Pay for Performance: Unintended Consequences

A cardiologist in yesterday's NY Times reports on a patient who received days of unnecessary antibiotics -- which he attributes to a Medicare Pay for Performance program encouraging rapid antibiotic treatment for community acquired pneumonia. The patient had a severe antibiotic complication, and spent two weeks in the hospital. Hence, pay for performance and public reporting, in this instance, led to higher costs.

There is good evidence that public reporting and pay for performance do increase performance in areas of measurement. Of course, they can distract attention from other important issues ("crowd out") and in many instances, these programs encourage correcting underuse - this will also increase costs.

Sometimes, these programs can encourage providers to rethink core processes. Let me give you an example. The sooner a patient with an acute heart attack has an angioplasty, the more likely she will have less damage to the heart. Nonetheless, historically "door to balloon" times (from the emergency room door to the catheterization lab) have been very high. For instance, the best hospitals are almost always able to complete an angioplasty within 90 minutes of a patient's presentation -- many other hospitals fail to do this more than half the time! Not many patients are reading the public reporting - but hospital administrators and physicians are - and care is improving as a result.

Back to Dr.Jauhar's example , I believe that hospitals can diagnose community acquired pneumonia in a hospital in less than 6 hours from patient arrival without resorting to indiscriminate use of antibiotics. Public reporting and pay for performance shines a bright light on this important issue. In many hospitals, clinicians are figuring out how to reduce steps, eliminate bottlenecks, delegate authority to non-physicians, and reduce the cycle time before treatment for pneumonia. In the end, that should improve quality AND reduce costs - but there are bumps along the way.