An analysis of labor and delivery data in the state of New York has revealed that hospital—but not neighborhood—characteristics are associated with severe maternal morbidity (SMM). The study identified several factors that predict SMM.
Lead study author Jean Guglielminotti, MD, PhD, an anesthesiologist at Hôpital Bichat-Claude Bernard, in Paris, and a postdoctoral research fellow at Columbia University, in New York City, said the study indicates that all pregnant women should be assessed with a comorbidity index tool, and that high-risk women should be referred to hospitals with levels of care adapted to their condition. He presented the study at the 2016 annual meeting of the Society for Obstetric Anesthesia and Perinatology.
Previous research suggests that neighborhood characteristics, such as long driving time to hospitals or low obstetrician density, are associated with adverse maternal outcomes (Am J Obstet Gynecol 2005;193:1083-1088). Similar associations have been suggested for hospital characteristics, such as rural location, low delivery volume, high proportion of minority patients or low staffing (Am J Obstet Gynecol 2014;211:647.e1-16).
Neighborhood, Hospital And Personal Factors
In the new study, Dr. Guglielminotti and his colleagues examined the association between neighborhood- and hospital-level factors and SMM to see whether they could identify potential avenues for improving maternal safety. They analyzed discharge records indicating labor and delivery in the State Inpatient Database for New York between 2009 and 2011.
In the analysis of 605,534 discharges from 139 hospitals, 1.26% indicated SMM, up from 1.1% in 2009 (P<0.001). The three most frequent SMMs were severe postpartum hemorrhage (35%), disseminated intravascular coagulation (27%) and heart failure (16%; Table). Similar to previous research, patient-level factors also were associated with SMM, including ethnic minority, pregnancy results from assisted reproductive technology, Bateman Risk Index of at least 1, induction of labor and cesarean delivery.
Contrary to what they expected, neighborhood-level factors did not appear to affect maternal morbidity. Driving time to hospital (per minute increase) and obstetrician density in the county of residence were not associated with maternal morbidity. Dr. Guglielminotti said one possible explanation for this is that, compared with the rest of the United States, New York has an increased density of active physicians (394 vs. 273/100,000), and more patients live within a 30-minute drive to the nearest hospital (94% vs. 87%).
The researchers also found that hospitals with higher cesarean delivery rates have better outcomes (cesarean delivery rate per 1% increase, adjusted odds ratio, 0.978; P<0.001). “This was a bit unexpected. We were thinking that hospitals with low cesarean delivery rates would have better outcomes,” Dr. Guglielminotti said. The findings, he said, suggest that cesarean delivery rates at the hospital level are a marker for unmeasured factors, including hospital practice patterns, protocol-driven care, and individual physician decision-making process and skills. He said a better understanding of what actually happens in hospitals with low cesarean delivery rates is needed.
Hospitals that had a high proportion of high-risk pregnancies had worse outcomes (percentage of deliveries of cases with Bateman Risk Index >1 [per 1% increase], 1.051; P<0.001). “We had expected better outcomes in hospitals that routinely manage high-risk pregnancies,” Dr. Guglielminotti said. “This strongly suggests that women with high-risk pregnancies and multiple morbidities are not seeking care at hospitals with appropriate levels of care adapted to their conditions.”
Dr. Guglielminotti said there were two main implications for clinical practice. “All women should be assessed for morbidity with a simple comorbidity index. We suggest that the Bateman Risk Index could be used,” he said. “The second recommendation is that women at high risk should be referred to hospitals with appropriate levels of maternal care to their condition, probably not rural or minority-serving hospitals.” The researchers plan to validate the findings in other states with lower physician and hospital density.
Minority-Serving Hospitals Avoidable?
After the presentation, Ivan Velickovic, MD, director of OB Anesthesia at SUNY Downstate, in New York City, said his hospital serves minority patients, many of whom are high risk. “Your second conclusion that we should refer patients to the nonminority hospitals is just impossible,” Dr. Velickovic said. “They would deliver on the Brooklyn Bridge. You can’t physically do that in New York City.”
Barbara Scavone, MD, professor of anesthesia and critical care at Pritzker School of Medicine at the University of Chicago, whose hospital is located on the South Side of the city, which is a poorer part of town, told Anesthesiology News she thought one of the difficulties in looking at these disparities at the facility level is that the facility may sometimes be a marker for other things. “In my own city of Chicago, I am going to posit that someone who leaves my neighborhood area to go up to the North Side for care is probably a different patient than one who comes to my hospital, where 30% are walk-ins with either no or very spotty prenatal care,” Dr. Scavone said. “When you compare the outcomes of those two facilities, I’m sure the outcomes are very, very different, but I’m not sure it is necessarily because the facility is a poorly performing facility.”