, prevent the injuries.
With so much at stake, it will be essential for Medicare to audit hospital performance and documentation by means other than voluntary data submission. Egregious events, such as removal of the wrong limb, are hard to hide, but less obvious failures can be hidden or not documented. These new rules will improve care, but at the price of more intrusive oversight.
Dr. Lucian Leape, a member of the Institute of Medicine committee that issued the report "To Err is Human," is an adjunct professor of health policy at Harvard School of Public Health. 
