Performance-based payments to doctors

Just as achievement of student learning outcomes has become the holy grail of education, alignment of incentives among doctors, patients, and insurers is arguably the most challenging problem in healthcare.  In particular, how do we ensure that doctors deliver the most cost-effective treatments - most effective treatment at the lowest cost?

In an interesting experiment, New York city public hospitals have initiated a project to reward doctors for their performance based on a set of parameters related to better patient outcomes, instead of the prevailing volumes or procedures-based remuneration of doctors. Doctors attached to hospitals are generally remunerated based on the income they generate for the hospital. The latter has been found responsible for generating several incentive distortions, primarily in promoting over-treatment.

The New York public hospitals, the largest public health system in the US with 11 large public hospitals and more than a million emergency room visits a year, proposes to link doctors pay increases to performance on bench-marked indicators. The Times, which runs the story, writes,

Under the proposal, bonuses of up to $59 million over the next three years would be distributed to about 3,300 doctors, and would be given to physicians as a group at each hospital, rather than as individuals, so that even the worst doctor would benefit. They would amount to up to 2.5 percent of salaries, which range from about $140,000 for entry-level primary-care physicians to $400,000 for experienced specialists... The public hospital system has come up with 13 performance indicators. Among them are how well patients say their doctors communicate with them, how many patients with heart failure and pneumonia are readmitted within 30 days, how quickly emergency room patients go from triage to beds, whether doctors get to the operating room on time and how quickly patients are discharged. 

However much I am enthused by such initiatives, I am not too optimistic about its success. Already the union of doctors have disputed the set of parameters being used for bench-marking. The selection of parameters will be critical. What constitutes patient outcomes will vary, often widely, based on the category of medical conditions being treated. It will also vary based on the nature of patients, and their expectations. Arriving at a set of parameters that accurately reflect all these and other variations and will be more or less universally accepted may be easier said than done.