P2A229: Racial Disparty in Pediatric Sepsis Mortality

Background: Despite improvements in early recognition and resuscitation, sepsis remains a major pediatric health issue with an estimated 40,000 hospitalizations and 5,000 deaths every year in the US. Many centers now use automated, real-time, algorithm-based detection of sepsis, severe sepsis, and septic shock incorporated into the electronic medical record (EMR). This method leads to earlier recognition, resuscitation, and improved outcomes. Recent data shows a continued discrepancy in sepsis outcomes based on race, despite improvements in outcomes among children overall. We hypothesized that mortality rates remain higher in black children with sepsis, severe sepsis, and septic shock compared to white children, despite incorporation of automated screening tools into the EMR.

Methods: A retrospective analysis of all patients at Arkansas Children’s Hospital (ACH) with sepsis, severe sepsis, or septic shock between January 2018 and April 2022 was conducted. ACH uses a best practice advisory (BPA) in the EMR for early detection in all hospital areas including the emergency department, all medical-surgical wards, all Intensive Care Units (ICU), and interfacility transport. EMR activation leads to a bedside huddle, followed by institution of clinical interventions. A sepsis episode (SE) was defined as BPA activation or a diagnosis of sepsis, severe sepsis, or septic shock in the EMR. Mortality rates, as well as demographic information and clinical outcome measures for children who died were compared between Black(B) and White(W) children. Student's t-test was used for categorical variables, chi-square for proportions, and odds ratio for overall mortality comparison.

Results: 3,514 patients had a SE during the study period; 2126(W) and 736(B). Overall mortality was 1.65% (40%(B); 47%(W)). Mortality was 3.13% (23/736) in Black children versus 1.27% in White children; OR 2.51 (1.43,4.40), p=0.001. Basic demographics including gender (Female-56.52%(B), 55.56%(W); p=0.95) and age in years (8.00+/-2.78(B), 7.87+/-2.66; p=0.97) were similar. Clinical interventions including total IV antibiotic days (23.83+/-8.36(B), 21.56+/-9.59(W); p=0.38), vasoactive infusion days (2.17+/-1.44(B), 2.63+/-0.90; p=0.18), and percent requiring Extracorporeal Membrane Oxygenation (26.07%(B), 18.52%(W), p=0.52) were similar. Black children who died had a longer length of hospitalization (16.70+/-6.47(B), 12.70+/-5.85(W) days; p=0.03) and longer ICU stay (7.57+/-2.57(B), 5,70+/-2.27(W) days; p=0.01). Percent over threshold for antibiotic administration (21.74%(B), 18.52%(W); p= 0.78; 1h-septic shock, 3h-sespis) and IV fluid bolus administration (8.70%(B), 14.81%(W); p=0.51; 20 minutes-septic shock, 1h-sepsis) were similar.

Conclusion: This single center, retrospective study shows that hospitalized Black children have a 2.5 times greater risk of death from sepsis compared to White children. Interestingly, this preliminary data suggests similar rates of timely resuscitation including antibiotic administration and IV fluid boluses. Further investigations are needed to identify biases (conscious and unconscious), potential socio-economic factors, and genetic predispositions leading to racial disparities in outcomes of children with pediatric sepsis, severe sepsis, and septic shock.