Compliance with the bundle is not enough: critical analysis of pediatric sepsis in a hospital ina middle-income country

Abstract

Pediatric sepsis is one of the leading causes of infant mortality worldwide. Early implementation of evidence-based management packages has been shown to improve survival; however, their application faces challenges, particularly in middle-income countries. This study evaluated whether adherence to  the  sepsis  protocol  within  the  f irst  hour  is  associated  with  better  clinical  outcomes  in  a  tertiary  public hospital in Ecuador. Methods: Observational, retrospective, and analytical study conducted between  2022  and  2025  in  children  under  f ive  years  of  age  with  a  clinical  diagnosis  of  sepsis  or  septic shock. Compliance with the sepsis protocol (blood cultures before antibiotics, antibiotics ≤1 h, and initiation of adequate fluid therapy) and its association with mortality, ICU admission, and length  of  hospitalization  were  evaluated  using  bivariate  analysis  and  logistic  regression  (crude  and  adjusted). Results: Ninety-six patients with a median age of two years (IQR: 1–3) were included. Overall mortality was 7.3%, and 32% of cases required admission to the ICU. Thirty-seven-point five percent were classified as adhering to the package (3 components) within the first hour. Compliance with the package was not associated with mortality (OR: 1.27; 95% CI: 0.27–6.04; p = 0.762), ICU admission (p = 0.857), or length of hospital stay (p = 0.241). After adjusting for age, sex, pulmonary focus, and use of vasopressors, the association remained nonsignificant (adjusted OR = 1.17; 95% CI: 0.21–6.40; p = 0.860). The use of vasoactive drugs was significantly associated with mortality (p = 0.0003). Conclusions: Compliance with the sepsis package was not associated with better clinical outcomes in this cohort. These findings should be interpreted considering the structural limitations of the local context, such as delays in recognition, difficulties in real-time recording, and partial implementation of protocols. In addition, the small sample size limits statistical power, so clinically relevant associations may not have reached significance. Contextualized, multicenter interventions are needed to strengthen pediatric sepsis care in vulnerable settings.