5 Calif. Hospitals Ranked High in Medicare Deaths

Five California hospitals rank among the 50 U.S. facilities with the highest risk-adjusted death rates for Medicare patients in 1986 and 1987, and four are among the 50 institutions with the lowest overall rates, according to a new analysis of recently released U.S. Health Care Financing Administration data.

Public hospitals in the United States make up a disproportionate number of the hospitals with the highest death rates, when compared to their predicted values. The public hospitals in California in this group are Martin Luther King Jr./Drew Medical Center in Los Angeles, Riverside General Hospital, and Valley Medical Center in Fresno; the others, both private, are American River Hospital, a nonprofit institution in Sacramento County, and Charter Community Hospital in Hawaiian Gardens.

Many of the hospitals with the lowest mortality rates for elderly and disabled patients are university hospitals and well-known private institutions. The California hospitals are Cedars-Sinai Medical Center in Los Angeles, Green Hospital of Scripps Clinic in La Jolla, Eisenhower Medical Center in Rancho Mirage, and UCLA Medical Center.

30% Greater Risk

A patient’s chance of dying within 30 days of admission to one of the 50 hospitals with the highest death rates is at least 30% greater than at an average institution, according to Dr. Michael Pine, president of Michael Pine & Associates Inc., the Chicago health care consulting firm that prepared the analysis. By comparison, the chances of surviving more than 30 days after admission to one of the hospitals with the lowest death rates is at least 20% better than at an average hospital.

Many high-mortality-rate hospitals have responded to the federal data by saying that their patients are more severely ill to begin with than patients at other hospitals.

But on the basis of his analysis, Pine said such hospitals need to provide fuller explanations. “I don’t think that hospitals (with the highest mortality rates) can summarily dismiss the data by saying that their patients are older and sicker,” he said.

Pine made available to The Times the mortality rate statistics for California hospitals and for what he identified as the highest- and lowest-ranking hospitals across the country.

Controversy Likely

The analysis is likely to prove controversial because the federal Medicare program has actively discouraged efforts to use its data to rank hospitals or to make comparisons between institutions. When the 14-volume federal report was released last month, the results were couched with so many qualifying statements that many physicians wondered if they had any meaning at all.

The federal government “is in the business of supplying information in a very neutral role,” said Pine, who was retained by the Health Care Financing Administration to advise the agency on the mortality rates report. He has analyzed similar data for the Veterans Administration and the Joint Commission on Accreditation of Healthcare Organizations. “My role is to interpret the data further and make the information more usable and understandable. . . .I don’t think you can take our analysis and say this a good hospital or that is a bad hospital. But one can ask serious questions that are entitled to serious answers.”

Outside of California, the hospitals with the highest mortality rates include urban public institutions in Atlanta, New York City, Seattle and Washington, D.C. The facilities with the lowest mortality rates include Massachusetts General Hospital and four other Boston hospitals, and the Mayo Clinic’s two hospitals in Rochester, Minn.

A senior Health Care Financing Administration official declined to comment on the specifics of the new analysis but said it was a “reasonable approach” to “moving the state of the art (in mortality rate research) forward.” John Spiegel, deputy director of HCFA’s Health Standards and Quality Bureau in Baltimore, said that “(Pine) is a smart guy and a responsible researcher.” As a consultant, he was “very helpful to us in a variety of aspects of the mortality rate analysis.”

Other health care experts and hospital officials cautioned that results of the new analysis might be misleading in some instances, and that the reputations of institutions could be unfairly boosted or tarnished as a result. “It probably does mean something, but it is a question of how much it means,” said Dr. David Auerbach, an assistant professor of medicine at UCLA who helped evaluate the HCFA data for the medical center.

Dr. Mark S. Blumberg, director of special studies for Kaiser-Permanente in Oakland, and Dr. Sidney M. Wolfe, of the Washington-based Public Citizen Health Research Group, both said that hospital mortality statistics for specific diseases, such as heart attacks and strokes, are more meaningful than overall mortality data.

Detailed Review Urged

Dr. Martin F. Shapiro, an associate professor of medicine at UCLA Medical Center, who is studying the differences in mortality between public and nonprofit hospitals in New York City, said detailed reviews of facilities are necessary to interpret the data. HCFA has already directed peer review organizations, private groups that monitor the quality of care for Medicare patients under contract to the agency, to conduct such reviews for all the hospitals with high mortality rates and to report on whether problems exist and what is being doing to correct them.

One striking finding is that while many academic medical centers, such as those at UCLA and UC San Francisco, have lower than predicted mortality rates, their respective public hospital affiliates, such as Harbor-UCLA Medical Center and San Francisco General Hospital, have higher than predicted mortality rates. A similar pattern is seen in New York City and Atlanta.

According to Pine, the analysis “raises questions” about the quality of care provided by physicians in training, who may receive less supervision at public hospitals than at university hospitals. It also raises questions about the adequacy of resources made available for public hospital patients, who may be sicker and poorer on average than Medicare patients treated at private hospitals. For example, such patients might receive inadequate care for common conditions such as heart disease or diabetes before entering the hospital, and develop more severe illnesses as a result.

Patient Concentration

There is a “concentration of the poor, the sickest of the sick and the dying” among Medicare patients cared for at public hospitals, according to Carol B. Emmott, executive director of the California Assn. of Public Hospitals in San Mateo.

In preparing their report, HCFA officials decided not to adjust the data to account for race or socioeconomic status. While such adjustments might have improved the statistics for some hospitals, Dr. William L. Roper, the HCFA administrator, said they would have obscured the fact that minority health status, in general, is significantly worse than that of whites.

Emmott acknowledged that public hospitals might now be able to use the HCFA data to dramatize some of the difficulties they face in caring for a disproportionate number of minority group members and the poor. “It is very difficult to bring the skeletons out of the closet without frightening away the private pay patients, (but) we have come a long way in urging our facilities to let the public hear how bad it is,” she said.

Pine also acknowledged that some of the highest mortality institutions may have an unusual and disproportionate number of nursing home or hospice-type patients, including patients with advanced cancers and other terminal diseases. Conversely, he added, some of the lowest mortality institutions may admit large numbers of patients who are not very sick to begin with.

Footnotes to Data

Several California hospitals, such as American River Hospital, AMI Valley Medical Center in El Cajon, Humana Hospital in Westminster and Riverside General Hospital, specifically cited such factors in explanatory comments about the data they prepared for HCFA.

Humana Hospital, for example, pointed out that 71% of the Medicare patients who died there in 1987 had “do not resuscitate” orders written in their medical charts, signifying that they would be allowed to die if their hearts stopped beating because of the severity of their underlying diseases.

A spokesman at the for-profit Charter Community Hospital said that 68% of its Medicare deaths in 1987 involved patients with such orders and that an additional 12% of the deaths involved patients for whom such orders might have been considered. “We feel that we provide excellent care at our hospitals,” said Steven Smith, vice president of hospitals for FHP Inc., a health maintenance organization that manages Charter Community.

Charter officials also acknowledged that they had submitted incomplete data to HCFA, which may help explain their unfavorable statistics. Dr. Sylvia Seamands, the hospital’s medical director, said the data problem was discovered after The Times requested an explanation of the facility’s statistics.

Dr. Richard Lockwood, director of medical affairs for Valley Medical Center in Fresno, said that, despite the results of the new analysis, the hospital is convinced that “the quality of care here is very good.”

Analysis Goes Further

The new analysis is similar to HCFA’s own, but extends it significantly in several key respects.

First, Pine and biostatistician David W. Smith calculated a “predicted death rate” for each hospital, based on the characteristics of its patients, such as age, sex, major diagnosis, complicating conditions and the number of serious hospitalizations within the last six months. The federal government has not revealed such predicted death rates, which are a standard reporting procedure for health care data. Instead, its statisticians have calculated a less precise range of predicted mortality.

For example, City of Hope National Medical Center in Duarte, which treats a large number of Medicare patients with cancer, had a predicted death rate of 24.8%. But because its actual death rate was only 18.9%, it showed up on the list of hospitals with the lowest mortality rates. By comparison, Riverside General Hospital had a predicted death rate for its Medicare patients of 10.6%. But because its actual death rate was 17.6%, it showed up on the list of hospitals with the highest mortality rates. (Hospitals with typical mixes of patients tended to have predicted death rates of between 11.5% and 12%, Pine said.)

Second, Pine and Smith combined the data for 1986 and 1987, providing more statistical precision than considering each year separately.

Finally, Pine and Smith used standard statistical tests to determine the probability that a hospital’s predicted death rate was in fact significantly better or worse than its actual death rate. The margin of error was 0.1%, compared to a margin of error of 5% for the federal government’s statistics. MEDICARE DEATH RATES IN CALIFORNIA

These California hospitals had the highest and lowest risk adjusted mortality rates for Medicare patients according to an independent analysis of the U.S. Health Care Financing Administration’s 1986 and 1987 mortality data. The analysis includes 5,577 hospitals from across the country; data for both years are combined. It was prepared by Michael Pine and Associates Inc., a Chicago-based health-care consulting firm. The data reflect the percentage of the hospital’s Medicare patients who died within 30 days of being admitted to the hospital, whether as hospital patients or after discharge. The predicted mortality takes into account characteristics of individual patients, such as age, sex, major diagnosis, other medical illnesses and prior hospitalizations, as well as the number of patients treated at each hospital. If a patients was hospitalized multiple times, only the last admission for each year was included. There is a less than one in a thousand probability that a hospital would be listed as having a high or low mortality rate because of chance alone, according to the study’s authors.

Predicted Actual Death Death Hospital Patients Rate Rate HIGH MORTALITY RATES Ranked among the 50 highest mortality facilities in U.S. American River Hospital, Carmichael* 4,242 12.2% 15.8% Charter Community Hosp, Hawaiian Gardens* 2,297 9.4% 13.4% Martin Luther King Jr./Drew Med. Center 865 13.5% 22.3% Riverside General Hospital* 1,559 10.6% 17.6% Valley Medical Center, Fresno* 2,324 13.3% 18.5% Ranked between 51 and 100 AMI Valley Medical Center, El Cajon* 2,540 14.2% 18.4% Humana Hospital, Westminster* 1,893 12.8% 17.8% Kern Medical Center, Bakersfield* 1,188 11.8% 17.9% San Bernardino County Med. Center 1,130 10.5% 15.9% Santa Clara Valley Med. Center, San Jose* 2,431 11.1% 15.1% Ranked between 101 and 200 Brookside Hospital, San Pablo* 3,452 13.2% 16.1% Canoga Park Hospital* 821 13.8% 19.2% Delta Memorial Hospital, Antioch* 1,418 14.1% 17.3% Greater Bakersfield Community Hospital 3,488 13.3% 16.2% Harbor-UCLA Medical Center, Torrance* 1,409 11.3% 15.0% Highland General Hospital, Oakland 964 11.3% 16.1% Kaiser Foundation Hospital, Oakland* 4,130 12.4% 15.2% Pico Rivera Community Hospital* 885 16.6% 23.3% Pioneer Hospital, Artesia* 1,148 11.7% 15.9% San Francisco General Hospital 2,363 11.8% 15.1% Sutter Solano Medical Center, Vallejo 2,000 13.7% 17.1% Ventura County Medical Center 1,106 10.1% 14.2% Verdugo Hills Hospital, Glendale* 3,105 11.6% 14.4% LOW MORTALITY RATES Ranked among 50 lowest mortality facilities in U.S. Cedars-Sinai Medical Center, Los Angeles* 15,002 12.5% 9.9% Eisenhower Medical Center, Rancho Mirage* 6,840 10.3% 7.8% Green Hospital of Scripps Clinic, La Jolla* 3,502 8.9% 5.4% UCLA Medical Center, Westwood* 6,002 10.9% 7.1% Ranked between 51 and 100 St. John’s Hosp. Health Center, Santa Monica 7,094 10.8% 8.7% St. Mary’s Hosp. Med. Cent., San Francisco 4,908 12.4% 9.9% Scripps Memorial Hospital, La Jolla* 6,142 11.3% 9.0% UC Medical Center, San Francisco* 5,090 10.5% 8.0% Ranked between 101 and 200 Century City Hospital* 1,521 11.8% 8.0% City of Hope National Med. Center, Duarte 1,131 24.8% 18.9% French Hospital, Los Angeles 1,239 14.7% 10.5% Hosp. of the Good Samaritan, Los Angeles* 6,842 12.9% 10.9% Kaiser Foundation Hospital, Los Angeles* 6,284 12.5% 10.6%

Veterans Administration and military hospitals are not included. * Indicates hospitals that prepared explanatory comments about their data for the U.S. Health Care Financing Administration. MEDICARE DEATH RATES IN OTHER STATES

These are the hospitals from across the country with the highest and lowest risk adjusted mortality rates for Medicare patients in 1986 and 1987, according to an analysis by Michael Pine and Associates Inc., a Chicago-based health-care consulting firm. Hospitals are listed alphabetically by state. HIGH MORTALITY RATES Cooper Green Hosp., Birmingham, Ala. Maricopa County General Hosp., Phoenix Dist. of Columbia General Hosp., Washington Polk General Hosp., Bartow, Fla. Grady Memorial Hosp., Atlanta Upson County Hosp., Thomaston, Ga. Oak Forest Hosp. of Cook County, Oak Forest, Ill. Provident Hosp. Training School, Chicago St. Joseph Memorial Hosp., Kokomo, Ind. Earl K. Long Memorial Hosp., Baton Rouge, La. Detroit Receiving Hosp.--Univ. Health Cen. Doctors Hosp., Detroit Bergen Pines County Hosp., Paramus, N.J. John F. Kennedy Memorial Hosp., Stratford, N.J. Kimball Medical Center, Lakewood, N.J. Community Hosp. of North Las Vegas Bronx Municipal Hosp. Center, New York City Hosp. Center at Elmhurst, Flushing, N.Y. Coney Island Hosp., Brooklyn, N.Y. Harlem Hosp., New York Jamaica Hosp., Jamaica, N.Y. Kings County Hosp. Center, Brooklyn, N.Y. Nassau County Med. Center, East Meadow, N.Y. Woodhull Med. Mental Health Cen., Brooklyn, N.Y. Metropolitan Hosp., Philadelphia Regional Medical Cen. at Memphis, Tenn. Harris County Hosp. District, Houston Tarrant Co. Hosp. District, Fort Worth Mesquite Physicians Hosp., Mesquite, Tex. Harborview Medical Cen., Seattle LOW MORTALITY RATES Miami Heart Institute, Miami Emory University Hosp., Atlanta Rush-Presbyterian-St. Luke’s Med Cen., Chicago Beth Israel Hosp., Boston Massachusetts Eye And Ear Infirmary, Boston Massachusetts General Hosp., Boston New England Baptist Hosp., Boston New England Deaconess Hosp., Boston Rochester Methodist Hosp., Rochester, Minn. St. Mary’s Hosp., Rochester, Minn. Barnes Hosp., St. Louis Medical Park Hosp., Winston-Salem, N.C. Overlook Hosp., Summit, N.J. Beth Israel Med. Cen., New York Lenox Hill Hosp., New York Manhattan Eye, Ear and Throat Hosp, New York Memorial Hosp. for Cancer & Allied Dist., New York Montefiore Medical Cen., Bronx, N.Y. Mt. Sinai Hosp., New York New York Eye and Ear Infirmary, New York New York Hosp., New York NYU Medical Cen.-University Hosp, New York Columbia-Presbyterian Hosp., New York Cleveland Clinic Hosp., Cleveland Ohio State Univ. Hospitals, Columbus Eye and Ear Hosp. of Pittsburgh, Penn. Pennsylvania Hosp., Philadelphia Thomas Jefferson Univ. Hosp., Philadelphia Wills Eye Hosp., Philadelphia St. Thomas Hosp., Nashville, Tenn. Univ of Texas--M.D. Anderson Hosp., Houston Virginia Mason Hosp., Seattle