For the second year in a row, none of Utah's 39 hospitals had unusually high death rates among the Medicare patients they treated, according to federal government statistics released Thursday.

The government figures show mortality rates for Medicare patients treated at nearly 6,000 U.S. hospitals during 1987. The Utah numbers profile 26,643 patients, more than half of them at three Salt Lake hospitals and one each in Provo and Ogden.Nationally, some 3.2 percent of hospitals had higher-than-predicted mortality rates based on an elaborate statistical model.

Ron Bouch, director of communications for the Utah Hospital Association, and some hospital administrators expressed concern that health-care consumers would give too much weight to the statistics compiled by the federal Health Care Financing Administration, which oversees Medicare and Medicaid.

"We believe it's a good attempt to help the consumer, but it still has inherent limitations in its data," Bouch said, because the figures don't differentiate between types of patients and institutions.

"It's a useful tool for health institutions but can be confusing to the consumer - even frightening," Bouch said.

HCFA Administrator William C. Roper emphasized the mortality report should not be used as the single source of information on the quality of health care at a particular hospital.

Seventeen sets of figures were reported for each hospital, including 16 diagnostic categories such as cancer or renal failure. Each category contains an actual mortality rate and a "range of mortality" that might be expected at that hospital given its particular patient mix.

For example, LDS Hospital in Salt Lake City treated 4,469 Medicare patients, 2,054 more than the runner-up, Utah Valley Regional Medical Center in Provo. LDS' actual overall mortality rate was 8 percent; its range of predicted mortality was 8-12 percent. The Provo hospital's actual death rate was 12 percent, the predicted range 9-13 percent.

While the 1987 figures show none of the state's hospitals exceeded its predicted range for overall mortality, there were 19 instances in which a hospital exceeded the mortality rate in one of the diagnostic categories, or about 3.1 percent of the cases. In 1986, there were 20 such instances.

But the percentage is statistically misleading since in 17 of the 19 higher-than-predicted categories, there were three or fewer patients. If those patients died, the predicted mortality for that category was exceeded.

None of the 39 hospitals exceeded the predicted range in more than two categories and 23 had no such instances.

Of the 10 hospitals with more than 1,000 Medicaid patients, four came within 2 percent of the high end of their predicted overall mortality rate. They were Utah Valley Regional Medical Center, St. Benedict's Hospital in Ogden and Cottonwood and St. Marks in Salt Lake.

The larger hospitals that came within 1 percent of matching the low end of their predicted overall mortality rates were LDS, Dixie Medical Center in St. George, Holy Cross Hospital in Salt Lake City, Humana Hospital Davis North in Layton, McKay-Dee Hospital in Ogden and the University of Utah Health Sciences Center in Salt Lake City.