Hospital Death Rates for Five Surgeries Vary

Death rates associated with heart bypass operations, gallbladder removals and three other common surgeries varied tremendously among California hospitals in 1985, according to a Times analysis of data covering nearly all patients hospitalized in the state.

At seven hospitals, more than 1 in 10 bypass surgery patients died after the operation, compared to the statewide average of fewer than one in 20. These hospitals included Hollywood-Presbyterian Medical Center, Daniel Freeman Memorial Hospital in Inglewood and UC Medical Center San Francisco. Five others, including two in San Diego, had significantly low bypass surgery death rates.

For gallbladder surgery, six hospitals had death rates more than four times the state average death rate of 1.6%. By comparison, 22 hospitals that performed large numbers of the operation, including County-USC Medical Center and three Kaiser Foundation hospitals, had death rates of less than 0.6%.

These are among the conclusions from a computerized hospital-by-hospital analysis of surgical death rate data by the Los Angeles Times Poll in collaboration with expert consultants from the Institute for Health Policy Studies at UC San Francisco.

Guide to Quality of Care

“These kinds of data are useful markers for the quality of care provided in hospitals,” said Jonathan A. Showstack, one of the three UCSF experts retained by The Times. “Hospitals should be expected to provide additional information to explain their high mortality rates.”

The data may help patients discuss with their physicians where to undergo specific surgical procedures. The federal Medicare program and private insurance carriers have said they are studying the use of such statistics to help identify the hospitals that will qualify for reimbursement for certain complex procedures, such as heart bypass surgery.

Some hospitals with above-average death rates vehemently denied that their statistics pointed to potential deficiencies in their quality of care. But some facilities acknowledged that they had been motivated to improve their performance.

“We are not happy with these numbers, looking at them, either,” said Sandi Chester, a Hollywood-Presbyterian vice president, of the hospital’s 14.6% death rate for heart-bypass surgery in 1985, more than three times the state average of 4.6%. Chester said the 395-bed facility is in the process of substantially retooling its cardiology and cardiac surgery programs.

Some Lessons Seen

“There is clearly in this some self-education,” said Dr. George Stefanik, a cardiac surgeon at Daniel Freeman. Hospital statistics show that the bypass surgery mortality rate decreased from 12.2% in 1985 and 13.0% in 1986 to 5.6% in 1987.

Bypass surgery death rates also decreased by about 50% between 1985 and 1987 at UCLA Medical Center and Pacific Presbyterian Medical Center in San Francisco--two hospitals that have heart transplant programs and reputations for operating on high-risk patients turned down by surgeons at other hospitals.

The Times analysis covers about 106,000 patients who underwent five common surgeries--heart bypass, gallbladder removal, prostate resection, total hip replacement and carotid endarterectomy--in 1985, the most recent year for which complete information is available.

It is based on computer tapes compiled by the Office of Statewide Health Planning and Development. The data is supplied and verified by individual hospitals and is “among the most accurate data of this type,” Showstack said.

The Times compared the mortality rates for individual hospitals to the statewide average, after correcting for variations between hospitals in the age, sex and race of patients. The death rates indicate the percentage of patients receiving a particular surgery who die in the hospital out of all such patients who are admitted.

This analysis attempts to improve upon other widely publicized hospital death-rate disclosures in recent years by such groups as the federal Health Care Financing Administration and California Medical Review. This is because it includes all adult patients, not just the elderly and disabled, and because it focuses on specific procedures, not broad heterogenous categories of illnesses.

For heart bypass and gallbladder surgeries, the analysis also corrected for some of the factors that may cause some patients to be sicker than others before an operation, such as the percentage of patients with acute heart attacks or the percentage with cancer.

Despite such improvements, the UCSF consultants and hospital officials caution that other factors are also important in judging the quality of surgical care. Among these are more complete data about the pre-operative condition of patients and post-operative complications, such as that contained in confidential medical records. A high death rate at a hospital does not necessarily indicate bad performance, nor a low death rate good performance.

Among The Times’ most significant findings are the high death rates at 11 California hospitals for heart bypass surgery, which was associated with about 800 deaths in 1985.

In this expensive and technically complex procedure, surgeons restore an adequate blood supply to heart tissue by grafting blood vessels from elsewhere in the body around blockages in the arteries supplying blood to the heart. During the surgery, the heartbeat is stopped and the patient is kept alive on a “heart-lung machine.”

The average patient undergoing bypass surgery in recent years is much sicker and older than patients operated on a decade earlier. This is due to improved medical treatments for heart disease and the increasing use of angioplasty to treat patients who previously would have required surgery. Angioplasty is a procedure performed by cardiologists in which clogged heart arteries are opened with a balloon-tipped hollow tube inserted through the skin and threaded through a vein to the heart.

Even though most surgeons have improved their skills in caring for sicker patients, some bypass patients have expected mortality rates of between 10% to 20% or more. In addition, patients who require emergency surgery because of angioplasty complications, such as ruptured heart arteries or worsened obstructions, are at particularly high risk of dying.

On the other hand, a particularly gifted surgical team, with good backup from other physicians and nurses, can save many more high-risk patients than colleagues with less talent and less organizational support.

A high percentage of very sick patients and a small number of elective cases appear to be important factors in the relatively high 9.0% bypass surgery death rate at the UCLA Medical Center in 1985.

About one-third of the 145 bypass surgery patients that year were referred from institutions that have their own cardiac surgery programs, including six of the 13 patients who died, according to Dr. Hillel Laks, UCLA’S chief cardiac surgeon. Dr. Donald Hill, the chief heart surgeon at San Francisco’s Pacific Presbyterian Medical Center, has a similar referral pattern.

The 13 UCLA patients who died included four who were already in acute circulatory collapse, or “cardiogenic shock,” after heart attacks. Others who died included a patient in shock after a failed angioplasty, a Jehovah’s Witness who died in need of blood transfusions that she had earlier rejected and a patient having repeat bypass surgery who had been turned down by the surgeons who performed the first operation, Laks said. In 1987, three of the seven patients who died were in cardiogenic shock and two had failed angioplasties.

Cardiogenic Shock

“The mortality rate at UCLA for patients in cardiogenic shock is 30%,” Laks said. “Without surgery the mortality is 90%. At the few other institutions that are willing to operate on patients in cardiogenic shock, the mortality is reported in the medical literature as being 50% or higher.”

Laks said that some patients referred to him “have been given an inflated mortality figure, so they will either not have the surgery or go elsewhere.” He speculated that some surgeons “may be less willing to take on high-risk patients even when surgery is indicated because of concern about the publication of their statistics.”

UC San Francisco had the highest mortality rate for heart bypass cases among all California hospitals in 1985, 14.7% or 24 deaths following 163 surgeries.

For 1986 and the first nine months of 1987, surgeons performed 281 such surgeries at UCSF with 29 deaths, for a 10.3% mortality rate, according to hospital statistics. By comparison, the hospital’s death rate in 1983 was 3.9%.

Dr. Stan M. Lindenfeld, chairman of UCSF’s quality assurance executive committee, said that a detailed review of all bypass deaths for 1985 through 1987 had revealed no quality-of-care deficiencies.

“We feel very comfortable with the statistics,” Lindenfeld said in a telephone interview. “We clearly have a (patient) population--because of referral patterns--who are very, very ill when they start.”

Other Factors Blamed

But several knowledgeable sources, who requested anonymity, said high-risk patients were only partially to blame for UCSF’s high mortality rates.

In particular, they pointed out that the division of cardiovascular surgery has been without a chief since June, 1986, when Dr. Paul Ebert, a world-renowned pediatric heart surgeon, left to become the director of the Chicago-based American College of Surgeons. Other university heart surgeons have also left or restricted their practices for a variety of reasons.

“It is a temporary crisis, but nobody here is going to deny it,” one widely respected UCSF physician said. “We don’t have as good a bypass surgery program as we had five years ago.”

Kaiser Foundation Hospital, San Francisco, had the lowest bypass surgery mortality rate among all California hospitals, 1.6%. Between 1983 and 1987, surgeons performed 1,875 bypasses with 32 deaths, for an overall 1.7% mortality rate, according to Kaiser statistics.

“A fairly mediocre cardiovascular surgeon can probably do bypass surgery and have a 90% survival rate,” said Dr. Terry McEnany, chief of cardiovascular surgery at the hospital. “But if you want to cut down to 1% to 2% to 3%, you need to have an entire approach to coronary surgery.”

McEnany said the key elements of the hospital’s approach were “surgical skill” and a system where nurses, anesthesiologists and other medical professionals involved in caring for heart surgery patients spend most of their time doing so.

In addition, a staff cardiac surgeon or a general surgeon in training to be a cardiac surgeon remains in the hospital 24 hours a day to care for post-operative patients in the intensive-care unit.

In most hospitals where open-heart operations are performed, such surgeons are not as available to coordinate post-operative cardiac care. “That probably pays off in three or four (lives-saved) every year,” McEnany said. “We can intervene quickly to prevent catastrophe.”

Surgical Exception

Kaiser heart patients are “as sick as the patients being operated upon” in most other California hospitals, McEnany added. The major exception is that surgeons do not have to treat angioplasty complications because Kaiser patients needing this procedure are treated at other hospitals.

Throughout the state, The Times also found that the 34 hospitals that had more than 200 bypass surgeries in 1985 had significantly lower death rates than the 44 that had fewer surgeries.

The average death rate varied from 3.94% for hospitals with more than 200 bypass surgeries, to 6.05% for facilities with between 100 and 200 surgeries, to 6.94% for those with fewer than 100 surgeries.

Four of the 11 hospitals with high death rates--Hollywood Presbyterian, Granada Hills Community Hospital, Mt. Zion Medical Center in San Francisco and Santa Clara Valley Medical Center in San Jose--had 68 or fewer surgeries. Five others had between 100 and 200 surgeries.

Because of this generally accepted relationship between high volume and low mortality, medical experts, such as The Times’ consultants, say that a significant number of deaths might be prevented by concentrating heart surgery in a smaller number of busier hospitals.

Hollywood Presbyterian, one of the low-volume hospitals, averaged about 40 bypass surgeries a year between 1983 and 1987, with an overall 10.7% death rate, according to hospital statistics.

Unlike most busy hospitals performing bypass surgery, Hollywood Presbyterian has neither a full-time heart surgeon nor a full-time surgical team. Hospital officials said that patients requiring heart surgery sometimes request to be sent to institutions with more comprehensive programs.

Improved Procedures

Chester, the vice president for planning and quality management, acknowledged the deficiencies, but said the program will continue while better cardiology and cardiac surgery services are developed.

If the program were suspended, it would be “much more difficult . . . to start up again,” Chester said in a telephone interview. She added: “If the numbers (don’t) improve, we would have to get out of the business of cardiothoracic surgery.” Dr. Roger Nahas, the chief surgeon for most of the recent cases, declined to be interviewed.

Daniel Freeman, which also had a high bypass mortality rate in 1985, said its heart surgery program is now back on track. Dr. G. Arnold Mulder, a heart surgeon at the hospital, attributed much of the improvement to a declining rate of angioplasty complications and improved pre-operative care for such patients, triggered in part by extensive discussion of the hospital’s mortality statistics.

In addition to high bypass surgery death rates, some hospitals had an unusually large number of survivors with prolonged hospitalizations after their surgery, The Times found.

A high percentage of patients hospitalized for 15 days or more after surgery suggests “an unusually high rate of complications,” such as irregular heart rhythms or severe bleeding, according to a paper by Showstack and other UCSF researchers published in the Journal of the American Medical Assn. last year.

In 1985, the median length of hospitalization after bypass surgery in California was eight days and only 9.2% of patients remained in the hospital for 15 days or more. By comparison, at Daniel Freeman, 25.3% of the survivors had such prolonged hospitalizations. The comparable figure was 22.5% at Hollywood Presbyterian and 17.9% at UCSF.

Other hospitals among the 10 with the greatest percentage of prolonged hospitalizations included Los Angeles County-USC Medical Center (41.1%); Lancaster Community Hospital (26.7%); Harbor-UCLA Medical Center (25.0%), and Cedars-Sinai Medical Center (21.1%). The raw mortality rates of these hospitals were 8.2%, 5.1%, 6.7%, and 3.2%, respectively.

Other Findings

The Times analysis also found:

- Death rates for gallbladder removal ranged from more than 7% at Ojai Valley Community Hospital, Humana Hospital, Huntington Beach and Robert F. Kennedy Medical Center in Hawthorne to 0.60% or less at 22 hospitals where large numbers of the surgeries were performed. The hospitals with low mortality rates included County-USC Medical Center (537 cases), Kaiser Foundation Hospital, Fontana (504 cases), and Kaiser Foundation Hospital, Bellflower (338 cases). In total, 39,292 gallbladder removals were performed, with 617 deaths, or a 1.57% mortality rate.

- More than 150 transurethral prostate surgeries were performed with no deaths at each of 17 hospitals. These hospitals included Cedars-Sinai Medical Center (624 surgeries), Mercy Hospital and Medical Center of San Diego (344 surgeries), and Memorial Medical Center of Long Beach (226 cases). The operation relieves a blockage in the flow of urine in men due to overgrowth of the prostate gland. In total, 29,023 transurethral prostate resections were performed with 136 deaths, or a 0.47% mortality rate.

- More than 150 total hip replacements were performed with no fatalities at each of six hospitals, including Green Hospital of Scripps Clinic in La Jolla (282 cases), Centinela Hospital Medical Center in Inglewood (153 cases), Santa Barbara Cottage Hospital (157 cases) and UCLA Medical Center (248 cases). In total, 10,594 total hip replacements were performed with 94 deaths, or a mortality rate of 0.89%.

Joint Is Replaced

In this operation, the hip joint is replaced with an artificial ball and socket joint. The surgery is used to treat serious cases of arthritis, improperly healed fractures or destruction of the hip joint from other causes.

Two hospitals with high mortality rates--Brotman Medical Center in Culver City and St. Jude Hospital and Rehabilitation Center in Fullerton--reviewed their deaths at the request of The Times. Both maintained that high-quality care was provided in all cases.

- Five hospitals had high death rates for carotid endarterectomy, in which fatty plaque is removed from arteries in the neck that supply blood to the brain. In total, 10,726 of these surgeries were performed, with 210 deaths, or a 1.96% mortality rate.

Detlef D. Matthies of the Los Angeles Times Poll did the computer programming for this story. Times researcher Tracy Thomas assisted in preparing the story.

ANALYZING HOSPITAL DEATH RATES

The Los Angeles Times Poll, directed by I. A. Lewis, compiled lists of hospitals with high and low in-hospital surgical death rates through a computerized statistical analysis called “multiple regression.”

The analysis compared mortality rates for individual hospitals to a statewide average for each operation. It corrected for variations between hospitals by age, sex and race of patients. There is a less than 5% probability that a hospital would be listed as having a high or low mortality rate because of chance variations in the death rate.

Some hospitals with high or low mortality rates based on raw data are not listed because after age, sex and race were taken into account, their mortality rates were no longer statistically high or low.

State data is not complete enough to make a comprehensive correction for variations in the sickness of patients undergoing surgery. But similar results were obtained for gall bladder and heart bypass surgery when some of the factors that influence the pre-operative sickness of patients were taken into account.

These factors included the percentage of patients with cancer, the percentage with acute heart attacks and the percentage of “non-scheduled” surgeries, which may be associated with higher death rates than scheduled cases.

The raw data used in the analysis was edited and verified by individual hospitals before being supplied to the state. The raw data was reconfirmed with about 35 facilities and generally found to be very accurate.

Three experts from the Institute for Health Policy Studies at UC San Francisco were hired by The Times for assistance in analyzing the data and interpreting the results. They are: Jonathan A. Showstack, associate professor of health economics; Harold S. Luft, professor of health policy, and Deborah W. Garnick, senior research associate.

SURGICAL MORTALITY RATES IN CALIFORNIA HOSPITALS These California hospitals have statistically significant high or low death rates for common surgeries, according to an analysis by the Times Poll of computer data tapes compiled by the Office of Statewide Health Planning and Development. The computer tapes contain non-confidential case-by-case data on 2.26 million hospital patients in 1985, the most recent year for which they are available. Patients under 18 and hospitals with fewer than 40 patients for a surgery were excluded. There is a less than 5% probability that a hospital would be listed as having a high or low mortality rate because of chance.

HEART BYPASS SURGERY *

Total Deaths Hospital Patients By % STATEWIDE 16,846 4.60% High Mortality Rates UC Medical Center, San Francisco 163 14.72% Hollywood Presbyterian Medical Center 55 14.55% Mount Zion Medical Center, San Francisco 68 13.24% Santa Clara Valley Medical Center, San Jose 46 13.04% Daniel Freeman Memorial Hospital, Inglewood 181 12.15% Granada Hills Community Hospital 64 10.94% Pacific Presbyterian Medical Center, San Francisco 156 10.89% UCLA Medical Center, Westwood 145 8.96% Northridge Hospital Medical Center 132 8.33% Samuel Merritt Hospital, Oakland 279 7.89% Seton Medical Center, Daly City 425 7.06% Low Mortality Rates Kaiser Foundation Hospital, San Francisco 449 1.56% Sequoia Hospital, Redwood City 294 2.04% Sutter Memorial Hospital, Sacramento 641 2.65% Sharp Memorial Hospital, San Diego 661 2.87% Alvarado Community Hospital, San Diego 411 2.92%

These hospitals had a low mortality rate but with a higher--5% to 10%--probability that the low rates resulted from chance.

Total Deaths Hospital Patients By % Kaiser Foundation Hospital, Los Angeles 476 2.31% Scripps Memorial Hospital, La Jolla 347 2.59% Cedars-Sinai Medical Center, Los Angeles 411 3.16%

* Statistics for heart bypass surgery exclude patients who also underwent heart valve replacement. GALL BLADDER REMOVAL

Total Deaths Hospital Patients By % STATEWIDE 39,292 1.57% High Mortality Rates Ojai Valley Community Hospital 41 9.76% Humana Hospital, Huntington Beach 52 9.62% Robert F. Kennedy Medical Center, Hawthorne 55 7.27% St. Helena Hospital and Health Center, Deer Park 42 7.14% Queen of Angels Medical Center, Los Angeles 87 6.90% Doctors Hospital of Pinole 61 6.56% Western Medical Center, Santa Ana 137 5.83% Scripps Memorial Hospital, Encinitas 76 5.26% Pomona Valley Community Hospital 222 4.95% Bellwood General Hospital, Bellflower 61 4.92% White Memorial Medical Center, Los Angeles 103 4.85% UC Medical Center, San Diego 125 4.80% Rideout Memorial Hospital, Marysville 85 4.71% UCLA Medical Center, Westwood 215 4.65% Loma Linda University Medical Center 195 4.62% Santa Ana Hospital Medical Center 44 4.54% Desert Hospital, Palm Springs 112 4.46% Samuel Merritt Hospital, Oakland 94 4.26% Roseville Community Hospital 125 4.00% Scenic General Hospital, Modesto 76 3.95% San Bernardino County Medical Center 105 3.80% Kern Medical Center, Bakersfield 132 3.79% Sharp Memorial Hospital, San Diego 242 3.31%

22 hospitals that performed more than 150 surgeries had mortality rates of less than 0.60%, including these Southern California hospitals. The differences from the state average were not statistically significant.

Total Deaths Hospital Patients By % Downey Community Hospital 178 0.56% Fountain Valley Regional Medical Center 188 0.53% Granada Hills Community Hospital 178 0.56% Harbor-UCLA Medical Center, Torrance 212 0.00% Kaiser Foundation Hospital, Bellflower 338 0.59% Kaiser Foundation Hospital, Fontana 504 0.20% Kaiser Foundation Hospital, San Diego 262 0.00% Little Company of Mary Hospital, Torrance 193 0.00% Los Angeles County-USC Medical Center 537 0.56% Queen of the Valley Hospital, West Covina 195 0.00% Santa Barbara Cottage Hospital 222 0.00% St. Jude Hospital and Rehab Center, Fullerton 181 0.55% Whittier Hospital Medical Center 192 0.52%

TRANSURETHRAL PROSTATE RESECTION

Total Deaths Hospital Patients By % STATEWIDE 29,023 0.47%

10 hospitals had statistically significant high mortality rates, including these hospitals with three or more deaths.

Total Deaths Hospital Patients By % San Bernardino Community Hospital 75 6.67% St. Francis Medical Center, Lynwood 104 4.81% Mercy Hospital, Merced 148 2.02%

17 hospitals that performed more than 150 surgeries had no deaths, including these Southern California hospitals. The differences from the state average were not statistically significant.

Total Deaths Hospital Patients By % Cedars-Sinai Medical Center, Los Angeles 624 0.00% Eisenhower Medical Center, Rancho Mirage 175 0.00% Huntington Memorial Hospital, Pasadena 194 0.00% Kaiser Foundation Hospital, Los Angeles 197 0.00% Little Company of Mary Hospital, Torrance 150 0.00% Memorial Medical Center of Long Beach 226 0.00% Mercy Hospital and Medical Center, San Diego 344 0.00% Saddleback Community Hospital, Laguna Hills 156 0.00% San Pedro Peninsula Hospital 215 0.00%

CAROTID ENDARTERECTOMY

Total Deaths Hospital Patients By % STATEWIDE 10,726 1.96% High Death Rates Desert Hospital, Palm Springs 67 8.96% Henry Mayo Newhall Memorial Hospital, Valencia 42 7.14% Sharp Memorial Hospital, San Diego 119 5.88% Anaheim Memorial Hospital 54 5.56% Mercy San Juan Hospital, Carmichael 120 5.00%

One hospital that performed more than 125 surgeries had no deaths. The difference from the state average was not statistically significant.

Total Deaths Hospital Patients By % Loma Linda University Medical Center 133 0.00%

TOTAL HIP REPLACEMENT

Total Deaths Hospital Patients By % STATEWIDE 10,594 0.89%

Five hospitals had statistically significant high mortality rates, including these hospitals with more than two deaths.

Total Deaths Hospital Patients By % Brotman Medical Center, Culver City 44 6.82% St. Jude Hospital and Rehab Center, Fullerton 53 5.66%

Six hospitals that performed more than 150 surgeries had no deaths. The differences from the state average were not statistically significant.

Total Deaths Hospital Patients By % Green Hospital of Scripps Clinic, La Jolla 282 0.0% Centinela Hospital Medical Center, Inglewood 153 0.0% Santa Barbara Cottage Hospital 157 0.0% Stanford University Hospital 165 0.0% UCLA Medical Center, Westwood 248 0.0% UC Medical Center, San Francisco 215 0.0%

FIVE YEARS OF HEART BYPASS DEATH RATES

The Times obtained heart bypass mortality statistics covering five years for five California hospitals with high death rates in 1985 and one hospital with a low death rate.

1983 Total Deaths Hospital Patients By % Daniel Freeman Memorial Hospital 270 4.4% Hollywood Presbyterian Medical Center 59 8.5% Pacific Presbyterian Med Center, San Francisco 216 6.5% UC Med Center, San Francisco 152 3.9% UCLA Medical Center 179 4.5% Kaiser Foundation Hospital, San Francisco 296 1.7%

1984 1985 1986 1987* Total Deaths Total Deaths Total Deaths Total Deaths Hospital Patients By % Patients By % Patients By % Patients By % Daniel Freeman Memorial Hospital 186 3.2% 181 12.2% 143 13.0% 160 5.6% Hollywood Presbyterian Medical Center 16 12.5% 55 14.5% 44 11.4% 31 6.5% Pacific Presbyterian Med Center, San Francisco 176 10.8% 156 10.9% 131 6.1% 132 6.1% UC Med Center, San Francisco 160 6.9% 163 14.7% 179 9.5% 102 11.8% UCLA Medical Center 152 5.3% 145 9.0% 172 8.7% 167 4.2% Kaiser Foundation Hospital, San Francisco 327 3.1% 449 1.6% 371 1.1% 432 1.4%

* 1987 statistics for UCSF are for the first nine months only.

Statistics exclude patients who also underwent heart valve replacement.

Sources: Office of Statewide Health Planning and Development, hospital statistics