High charges kept company with high mortality rates at some Colorado hospitals in 1992, according to a state report released last week.
Colorado's first look at hospital outcomes found the state's highest average charge for the conditions tracked-$17,910-as well as higher-than-expected inpatient deaths at 400-bed Presbyterian-St. Luke's Medical Center in Denver. North Suburban Medical Center, a 180-bed for-profit hospital in Thornton, Colo., also showed up with higher charges and more deaths than expected.
"It is old data, but we have to accept the fact that the data was real for 1992," said Wag Schorr, M.D., medical director at Presbyterian-St. Luke's and vice president of medical affairs at its parent system, HealthOne. "We are a different institution in 1994, and we're going to be a vastly different institution as things evolve this year."
In fact, the state outcomes report justifies the steps HealthOne has taken to control its costs, Dr. Schorr said. In April, the system announced plans to chop its charges 5% to 10% after another state report ranked Presbyterian-St. Luke's the highest-priced hospital in Denver (April 11, p. 30). This year, the system cut about 800 positions to bring its total work force to 8,000 at its six hospitals and 30 out-patient facilities. More cutting is ahead as each facility tries to reduce its full-time employees to a ratio of five per operational bed.
Other hospitals with results outside normal ranges also said they would use the information to improve their performance. For example, 326-bed St. Mary-Corwin Regional Medical Center in Pueblo, Colo., had higher-than-expected lengths of stay and charges for major joint replacements.
The hospital suspected as much in 1992 and took steps to improve, said Dean Barnett, its quality-resource coordinator. Now, St. Mary-Corwin knows which hospitals perform those procedures most efficiently, and it can turn to them for help, Mr. Barnett said.
Colorado's report proves again that mortality, charges and lengths of stay vary significantly at hospitals-even after the data is adjusted for patients' severity of illness, said Reid Reynolds, director of the Colorado Health Data Commission, which released the report. The results cover one in five patients admitted to Colorado general hospitals in 1992. Some 24 hospitals submitted information on 28 conditions under a 1990 state mandate.
The commission is using a somewhat controversial program from Westborough, Mass.-based MediQual Systems to adjust data so it reflects the greater odds that more severely ill patients will die.
The MedisGroups program, however, can't account for a number of factors that might contribute to higher death rates at some hospitals. Discharging practices, for example, affect lengths of stay and inpatient deaths. Most importantly, the validity of data depends on how well physicians document patients' illnesses.
In addition, the system's cost has irritated many hospitals required to use it. Last month, Iowa replaced the MedisGroups program with a less-expensive system. Iowa hospitals spent about $2.5 million each year preparing data for the program, while employers weren't using the information because of its complexity, said Perry Meyer, vice president of the Iowa Hospital Association information center.