WASHINGTON Physicians may never embrace pay for performance with open arms, but they do need to get in the game.
That was the message delivered by policy experts speaking at the annual research meeting of AcademyHealth.
Hospitals have viewed pay for performance "as something that is coming down the pike, and they're getting ready for that," said Melony Sorbero, Ph.D., a researcher with the RAND Corporation.
In recent interviews that were conducted by RAND as part of the organization's studies of existing pay-for-performance programs, hospital staff expressed much less resistance than did physicians.
"Hospitals have an organizational framework, staff, and systems to be able to respond to these programs," said Cheryl Damberg, Ph.D., a senior researcher with RAND.
For hospitals, the question about pay-for-performance programs is how many measures are being requested and what the technical requirements are for reporting the data.
For physicians, the problem is a fundamental one: How will they collect the data in the first place?
"Physicians for the most part lack the infrastructure. Their data systems aren't anywhere near what hospital data systems are," said Dr. Damberg.
However, physicians do have opportunities to get involved with the development of pay-for-performance measures. There are hundreds of pay-for-performance experiments currently engaging physicians, while a total of only about 40 programs are aimed at hospitals, said Dr. Sorbero.
The American Board of Internal Medicine is behind one of the efforts targeting physicians. The organization recently completed a study to determine whether physicians can be ranked based on a combination of chart reviews, patient surveys, and practice system surveys. They assessed the consistency of those data individually and together.
"We want to make sure that the measures that are going into our composites are fair and reliable," said Rebecca Lipner, Ph.D., vice president of psychometrics and research analysis at ABIM.
The study of physician ranking looked at the treatment of a single medical conditionhypertensiona focus that was key in formulating the patient survey questions, she said.
The questions that are used to survey patients aren't "the general 'do you like your physician?' or 'do you get good access to care?' They're all about how does the physician give care for your specific disease," said Dr. Lipner.
However, ABIM found that there was wide variation across the sets of measures and, depending on how they were combined, an individual physician's rank could swing by more than three quartiles. For example, a physician could do well based on his chart and systems data, but do poorly based on the patient surveys, she said.
One lesson of the study may be that devising a reliable measure of physician performance is not a simple thing to do, Dr. Damberg suggested.
Another lesson may be that it is important for physicians to have a structure within which these performance measures become relevant.
In that sense, medical homes can be seen as an attempt to give a framework to practice settings outside the hospital, said Dr. Lipner.
"A lot of what we have learned from hospital systems is transferring over to medical homes. But it is a big challenge," she noted.
"We have … quite a few physicians who are in solo practice. They are really by themselves, and we always underestimate how many physicians are working by themselves without an infrastructure, without peer connections," Dr. Lipner said.