Nearly 1 in 5 Pediatric Hospital Deaths in the U.S. Involve Sepsis

CHICAGO — More than 18,000 non-neonatal sepsis cases occur annually in U.S. kids, and nearly one in five pediatric in-hospital deaths involve sepsis, researchers estimated based on findings of a retrospective cohort study.

Among nearly 4 million U.S. pediatric hospitalizations from 2016-2023 analyzed, over 51,000 cases of sepsis were identified (average incidence 1.3%), reported Chanu Rhee, MD, MPH, of Harvard Medical School in Boston, and Scott Weiss, MD, of Nemours Children’s Hospital in Wilmington, Delaware.

In-hospital mortality for the pediatric sepsis cases reached 10.1%, and sepsis was involved in 17.8% of the hospitalizations that resulted in death, according to findings presented at the Society of Critical Care Medicine annual meeting in Chicago and simultaneously published in JAMA.

The researchers derived their estimates after developing the Pediatric Sepsis Event (PSE) definition — which was adapted from the 2024 Phoenix criteria and requires presumed infection with concurrent organ dysfunction — as a CDC framework for tracking sepsis in children. The tool can be used with routine electronic health care (EHR)-captured surveillance data.

“These findings underscore a very high burden of pediatric sepsis and establish the PSE as a scalable framework that can be used to advance prevention, quality improvement, and policy initiatives,” Weiss told attendees.

On record review, the PSE had a 69.9% sensitivity (95% CI 58.1-79.8) and 93.1% specificity (95% CI 89.6-95.7) for identifying sepsis based on the EHR data.

“It’s good for the field that we have a reliable definition of sepsis that can be used across multiple different EHR platforms,” Nadir Yehya, MD, of the Children’s Hospital of Philadelphia, told MedPage Today.

Yehya, who was not involved in the research, said he was impressed with how thorough the study was and how exhaustive the datasets were in making the first steps toward reliable mortality estimates for pediatric sepsis. “That is a real advance in critical care.”

The findings represent the first national estimates for pediatric sepsis, and the PSE definition addresses an unmet need, noted authors of an editorial that accompanied the JAMA publication.

“While many countries maintain public surveillance systems for common infectious diseases such as influenza, the absence of systematic sepsis surveillance obscures the full societal burden of disease, limits benchmarking of care, and hinders effective public health measures,” wrote Luregn Schlapbach, MD, PhD, of University Children’s Hospital Zurich, and colleagues.

The PSE definition “sets a pragmatic standard to facilitate [EHR]-based surveillance of pediatric sepsis in the U.S.,” they added. “One of the leading causes of death as well as short- and long-term morbidity in children can thereby be robustly monitored for public health. This should spur advances in quality improvement and research both in the U.S. and internationally to reduce the excessive burden of sepsis on child health.”

For the epidemiological estimates, Weiss, Rhee, and colleagues used Epic Cosmos records from 245 health systems across 50 states from 2016-2023 and HCA Healthcare records from 146 hospitals across 20 states from 2018-2023 on children ages 30 days or older admitted as inpatients or who died in the emergency department.

They identified 3,926,809 pediatric hospitalizations and 51,542 sepsis encounters, including 61.6% with shock. Community-onset sepsis comprised 72.6% of cases. In-hospital mortality was 4% for cases without septic shock and 14% for septic shock cases.

Incidence of sepsis was highest among infants, at 1.8%, and decreased among older age groups (1% among those ages 12-17 years). Cases were more common among boys (1.4% vs 1.2% in girls) but mortality was similar.

Mortality was highest in infants (11.1%) and adolescents (11.7%), and higher in hospital-onset (11.4%) versus community-onset (9.6%) cases.

The most common organ dysfunction was respiratory requiring invasive ventilation, which occurred in 67.4% of cases, followed by hypotension (36.9%) and neurologic dysfunction (31%). Most cases were admitted to the intensive care unit (79.2%), with a median 13 days of hospital stay.

With the Epic Cosmos and HCA Healthcare datasets capturing 44.5% of pediatric hospitalizations in 2022, the national estimates for that year were 18,231 sepsis cases and 1,877 sepsis deaths.

The PSE definition was validated with physician medical record review of 581 encounters at three geographically diverse hospitals. The positive predictive value was 70.7%, and the negative predictive value was 92.9%. Most false negatives were due to short antimicrobial courses, often with viral infections that were rarely fatal, and most false positives resulted from cases of organ dysfunction deemed unrelated to the infection.

Clinical characteristics remained consistent in secondary datasets used to assess the feasibility and validity of using the PSE definition across heterogenous settings, said Weiss, supporting the generalizability of the PSE framework across diverse settings.

The study was limited by the exclusion of neonates, for whom Weiss suggested would require distinct surveillance criteria. Other limitations included that validation was only in academic children’s hospitals, though the researchers said they expect the findings to be generalizable to community hospitals as well.

Please enable JavaScript to view the comments powered by Disqus.