Cardiac Deaths in Hospitals Starkly Higher Among Children From Low-Income Neighborhoods

Pediatric patients from low-income neighborhoods died at rates 18% higher while hospitalized, and had higher hospital costs after cardiac surgery, compared with those from higher-income neighborhoods, according to results of a national study that the lead investigator called “shocking.”

The NIH-supported study, published Friday in Pediatrics, retrospectively examined the records of more than 86,000 children, mostly infants, with congenital heart disease (CHD).

The size of the neighborhood effect, which persisted even after accounting for race, type of insurance, and hospital, was similar for children of all disease severities.

The study, titled Disparities in Outcomes and Resource Utilization after Hospitalization for Cardiac Surgery by Neighborhood Income, left more questions than answers. The researchers said that more study is needed to understand the causes of these disparities and to test interventions to standardize outcomes.

“The fact that disparities exist in healthcare is nothing new,” said Brett Anderson, MD, MBA, an attending pediatric cardiologist at New York-Presbyterian Morgan Stanley Children’s Hospital and assistant professor of pediatrics at Columbia University Irving Medical Center, in a statement.

“But the fact that we see such a big effect in this population is shocking. We think of this group of children as being particularly well integrated into the healthcare system, regardless of their background.”

Most young patients with congenital heart disease are diagnosed prenatally or as newborns, she said.

“While we expected to see some differences, we assumed the effect would be minor compared to what is seen in general pediatric populations,” said Anderson, the study’s lead investigator. “In fact, the effect was essentially identical to that observed in general pediatric patients.”

CHDs are the most common birth defects managed in the US.

The researchers used data from the Pediatric Health Information System (PHIS) to evaluate post-surgical mortality, length of stay, and standardized hospital costs in 86,104 patients with congenital heart defects at 46 pediatric hospitals between 2005 and 2015. The PHIS is the largest US database of pediatric discharges.

The results were combined with census data on median household income by zip code.

Overall, 2.9% of the patients who had heart surgery died. Length of hospital stay and costs were both 7% higher for patients from the lowest-income neighborhoods compared to those from the highest-income neighborhoods.

The researchers also looked at outcomes in 857,833 children who were hospitalized for other conditions between 2013 and 2015. About half of this group had a chronic condition.

Similarly, patients from the lowest-income neighborhoods had a 22% greater chance of dying in the hospital compared with those from higher-income neighborhoods. Length of stay and in-hospital costs were about 3% higher for patients from the lowest-income neighborhoods.

“When neighborhood disparities have been described in other studies, they have been largely attributed to differences among hospitals or in environmentally-mediated differences in behavioral health,” said Anderson. “In our study, even when the hospital effect was taken into account, neighborhood remained an important predictor of outcomes.”

The study found a higher incidence of severe heart disease in children from low-income neighborhoods, but the neighborhood effect remained after controlling for disease severity.

“Certain environmental factors—such as maternal stress, nutrition, or health expectations—might have contributed to the differences in outcomes that we saw in children from low-income neighborhoods,” said Anderson. “But until we conduct detailed qualitative studies, we can’t be sure why these disparities persist. Ideally, such studies would examine the role of both families and providers—such as how long it takes before a family makes or obtains an appointment with a subspecialist, and whether providers knowingly or unknowingly make different care choices based on a family’s income level or the socioeconomics of the neighborhood from which the child comes.”

The possible reasons cited for neighborhood disparities in previous discussions in other areas of healthcare—differences in the hospitals used or environmentally-mediated differences in behavioral health–seem unlikely in an infant population with preplanned surgeries, the authors noted.

In an interview with The American Journal of Managed Care^®, Anderson said there are 2 possiblilities to explain this study; both would have to be borne out by further research. On the one hand, it is possible that “the patients are fundamentally different,” maybe because of effects from maternal stress or nutrition during the in utero period.

On the other hand, she said providers need to ask themselves, “Are there conscious or unconscious biases that are influencing our care, and might they may be influencing parent-provider dynamics?”

“One real possibility is that patients from different neighborhoods advocate differently for their children,” Anderson said. And in turn, “we as clinicians alter the care that we provide accordingly.”

This does not mean that caregivers should advocate for more care, since care could be adversely altered as such, but it may cause differences in outcomes.

In addition, differences could arise from the time period before a patient gets referred to a tertiary care center. Are there differences in access in prenatal access for these neonates, or healthcare access, or in the timing of diagnosis? For older pediatric patients, what was the timing of the diagnosis, when were they referred for surgery, and what was the preoperative management?

“In this field we think that things are fairly protocolized,” Anderson said.

Efforts like the National Pediatric Cardiology Quality Improvement Collaborative aim to standardize care for the tiniest, most critically ill cardiac patients, and yet these disparities still exist.

“We think we need to think about whether there are subtle differences in what we are doing as clinicians and how we are reacting to the interactions with patients and that we are changing our management accordingly,” she said.

Sometimes doctors may think they are making better choices out of concern for a family, and maybe they are and maybe they aren’t, she noted.

For instance, providers will routinely, if there are concerns about whether or not caregivers have the resources at home to care for a pediatric patient, decide to keep the patient hospitalized a little bit longer. That drives up resource utilization, and if you are judging a hospital based on the length of stay, they may actually be acting in the patient’s best interest, Anderson noted.

However, it doesn’t explain the mortality differences, and this study will likely prompt some introspection.

“What clinicians can do while they are waiting for the results of those studies is to look at their own management and say, is there anything that I am doing differently, is there any way that this could be affecting my patients?” Anderson asked.” “Or, is there any way that I could be helping my patients?”

Reference

Anderson BR, Fieldston ES, Newburger JW, et al. Disparities in outcomes and resource utilization after hospitalization for cardiac surgery by neighborhood income. Pediatrics. 2018;141(3):e20172432. doi: 10.1542/ peds.