Study: Nearly 1 in 5 pediatric hospital deaths involve sepsis

Sepsis is a factor in more than 10% of non-neonatal pediatric hospitalizations as well as almost 18% of those patients’ in-hospital deaths, according to a recently published estimate built on electronic health record data. 

Those rates translate to over 18,000 estimated nationwide hospitalizations and 1,800 deaths in a hospital in 2022, according to the study. 

Sepsis incidence and mortality show no statistically significant changes from 2016 to 2022. 

Children with sepsis also had substantially longer average hospital stays, very often received care in the ICU and were frequently discharged to post-acute care.

Researchers behind the study, as well as physicians commenting on the findings in an accompanying editorial, describe the tally as the first national estimates of non-neonatal pediatric sepsis built on a standardized, scalable framework. Compared with prior estimates, the updated approach indicates a lower national incidence of pediatric sepsis but a higher rate of mortality.

“These findings underscore the substantial burden of pediatric sepsis and its major contribution to childhood morbidity and mortality,” researchers wrote of the estimates in JAMA. 

The analysis included nearly 4 million admissions of children aged one month to 17 years between 2018 and 2023. Researchers collected these data from Epic Cosmos, which includes deidentified electronic health records from 245 health systems, and from HCA Healthcare’s 146 hospitals.

Over 51,000 of these admissions, or 1.3%, involved sepsis as defined by criteria the researchers adapted from standards laid out in 2024 by a multispecialty task force—an effort to update prior sepsis criteria established in 2005. 

Whereas all hospitalized children had an average length of stay of four days, those with sepsis averaged 24.2 days. The 10.1% in-hospital mortality of pediatric sepsis cases was substantially higher than the 0.7% rate of hospitalized children without sepsis. 

Hospital-onset sepsis represented just over a quarter of cases, but had a higher rate of mortality than community-onset cases (11.4% versus 9.6%). Both incidence and mortality were highest among infants. Septic shock was identified in just over three-fifths of the cases, and those with septic shock had higher rates of in-hospital mortality (14% versus 4%). 

A record review showed the new definition had 69.9% sensitivity and 93.1% specificity when validated in a record review. “Although imperfect,” the researchers said that sensitivity was an improvement over prior sepsis estimate studies that rely on administrative diagnostic codes. 

Such approaches “have been shown to be variably applied, resulting in a number of limitations and biases,” noted physicians who authored the accompanying editorial. The new analysis and its estimates still have limitations, authors of both works acknowledged, but are a clear step forward in surveillance for a severe health threat.

“[The new approach] sets a pragmatic standard to facilitate electronic health record–based surveillance of pediatric sepsis in the U.S.,” authors of the editorial wrote. “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.”