Seven days into the patient's hospital stay, his doctors realized they had fought a losing battle. The patient, an overweight smoker with a touch of diabetes, had come to the emergency department with shortness of breath. After a series of tests in the emergency room, he was given a dose of antibiotics for possible pneumonia and admitted to the hospital. His condition deteriorated quickly and now, on a ventilator in intensive care, he was unconscious and a chest x-ray showed that his lungs had completely filled up with fluid. Over the next couple of days, after long discussions with his family, the doctors and nurses withdrew his ventilator and the patient passed away.
Should the hospital be paid for taking care of this patient? Traditionally, hospitals in the U.S. are paid for services rendered, regardless of whether the patient benefits from the services. Intensive care stays cost money, and many patients in such units die despite the care they receive. Hospitals can't afford to be paid only for those patients who survive such ordeals.
More recently, however, healthcare reformers have questioned this payment mechanism, asking why healthcare providers aren't paid more often according to the quality of their care, rather than the intensity of this care. Common business practice holds that when you buy a faulty product, you get your money back. The same doesn't seem to hold for hospital care.
Economics has taught us a number of simple but profound truths about how markets behave. For instance, as the supply of a good increases, the price of that good should drop. When looking at the economic mess we Americans call a "health system," then, it seems only reasonable to turn toward these simple truths when considering how to reform health care.
Hence the power of the pay-for-performance movement in medicine. The idea is simple -- human beings, as economists have taught us, respond to incentives. If we reward doctors for providing high quality care, or punish them financially for providing less than optimal care, then doctors will do better at their jobs.
Unfortunately, as shown in an article in July's issue of the prestigious Annals of Internal Medicine, simple truths often don't work in healthcare. The authors of this study focus on the problem of pneumonia, the illness that killed the patient whose story opened up this essay.
Pneumonia is an infection of the lungs, often caused by bacteria. Pneumonia is often deadly. Therefore, if patients are going to survive this disease, they need rapid treatment with antibiotics. Indeed, when a group of pneumonia experts reviewed the medical records of several thousand pneumonia patients, they discovered that those who had received antibiotics within four hours of arriving at the emergency department were significantly more likely to survive than those who did not receive them so quickly.
It wasn't hard for reformers to look at this evidence and conclude that those doctors who give pneumonia patients antibiotics within four hours should be paid more than those who do not. Pay for performance, not for poor performance.
Only one problem with this plan however -- it creates a hoard of problems.
Pneumonia, you see, is often difficult to diagnose. When a patient comes to an emergency department with shortness of breath, doctors like me are going to wonder whether it's from pneumonia, asthma, or congestive heart failure. We'll also consider more urgent problems, like a potentially-fatal blood clot in the patient's lungs, or a blockage of one of his coronary arteries, leading to a heart attack.
What should we doctors do in these situations? Well of course, we should move quickly to make the diagnosis. If the patient looks relatively stable, we should hold off on any preliminary treatment for a short time, while obtaining the kinds of tests that will point them toward the right diagnosis.
Unfortunately, the right diagnosis doesn't necessarily announce itself in the emergency department. A good chunk of people with pneumonia come to the emergency department without fevers. They are often so dehydrated from their illness, that their pneumonia doesn't show up on chest x-rays. It's pretty common, in fact, for emergency department doctors to treat patients for multiple problems at the same time, because they aren't sure of the diagnosis. I've taken care of many patients in the hospital who, even after a few days of treatment, leave me befuddled as to whether they were short of breath because of congestive heart failure or pneumonia.
Why this brief introduction to the difficult diagnostic life we physicians face in the hospital? To reveal the pitfalls of simple economic schemes (pay for performance, dude!) on the practice of medicine. You see, when insurance companies began rewarding doctors for administering antibiotics to pneumonia patients within four hours of arriving at the emergency department, many physicians raised a proper and justified stink. They pointed out the strange behaviors that would follow from this incentive scheme -- doctors would give too many patients antibiotics, even stable ones who could have waited another hour for test results. Overuse of antibiotics leads not only to antibiotic resistance, but also creates potentially life-threatening side effects. And this payment scheme, by making doctors so concerned about treating pneumonia quickly, diverts their attention away from other possible diagnoses. In other words, the already complicated job of diagnosing sick patients gets even more complicated, thanks to these simplistic reforms.
Beware of business people and politicians who promise that a little common sense -- pay for performance -- will improve our healthcare system. The same kind of quality improvement processes that make companies more efficient at producing widgets, and the same kind of financial incentives that maximize the performance of door-to-door salesmen, won't improve the way doctors like me treat patients who come to us struggling for air.