The pitfalls of Medicare physician profiling

In 1980, Medicare spent an average of $1,000 per patient. In 2015 health-care expenditures are projected to increase up to $12,000 per person. To help reduce those expenditures, Medicare is turning to physician profiling.

Michael X. Repka, MD, professor of ophthalmology and a professor of pediatrics at the Wilmer Eye Institute, Johns Hopkins University School of Medicine, Baltimore, reviewed the nature of physician profiling.

A variety of profiling agencies can be found on the Internet, including HealthGrades and Leapfrog, which offer quality reports on physicians or hospitals. The trend, Dr. Repka said, is toward consumer-directed searches of physicians by for-profit entities.

"They're not for the greater good, but for getting health care right," Dr. Repka said.

Economic profiling gives a certain level of cache to a commercial health-care plan. It works by profiling every provider in the system for at least 3 years. Most plans are using grouper software to look at performance. It takes the claims database and converts it to cross efficiency, which typically disregards outcomes. The goal is to find providers that cost the system less.

Regional providers can use economic profiling to lump physicians into "cheap" or "expensive" categories, or tiered provider networks. Patients will end up choosing providers the same way they choose a generic versus a proprietary physician today, Dr. Repka said. Paying a higher co-payment means going to a provider of choice. A lower co-payment means choosing a physician in a narrow tier.

"It's differentiation based on cost efficiency," Dr. Repka said.

Grouper software takes claims data and uses a proprietary program to generate answers. It takes all costs-hospitalization, medications, wheelchairs, walkers, ambulatory care centers-and identifies what proportion of those costs is assigned to every physician who submitted a claim for every episode of care. That information is used to calculate an episode of cost per patient.

The problem, Dr. Repka said, is that the calculated actual costs may not really be the actual costs for that episode of care. Aspects of care, which physicians don't have much choice in, still show up as costs under the physician-patients may be sent to more expensive labs for lab work rather than the less expensive, good quality labs on the other side of town, or patients may choose more expensive drugs.

Another problem is small sample size. Smaller samples mean wider variability and the inability to calculate true and accurate costs per care.

Tiered networks

Physicians in subspecialties or specialties have bigger issues with tiered networks. Typically there are too few providers in the favorable tier and those physicians are bombarded with patients. It means patients cannot get appointments and longstanding patient-doctor relationships are broken.

Clear incentives exist for patients to use physicians in the lower-cost tier. Lower co-payments may attract patients to one tier over another. Insurers tend to drive companies or employers into using the tiered-network system because it reduces the employers' premiums.

"Patients believe if a doctor is in this effective care tier, it means there must be a quality effect," Dr. Repka said. "That's not truth in advertising there."

He added that tiered networks carve out the sickest patients and place them in the most expensive tiers. The process makes it difficult to manage referrals, erects barriers to care, and controls costs for the insurer.

Physician profiling and the tiered network it creates, Dr. Repka said, are "an administrative nightmare for specialists not deemed preferred." He recommended that ophthalmologists review his or her contract to see if it allows profiling. If it does, request your profile, review it, and determine if the methodology is sound.