National Outcomes of Sepsis in Publicly Insured Children: 2021–2023

OBJECTIVE To identify incidence and factors associated with sepsis mortality. METHODS We conducted a cohort study of all US publicly insured children younger than age 19 years hospitalized with a diagnosis code–based definition of sepsis in 2021–2023 using the Transformed Medicaid Statistical Information System. The outcome was death within 30 days of sepsis onset. Regions were assigned using the Atlas of Pediatric Acute Care. We determined crude mortality and expected death counts by region using a patient-level model including demographics and illness severity. Regional standardized mortality rates (SMRs) were calculated as observed-to-expected ratios. Outlier regions had 95% CIs of SMR not crossing 1. Expected numbers of outlier regions were simulated. We evaluated region-level mortality risk factors using hierarchical models, then determined 30-day rehospitalizations. RESULTS There were 36 552 sepsis episodes with an incidence of 45.1 episodes per 100 000 person-years. Among 40 524 community-acquired sepsis episodes, 1562 (3.9%) died within 30 days. Patient age, complex chronic condition type, and organ dysfunction were associated with mortality. Regional SMRs ranged from 0.00 to 3.82; 14 of 104 regions were low-mortality and 12 of 104 high-mortality outliers, exceeding simulated expectations. Regions with a children’s hospital (adjusted odds ratio [aOR], 0.76; 95% CI, 0.61–0.94) and higher sepsis incidence (aOR, 0.91; 95% CI, 0.86–0.97 per sepsis episode per 10 000 person-years) were associated with lower mortality. Thirty-day rehospitalizations occurred after 16.7% of sepsis episodes. CONCLUSIONS There was substantial regional variation in pediatric sepsis mortality. Regions with a children’s hospital and higher sepsis incidence had lower mortality. There were more high- and low-mortality outliers than expected.