HMO's report card has some bad marks.

The nation's largest managed health care company has run a battery of tests on its own cardiologists and internists, finding that many failed to follow common medical practices when treating their patients.

United HealthCare Corp.'s pilot evaluation program is designed to make its physicians aware of how their peers most commonly treat such conditions as congestive heart failure, diabetes, and heart attacks. United said its goal is to give doctors a benchmark for determining proper care.

"We want to provide the doctors in our network with information as to what the best practice standards are," said Phil Soucheray, a United HealthCare Corp. spokesperson. `We found there's room for improvement."

For example, the American College of Cardiology recommends heart attack patients routinely be prescribed beta blockers to avoid a second attack, but only 37 percent of recent heart attack patients treated by United cardiologists in Ohio received the drugs. (Thomas M. Burton, An HMO Checks Up on Its Doctors' Care and Is Disturbed Itself, Wall St. J., July 8, 1998, at A1.)

Likewise, Dr. Kenneth Polonsky, a diabetes specialist at the University of Chicago Medical Center, told The Wall Street Journal that all diabetics should get a glucose-monitoring test every year. United internists in Ohio gave these annual tests to only 59 percent of patients, while the percentages were 60 percent in North Carolina, 65 percent in Colorado, and 67 percent in Texas.

These findings are "not so much an indictment of the doctors as [an indicator of] how medicine is being practiced," Soucheray said, adding that veterinarians and dentists may do a better job of tracking patient treatment than physicians.

"The vast majority of veterinary and dental business is based on return visits," he said. "How do you know when it's time for your six-month checkup at the dentist? You get a card in the mail. That's not how doctors have been taught to practice. They are used to treating whoever comes in the door with whatever condition. We are providing them with a spotlight to show that maybe this is a system that needs to be examined."

United patients--and plaintiff attorneys seeking information on a specific doctor's practice, for that matter--should not expect to come by these clinical profiles any time soon. They are for the physicians' eyes only, according to a company-issued fact sheet on the program.

"The goal is to give doctors feedback that will let them improve care, [so] patients won't have access [to the profiles] initially. That could change after doctors have had time, perhaps two years, to make improvements," Soucheray said.

United evaluated the claims records of 1,600 of its cardiologists and internists in Denver; Columbus, Ohio; Greensboro, North Carolina; and Dallas. The company generated clinical profiles to show physicians how often they chose the most commonly recommended treatment for their patients.

For example, a doctor's claims report may show that he or she conducted routine mammograms on 20 of 28 patients, or 71 percent. On average, 74 percent of patients evaluated by United doctors were given the test. According to this profile, the individual doctor would have fallen only a few percentage points short of United's national average.

Lee Newcomer, United's chief medical officer, said the profiles are not meant to be used punitively or to tell doctors how to treat their patients.

"Sharing objective, valuable information about clinical performance and well-established standards of care should benefit all patients, not just those who receive health care coverage from United HealthCare," he said.