MEDICINE HOSPITALS : Medicare Death Rates: 72 Facilities Exceed Predictions
More than 70 California hospitals, including Humana Hospital-Westminster, had significantly higher than predicted death rates for Medicare patients between 1986 and 1988, and 28 had rates significantly lower than predicted, according to an analysis of the U.S. Health Care Financing Administration’s third annual report on mortality among Medicare patients at the nation’s hospitals.
Also among the 72 California hospitals with higher than predicted rates were California Medical Center and Martin Luther King Jr./Drew Medical Center in Los Angeles, Valley Hospital Medical Center in Van Nuys and two Kaiser Foundation hospitals.
Those with lower than predicted mortality rates included City of Hope National Medical Center in Duarte, Green Hospital of Scripps Clinic in La Jolla, seven Kaiser Foundation Hospitals, Cedars-Sinai Medical Center and UCLA Medical Center in Los Angeles and UC San Francisco Medical Center.
The HCFA report was made public Wednesday. The analysis of the report was prepared by Michael Pine & Associates Inc., a Chicago health care consulting firm that was a consultant to HCFA in preparing its report.
Dr. Michael Pine, the president of the firm, made available to The Times a list of the California hospitals with the highest and lowest overall mortality rates for Medicare patients for the three-year period, as compared to their predicted death rates.
The statistics reflect the percentage of a hospital’s Medicare patients who died within 30 days of being admitted to the hospital, whether as hospital patients or after discharge, as compared to its predicted percentage.
A hospital’s predicted mortality rate takes into account patient characteristics such as age, sex, chronic diseases and previous hospitalizations to assess their relative risk of dying there.
The mortality information “is not a direct measure of the quality of care, but it is a valuable tool to identify potential problems for review and action by hospital administrators and medical staff,” said Louis B. Hays, HCFA acting administrator. Hays said consumers should “consult their physicians when evaluating the report on a particular hospital.”
Hospitals should “begin to incorporate not only this data but also statistical techniques in day-to-day quality improvement activities,” Pine said.
Officials of Humana Hospital-Westminster Wednesday expressed disappointment with the conclusions of the analysis and defended the hospital’s patient care.
“I think we are doing everything that can possibly be done in the area of quality assurance and morbidity and mortality,” said Executive Director Joel Premselaar. “We have evaluated every Medicare death and have found no unusual circumstances other than the fact that the patients are very elderly and have multiple illnesses.”
The Westminster hospital was among 32 nationwide to exceed predicted death rates for three straight years--1986, 1987 and 1988--according to the HCFA report.
Premselaar said the statistics are in part a reflection of the hospital’s high-risk patient population. Of the patients admitted to the hospital in 1988, for example, he said, 50% were admitted to the intensive care unit. Seventy-four percent of those who died at the hospital that year were admitted through the emergency room, and 62% of those who died were admitted in such a condition that their families and physicians had decided to avoid heroic efforts to save them.
Premselaar also disputed the accuracy of some of the data on which the HCFA conclusions are based.
The hospital treated about 1,300 Medicare patients last year, many more than the 789 the HCFA study identifies, he said. One death the HCFA reports occurred at Humana actually occurred at a Veterans Adminstration hospital, he said.
Premselaar’s defense of care at the hospital was backed up by a recent survey of the facility, state health officials said.
Humana Westminster was one of 50 hospitals nationwide that, because it had higher than expected death rates for two years in a row, the HFCA recommended undergo more detailed evaluation this year by state health agencies. A team composed of representatives from the state Health Services Department and the California Medical Assn. conducted a review of the facility in May and found no major code deficiencies, said Jacqueline Lincer, district administrator of the Health Services Department licensing and certification division in Santa Ana.
“There was a full inspection of Humana Westminster, and with the exception of some fire safety code deficiencies, we did not identify anything that would lead to termination of (Medicare) funding,” Lincer said.
Premselaar said that all of the code violations are being corrected.
The HCFA statistics released Wednesday allow a hospital’s actual performance to be compared to its expected performance; they are not intended to be used to make comparisons between institutions. In addition to overall death rates, the HCFA report includes data on common causes of death in the elderly such as heart disease and strokes.
HCFA also released the results of studies that it said demonstrate the value of the report as an indicator of the quality of medical care.
For example, HCFA said that hospitals with high mortality rates are more likely to have quality problems identified after detailed case-by case reviews than hospitals with lower ones.
On the other hand, hospitals with low mortality rates are more likely to have favorable characteristics, according to a report by Medical College of Wisconsin and HCFA researchers published in today’s New England Journal of Medicine.
Compared to the average hospital, these favorable characteristics include a higher percentage of physicians who are board certified specialists, a higher percentage of registered nurses, higher occupancy rates, and better availability of high-technology equipment.
But the researchers caution that these findings, which are based on “patterns of performance in a large number of hospitals,” do not imply that a detailed review of a specific hospital with a high mortality rate will find quality problems.
As in previous years, hospital associations downplayed the significance of HCFA’s data and consumer groups applauded its release.
“The mortality data only tells you a very little bit about the quality of care that is delivered,” said Carol McCarthy, president of the Chicago-based American Hospital Assn.
Dr. Sidney Wolfe, director of the Public Health Research Group in Washington, said the federal government is “clearly moving in the right direction” but should continue to improve the annual reports by including data on individual surgeries and medical illnesses and on individual physicians.
The 14-volume report is based on a computerized analysis of about 10.5 million Medicare admissions at nearly 6,000 acute-care hospitals during 1988.
The 1988 report is similar to the two previous reports, although HCFA said the statistical methods have been “refined and improved.” For example, the percentage of emergency cases and transfers from nursing homes are taken into account. Such patients are at greater risk of dying than the average Medicare patient.
Mortality data for 1986 and 1987 have been reanalyzed with the improved techniques. These results are included in the new report, as is summary statistical information on each hospital’s services and staffing. Mortality is listed for “all causes” as well as for 16 individual “diagnostic categories.”
Hospitals were given a chance to review their statistics in advance and to provide explanations, which were published as part of the report. About 1,360 hospitals, or 24% of the total, prepared such comments, compared to 51% of the hospitals for the report released in 1986.
In general, hospitals such as Humana Westminster that had higher than expected mortality rates defended the quality of their care. The most common explanations included the high-risk nature of their cases or discrepancies in data.
“Hospitals initially felt threatened by the publication and commented in response to that apprehension,” HCFA officials said in a written statement made available to reporters. “Hospitals now realize that the publication is perhaps of greater use to them than they originally believed. Rather than taking the time to object, hospitals have studied the results.”
Thomas Morford, executive director of the HCFA Health Standards and Quality Bureau, which prepared the report, said that hospitals with consistently high Medicare mortality rates would be scrutinized by federal survey teams and Medicare watchdog organizations such as the California Medical Review.
“We don’t take punitive action” because a hospital has a high mortality rate, Morford said. “When presented with this information, I say, ‘Let’s look further.’ ”
Such a survey of King Hospital in Los Angeles, based on its high 1987 mortality rates, discovered scores of patient care deficiencies and led to a threat that federal funds would be cut off. On Monday, the federal government, citing a “dramatic turnaround” at King, said Medicare and Medi-Cal funds would be continued.
The report, “Medicare Hospital Mortality Information, 1986-1988” can be ordered from the Health Care Financing Administration, Health Standards and Quality Bureau, 2-D-2 Meadows East Building, 6325 Security Blvd., Baltimore, MD 21207, or by calling (301) 966-1133.
Steinbrook reported from Los Angeles and Vickers from Washington. Staff writer Carla Rivera contributed from Orange County. Times researcher Tracy Thomas in Los Angeles assisted in preparing this story.
MEDICARE DEATH RATES
These California hospitals had the highest and lowest mortality rates for Medicare patients as compared to their predicted death rates between 1986 and 1988, according to an analysis of U.S. Health Care Financing Administration data by Michael Pine & Associates Inc., a Chicago health care consulting firm. The HCFA data was compiled earlier this year and made public Wednesday. Pine & Associates was a consultant to HCFA on the preparation of the report.
The data reflects the percentage of a hospital’s Medicare patients who died within 30 days of being admitted, whether as patients or after discharge. The national average 30-day mortality rate is about 12%.
This list of high and low mortality rate hospitals, which was prepared by Pine & Associates, differs in some instances from the report released by HCFA, although it is based on the same data. This is because different statistical techniques were used and data for the three years was combined.
Hospitals listed in boldface type had statistically significant high or low mortality rates for 1988, as well as for the entire three-year period.
An asterisk indicates hospitals that sent explanatory comments about their data to HCFA.
HIGHER THAN EXPECTED MORTALITY RATES
PATIENT DEATH HOSPITAL PATIENTS RISK RATE American River Hosp, Carmichael* 6,643 1.00 16.0% AMI Valley Med Center, El Cajon* 3,553 1.39 19.0% California Med Center, LA* 5,701 1.13 16.2% Canoga Park Hospital 1,205 1.24 18.5% Charter Comm Hosp, Hawaiian Gardens* 4,821 .81 12.7% Charter Suburban Hosp, Paramount 1,780 1.35 19.0% Christian Hosp Med Center, Perris 2,010 1.19 17.2% Coastal Comm Hosp, Santa Ana 1,710 1.24 17.6% Dameron Hospital, Stockton 5,329 1.12 15.3% Doctors Hosp of Lakewood 4,686 1.17 16.1% Doctors Hosp of Montclair 2,006 1.07 15.6% Fremont Med Center, Yuba City 3,792 .85 12.6% Friendly Hills Reg Hosp, La Habra 2,617 1.01 14.6% Granada Hills Comm Hospital* 2,575 .99 13.9% Greater Bakersfield Mem Hosp 5,354 1.12 15.4% Hemet Valley Hospital* 14,593 .91 13.1% Humana Hosp, West Anaheim* 3,008 1.26 17.4% Humana Hosp, Westminster* 2,672 1.17 18.5% Inter-Community Hosp, Covina* 6,157 .90 12.8% Kaiser Foundation Hosp, Oakland* 6,359 1.09 14.6% Kaiser Foundation Hosp, Walnut Creek* 4,258 1.06 14.5% Lodi Mem Hospital* 4,071 .96 13.0% Marian Hospital, Santa Maria 4,010 1.11 15.4% ML King Jr./Drew Med Center, LA* 1,323 1.23 22.7% Memorial Hospital, Modesto* 6,124 .98 13.8% Mercy Medical Center, Redding* 6,327 .95 13.7% Methodist Hosp of So Ca, Arcadia 9,250 .90 11.9% Methodist Hospital, Sacramento 4,132 1.17 16.1% Ontario Community Hospital 1,400 1.23 18.3% Parkview Comm Hospital, Riverside* 4,286 1.12 14.9% Pico Rivera Community Hospital 1,328 1.71 24.8% Pioneer Hospital, Artesia* 1,786 .93 14.6% Queen of Valley Hosp, West Covina* 4,966 1.23 16.3% Redlands Community Hospital* 6,364 .91 14.0% Rideout Mem Hospital, Marysville 4,076 1.01 14.6% Riverside Community Hospital 7,347 1.01 13.7% Riverside General Hospital 2,318 1.05 16.9% RF Kennedy Med Cent, Hawthorne 3,203 1.22 16.8% San Antonio Comm Hospital, Upland 7,129 1.01 13.6% San Bernardino County Med Cent* 1,601 1.03 16.8% San Joaquin Comm Hosp, Bakersfield* 6,664 .99 13.9% Santa Clara Valley Med Cent, San Jose* 3,510 1.01 14.6% St. Francis Med Cent, Lynwood 7,712 1.08 15.1% St. Mary Desert Valley Hosp, Apple Valley 4,346 .90 12.5% Sharp Mem Hospital, San Diego 9,805 .81 10.8% Valley Hosp Med Cent, Van Nuys 2,870 1.20 17.2% Valley Med Center, Fresno* 3,385 1.25 17.8% Verdugo Hills Hospital, Glendale* 4,725 1.04 14.1% Whittier Hospital 2,916 1.18 16.9%
% HIGHER THAN HOSPITAL EXPECTED American River Hosp, Carmichael* 35.2 AMI Valley Med Center, El Cajon* 14.7 California Med Center, LA* 19.4 Canoga Park Hospital 24.7 Charter Comm Hosp, Hawaiian Gardens* 32.0 Charter Suburban Hosp, Paramount 17.7 Christian Hosp Med Center, Perris 21.0 Coastal Comm Hosp, Santa Ana 19.2 Dameron Hospital, Stockton 14.0 Doctors Hosp of Lakewood 15.1 Doctors Hosp of Montclair 22.6 Fremont Med Center, Yuba City 23.9 Friendly Hills Reg Hosp, La Habra 20.7 Granada Hills Comm Hospital* 17.8 Greater Bakersfield Mem Hosp 15.9 Hemet Valley Hospital* 20.6 Humana Hosp, West Anaheim* 15.7 Humana Hosp, Westminster* 32.6 Inter-Community Hosp, Covina* 19.5 Kaiser Foundation Hosp, Oakland* 12.9 Kaiser Foundation Hosp, Walnut Creek* 14.9 Lodi Mem Hospital* 13.3 Marian Hospital, Santa Maria 16.4 ML King Jr./Drew Med Center, LA* 54.4 Memorial Hospital, Modesto* 18.1 Mercy Medical Center, Redding* 20.8 Methodist Hosp of So Ca, Arcadia 10.4 Methodist Hospital, Sacramento 15.1 Ontario Community Hospital 24.6 Parkview Comm Hospital, Riverside* 11.8 Pico Rivera Community Hospital 21.5 Pioneer Hospital, Artesia* 31.4 Queen of Valley Hosp, West Covina* 10.8 Redlands Community Hospital* 28.3 Rideout Mem Hospital, Marysville 21.0 Riverside Community Hospital 13.5 Riverside General Hospital 34.8 RF Kennedy Med Cent, Hawthorne 15.5 San Antonio Comm Hospital, Upland 12.6 San Bernardino County Med Cent* 36.4 San Joaquin Comm Hosp, Bakersfield* 17.8 Santa Clara Valley Med Cent, San Jose* 20.7 St. Francis Med Cent, Lynwood 17.6 St. Mary Desert Valley Hosp, Apple Valley 16.5 Sharp Mem Hospital, San Diego 12.3 Valley Hosp Med Cent, Van Nuys 20.8 Valley Med Center, Fresno* 19.1 Verdugo Hills Hospital, Glendale* 13.7 Whittier Hospital 20.0
LOWER THAN EXPECTED MORTALITY RATES
PATIENT DEATH HOSPITAL PATIENTS RISK RATE Cedars-Sinai Med Center, LA* 22,529 .98 10.1% Century City Hospital, LA 2,321 .99 9.2% City of Hope Nat Med Cent, Duarte 1,608 2.02 18.6% Eisenhower Mem Hosp, Rancho Mirage 10,613 .84 8.4% French Hospital, LA 1,771 1.26 10.8% Hoag Mem Hosp, Newport Beach 9,801 1.06 11.3% Kaiser Foundation Hosp, Anaheim 3,846 1.06 10.5% Kaiser Foundation Hosp, Bellflower 5,729 1.03 10.9% Kaiser Foundation Hospital, LA 9,899 1.17 10.8% Kaiser Foundation Hosp, Panorama City 6,693 .92 9.1% Kaiser Foundation Hosp, San Diego 7,257 .97 10.3% Kaiser Foundation Hosp, West LA 5,939 1.08 10.1% Kaiser Foundation Hosp, Woodland Hills 3,457 .95 9.4% Pacific Presbyterian Hosp, SF 6,860 .85 8.4% Queen of Angels Med Center, LA 3,868 1.56 16.5% Green Hosp, Scripps Clinic, La Jolla* 5,460 .81 .53% Scripps Mem Hospital, La Jolla* 8,997 .90 9.2% Sierra Nevada Mem Hosp, Grass Valley 4,539 1.14 11.3% South Coast Med Cent, S. Laguna* 2,764 1.07 10.3% St. John’s Hosp, Santa Monica 10,713 .90 8.8% St. Mary’s Hosp, San Francisco 7,557 1.01 10.0% St. Vincent Med Cent, LA 8,023 .92 9.7% UCLA Med Cent, LA* 9,023 .86 7.6% UC San Francisco Med Center* 8,696 .93 8.1%
% LOWER THAN HOSPITAL EXPECTED Cedars-Sinai Med Center, LA* 13.3 Century City Hospital, LA 22.2 City of Hope Nat Med Cent, Duarte 23.1 Eisenhower Mem Hosp, Rancho Mirage 16.2 French Hospital, LA 28.1 Hoag Mem Hosp, Newport Beach 10.9 Kaiser Foundation Hosp, Anaheim 17.3 Kaiser Foundation Hosp, Bellflower 11.0 Kaiser Foundation Hospital, LA 22.5 Kaiser Foundation Hosp, Panorama City 16.7 Kaiser Foundation Hosp, San Diego 10.9 Kaiser Foundation Hosp, West LA 21.6 Kaiser Foundation Hosp, Woodland Hills 17.1 Pacific Presbyterian Hosp, SF 17.4 Queen of Angels Med Center, LA 11.7 Green Hosp, Scripps Clinic, La Jolla* 45.7 Scripps Mem Hospital, La Jolla* 13.6 Sierra Nevada Mem Hosp, Grass Valley 16.8 South Coast Med Cent, S. Laguna* 19.2 St. John’s Hosp, Santa Monica 17.9 St. Mary’s Hosp, San Francisco 17.3 St. Vincent Med Cent, LA 11.2 UCLA Med Cent, LA* 25.6 UC San Francisco Med Center* 26.6
NOTES: “Patient risk” is a measure of the relative risk of dying for a hospital’s patients as compared to the national average risk. A value of “1” is average. A value over “1” is above average. A value under “1” is below average.
The “% Higher” or “% Lower” than expected statistic is a comparison of the hospital’s actual and predicted mortality rates. The predicted mortality rate takes into account characteristics of individual patients. These include age, sex, principal diagnosis, other chronic illnesses, previous hospitalizations, emergency or non-emergency admission status, admission source, such as a nursing home or hospital transfer, as well as the number of patients treated at each hospital. If a patient was hospitalized multiple times, only the last admission for each year was included.
Southern California hospitals with more than 1,000 cases over the three-year period are listed, as well as hospitals from other parts of the state with more than 3,000 cases. There is a less than a one-in-1,000 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.
