MedPAC starts testing doctor efficiency

Washington -- The panel that advises Congress on Medicare has taken the first steps toward comparing physicians based on how many program resources they use, with the hope that officials some day can identify more-efficient and less-efficient doctors.

In its June report, the Medicare Payment Advisory Commission analyzed 5% of Medicare claims in 13 major metropolitan areas to see how much the program spent on physician services and other medical care for certain types of conditions. MedPAC used commercially available software to organize results into groups based on each episode of care, or single full course of treatment.

Analysis of the numbers determined that the average cost to Medicare of a single episode of care varied widely, depending on region. For instance, an episode for a patient with type 1 diabetes cost Medicare an average of $600 in Minneapolis but $1,175 in New York. The national average is $833.

Not all of this money is due to physician services. The tabulation for one episode of care includes inpatient services, imaging and other tests, and post-acute care plus physician exams and procedures.

But MedPAC hopes that Medicare officials someday may use these types of total dollar figures to determine how many resources can be linked to the one physician who is most responsible for directing or coordinating the care for the patient. The panel recommended more than a year ago that the Centers for Medicare & Medicaid Services devise a system under which the agency collects episode-of-care information and confidentially shares the results with individual physicians. If some doctors after receiving these reports find that they are ordering many more services and complex procedures than their peers, they may be convinced to be more efficient in the provision of care, MedPAC said.

In most cases, a particular episode in the MedPAC analysis could be attributed to one physician who was shown to be directing the patient's care. In some cases, the evidence was not definitive enough to shine the spotlight on a specific doctor. The panel's next step is to examine 100% of Medicare claims in six metropolitan areas to see if it can develop efficiency measures at an individual physician level.

CMS Administrator Mark McClellan, MD, PhD, said preliminary work by the commission shows promise amid recent trends of rising Medicare costs and growing evidence that more care isn't necessarily better for the patient.

"We want to be able to deliver better care at a lower overall cost, and this is one of the ways that we can encourage that," he said. By linking such information to patient outcomes data and comprehensive quality reports, Medicare eventually could work with doctors to determine when care is needed and when it is not necessary, he said.

But preliminary cost comparison examples from MedPAC don't take into account whether doctors in some areas are seeing relatively sicker patients, a factor that could explain some variances in efficiency. American Medical Association Trustee J. James Rohack, MD, said any program that CMS launches in this area must allow for these considerations.

"Efforts by Medicare to provide confidential feedback to physicians on care provided may help physicians as they work to improve," he said. "However, it is critical that the data be risk adjusted and take into account additional factors beyond what can be gleaned from claims data to be truly useful."

In the meantime, physicians will continue to work on improving care through education, training and consultations with their peers, he said.

A skeptical audience

If the concept of measuring physician efficiency sounds familiar to many doctors, it may be because some private insurers already have gotten in on the act. While neither MedPAC nor CMS have brought up the possibility of linking efficiency -- which the commission dubs "physician resource use" -- to the amount Medicare pays for care, some managed care companies have done just that.

One of the most prominent examples of this comes in the form of tiered physician networks, said J. William Thomas, PhD, a professor emeritus at the University of Southern Maine and an expert on efficiency measurement.

An insurer uses episode-of-care grouping software to determine which physicians are prescribing what it deems as too many services based on patient conditions. It then places those doctors in a less-preferred tier in the network. By charging beneficiaries higher co-payments for using these lower-efficiency doctors, the plans try to drive physicians to improve their scores so they can move to a more-preferred tier.

Such initiatives by health plans have made some physicians skeptical of the entire concept of measuring doctors' efficiency, Dr. Thomas said. While MedPAC and CMS say efficiency reports would be confidential and serve only an educational and not a punitive role, some insurers started out saying the same thing but eventually moved to tiered networks under pressure from employers facing rising health care costs.

"Right now we're dancing on the edge of a knife," Dr. Thomas said. "It has the potential for being very useful to physicians, but it could be very harmful to physicians both in terms of their levels of reimbursement and in terms of their reputations if it is done incorrectly."

Dr. McClellan said the system would benefit doctors who complain that they are blamed for increasing the amount or complexity of services but are not given credit for reducing costs elsewhere. Because the efficiency measurement would take into account all services involved in an episode of care, investigators could identify areas where money is saved.

"Physician decisions are a big focus in resource measurement, but if a physician is seeing a patient more often or ordering a few more tests and that's translating into overall lower costs of care and fewer complications, that's a good thing," he said. "That's what we ought to be focusing on in our payment system."