Analysis Reveals Perinatal Death Rates : Care for Newborns Varies, Studies of Hospitals Show

California’s hospitals vary widely in their ability to provide quality medical care to newborn babies, according to a sophisticated hospital-by-hospital analysis of perinatal death-rate data by researchers at the University of California, Santa Barbara.

In Southern California, 17 hospitals, including County-USC Medical Center and Martin Luther King Jr./Drew Medical Center in Los Angeles County and Chapman General and St. Jude in Orange County, have significantly higher-than-expected adjusted death rates when compared with the statewide average.

Ten others, including Cedars-Sinai Medical Center in Los Angeles, Loma Linda University Medical Center in Loma Linda and Martin Luther in Anaheim, have significantly lower-than-expected adjusted death rates from 1980 to 1984, the period studied. The UC Santa Barbara “Maternal and Child Health Data Base” is considered by many health care experts to be one of the most valid statistical measures of the effectiveness of medical care ever developed. It has been published annually since 1980 and distributed to hospitals, county health departments and major public libraries throughout the state, although it has essentially escaped public notice.

“This is not a one-dimensional index that can be used to rank hospitals,” said Ronald L. Williams, the health policy researcher at UC Santa Barbara’s Community and Organization Research Institute who devised the measure as a graduate student. “But for the most part, lower-than-average adjusted death rates equal better-than-average care.”

Williams and other experts believe that such data, properly interpreted, may be useful to parents in choosing a hospital to deliver their babies and to physicians and hospitals in improving the quality of care.

Differences between hospitals are most likely to be significant for mothers who are at high risk of delivering a sick or premature baby because these are the newborns who may require the greatest medical care.

But physicians caution that the perinatal death rate--deaths around the time of birth and during the first 28 days of life--primarily reflects hospital performance, not other important factors that influence infant survival and well-being, such as prenatal care and inherited birth defects.

“On balance, it is very useful information,” said Dr. Samuel Sapin, the associate medical director of clinical services for Kaiser Foundation Hospitals in Southern California. “It can serve as a flag for hospitals to make sure they are not having problems.”

According to Sapin, the data base helped Kaiser pinpoint problem areas with perinatal care at its Panorama City hospital in the early 1980s, such as delays in some cases in performing Caesarean sections and in putting babies with breathing difficulties on ventilators. The data also spurred Kaiser’s nine hospitals, which deliver more than 25,000 babies each year, to develop cooperative arrangements, including the transfer of high-risk mothers or babies to specialized facilities.

Methodology Rated ‘Valid’

Kaiser considers Williams’ methodology “sufficiently valid” to adopt it as part of the health maintenance organization’s computerized early warning system to monitor the quality of perinatal care, Sapin said.

Among the study’s most striking findings are the high perinatal mortality rates at County-USC Medical Center and Martin Luther King Jr./Drew Medical Center. The two giant county hospitals, which account for more than 15% of all the births in Los Angeles County each year, both have “standardized” death rates of about 20% higher than their predicted values, which are calculated on the basis of the statewide average. The death rate at the other large county hospital, Harbor-UCLA Medical Center, is equal to its predicted value.

In interviews, physicians at both County-USC and Martin Luther King expressed strong criticisms of the study.

“Given the fact that our nurseries are full, that we have patients delivering in the halls, that we don’t have enough monitors, given all those problems we do remarkably well,” said Dr. Richard Paul, chief of obstetrics at County-USC. “I don’t believe (the Williams study) is a fair appraisal because he can’t factor in for our limited resources and our size.”

Dr. Ezra Davidson, the chief of obstetrics at Martin Luther King Jr./Drew Medical Center, termed the study “warped” because the hospital comparisons were made “without factoring in whether the (mother) received prenatal care.” He added, “dollars to improve the availability and accessibility of good prenatal care” are just as important in saving the lives of newborns as improvements in the quality of hospital care.

Williams Responds

In response to these criticisms, Williams maintained that the effectiveness of perinatal medical care may be “less than average” at these hospitals. “It is unlikely that it is just a lack of prenatal care that is responsible,” he said. “Year after year after year, the adjusted perinatal mortality rate at these hospitals is higher than we would expect.”

Also critical of the study was John Kramer, administrator of Chapman General Hospital, which, according to the study, had a mortality rate that was 24.9% above the statewide average.

“You can’t always take the numbers and look at them at face value,” Kramer said, noting that his hospital’s infant mortality rate may have been high in the early ‘80s. But by 1983, the hospital became part of a regional perinatal network run from UCI Medical Center and had a much lower infant mortality rate. As part of the new perinatal network, neonatologists began holding monthly seminars at the hospitals and taking care of Chapman’s high-risk babies.

Fetal Monitor Used

Also, rather than focusing just on high-risk mothers, the hospital now places all mothers on a fetal monitor. “We didn’t always do that,” Kramer said. Does that make a difference in the outcomes of small babies? “You bet it does,” he said.

At St. Jude Hospital and Rehabilitation Center in Fullerton, Assistant Administrator Doreen Dann said she had not reviewed results of the study that showed an infant mortality rate that was 12.1% above the state average. But since 1983, that hospital too has joined the county’s regional perinatal network and increased its fetal monitoring, she noted.

Martin Luther Hospital in Anaheim, however, has one of the lower infant mortality rates in the state--12.1% below the state average, according to the UC Santa Barbara study. Dr. Leonard Fox, director of the hospital’s neonatal intensive care unit, said the unit was designated in 1981 as one of the more sophisticated such units in the county. Fox said the unit was staffed with four specialists in neonatology--”no residents or nurse practitioners”--and that the unit provided one-to-one nursing care for its 15 patients.

The 1987 UC Santa Barbara study is based on a more than 98% complete set of birth and death certificate data for 2.15 million babies born in California between 1980 and 1984 at about 340 hospitals.

Such large numbers are necessary to analyze death rates because the actual number of perinatal deaths at any one hospital is very low. Over the five-year period, there were 24,229 perinatal deaths throughout the state, an average of 11.3 deaths for every thousand births.

The 764-page study, prepared under a contract from the state Department of Health Services by Williams and colleagues Frank P. Rust and Kam J. Rust, is designed to compare the mortality rate for individual hospitals to the statewide average. But it substantially improves upon the raw mortality data by adjusting for factors which influence the perinatal death rate that are beyond the control of the hospitals.

Birth Weight Factor

By far the most important of these factors is the infant’s birth weight, which is recorded on the birth certificate 99.8% of the time, according to the report. For example, despite good medical care, very low birth weight babies--those weighing less than 3.3 pounds--are about 130 times more likely to die around the time of birth than babies of normal weight. Such babies account for about 44% of all perinatal deaths in the state.

The three other factors are the sex and race of the baby and multiple births, such as twins. Together, the four factors account for more than 80% of the variation in the raw mortality rates, Williams said.

Finally, the analysis corrects for the number of babies delivered at each institution. This is necessary because in some cases high or low death rates may have occurred due to chance alone.

Critics contend that this analysis still falls short, because it does not include other potential factors that may significantly influence mortality rates, such as variations in prenatal care, socioeconomic status and the number of babies with fatal birth defects.

But including such factors has limited value, according to Williams. In part these factors are accounted for by correcting for birth weight variations. In addition, they are more difficult to measure and less accurately recorded. “It is fortuitous that birth weight is not only the most important predictor (of perinatal mortality) but it is the most accurately measured,” he said.

Other Findings Summarized

The 1987 UC Santa Barbara study also found:

- Perinatal mortality has fallen in California each year since 1970. From 1980 to 1984, the raw perinatal mortality rate fell from 12.3 to 10.2 per thousand births. About 80% of this improvement resulted from better medical care for low birth weight babies and the rest from a decrease in the number of low birth weight babies.

- Throughout the state, hospitals that deliver fewer than 1,000 babies a year have “standardized” death rates that average more than 7% above what would be expected. By comparison, hospitals which deliver more than 2,000 babies a year have “standardized” death rates that average 3.3% less than predicted values.

- Federal hospitals, county hospitals and for-profit hospitals all have perinatal death rates that average 7% or more above predicted levels. Death rates for home births average 9% above predicted levels. By comparison, University of California hospitals and private nonprofit hospitals have “standardized” death rates that average about 5% less than expected.

- “Standardized” perinatal mortality in Los Angeles, Riverside, Santa Barbara, San Diego and Ventura counties is significantly higher than the predicted levels. But “standardized” mortality is significantly lower than expected in such Northern California counties as Alameda, Sacramento and San Francisco. In Orange County, 1980-84 infant mortality rates were slightly below the statewide average, with a value of 98.2%. The explanation for these regional differences is not clear.

Specialized Nurseries

- Eight of the 10 Southern California hospitals with low “standardized” perinatal death rates have specialized “Level II” or “Level III” intensive-care nurseries for critically ill newborns. But five of the 36 hospitals with high “standardized” deaths rates also have such special nurseries--County-USC Medical Center, Kaiser Hospital, Panorama City, Martin Luther King Jr./Drew Medical Center, Presbyterian Intercommunity Hospital in Whittier and St. Francis Medical Center of Lynwood.

Overall, the state’s 60 hospitals with Level II or Level III nurseries have “standardized” perinatal mortality rates that are about 5% lower than predicted. Those with basic nurseries average mortality rates about 4% higher than predicted.

Another noteworthy finding is the “standardized” mortality statistics for Dr. Milos Klvana. Last month, Los Angeles Municipal Judge James F. Nelson ordered the Valencia physician and his midwife assistant to stand trial on second-degree murder charges based on “a strong suspicion” that their “gross negligence” had contributed to some of the deaths.

The six murder charges against Klvana stem from the deaths of babies he delivered at their mother’s homes or in his medical offices between December, 1982, and September, 1986. In 1984, the report says that Klvana’s personal perinatal mortality rate was 13.5% higher than his expected value, a significant difference.

Blythe Hospital Highest

Palo Verde Hospital in Blythe had the highest “standardized” perinatal mortality rate among all California hospitals, more than 40% above its predicted value, according to the report. “We have one of the highest percentages in the state of mothers who deliver with no prenatal care, including many women from Mexico,” said Dr. John Coddington, a general practitioner on the staff of the hospital.

Downey Community Hospital also had a “standardized” perinatal mortality rate that was more than 30% above its predicted value. A hospital spokeswoman attributed the statistics to a high number of babies who were born with fatal birth defects. After reviewing their data at the request of The Times, the physicians “felt very comfortable that the overall quality of care remained very high,” said Kathy Leaf, the hospital’s risk manager.

Sonora Community Hospital in the Sierra foothills had the lowest “standardized” mortality rate among all California hospitals, more than 40% below its predicted value. In Southern California, Memorial Medical Center in Long Beach had the lowest “standardized” death rate, almost 25% below its predicted value.

Reasons Ascribed

Dr. Houchang D. Modanlou, the hospital’s director of neonatal-perinatal medicine, credited Memorial Medical Center’s low mortality rate to good organization and extensive educational programs for physicians and nurses. “This is the best data available and not just because we rank very low,” he said. “To our knowledge, Williams did a very good job.”

Another hospital with favorable mortality statistics is Cedars-Sinai Medical Center in Los Angeles. Many Cedars-Sinai patients are affluent whites who receive comprehensive prenatal care from private physicians. In contrast, the majority of county hospital patients are poor and Latino or black. They often receive inadequate prenatal care.

Dr. Jeffrey Pomerance, a Cedars-Sinai neonatalogist, was asked how he believes Cedars would have rated if it cared for patients similar to those at County-USC and Martin Luther King Jr. “I think we’d do a better job than the county does but worse than we do with our own patient pool,” Pomerance said. “I think we’re privileged, for example, to have more personnel per baby than (those hospitals) are funded for.”

Times staff writers Lanie Jones and Claire Spiegel contributed to this story.