Pediatric Sepsis Interventions and Mortality in a Tertiary Emergency Medicine Department in Tanzania

Background: About 70% of pediatric deaths in Tanzania are attributed to preventable and curable diseases, including sepsis. Emergency medical care and early care play a critical role in improving the survival of pediatric sepsis patients. The Emergency Medicine Department (EMD) at Muhimbili National Hospital (MNH), a tertiary care hospital in Dar es Salaam, Tanzania, developed and implemented a Pediatric Severe Sepsis and Septic Shock Protocol for early recognition and treatment of children presenting with sepsis. This study sought to describe the pediatric sepsis protocol interventions that were received. This study also sought to explore the association between antibiotic administration and mortality.

Methods: This was a secondary analysis from a prospective cohort study of children aged 28 days to 14 years (N=2032) with sepsis presenting to MNH between July 1, 2016 and June 30, 2017. Descriptive statistics were utilized to characterize and describe the cohort, the administration of pediatric sepsis protocol interventions, and reasons for non-administration of antibiotics. A multivariable logistic regression model was performed to test the association between antibiotic administration and pediatric mortality while adjusting for age, severity of illness as measured by the Lambaréné Organ Dysfunction Score (LODS), human immunodeficiency virus (HIV) status, malaria status, malnutrition status, and referral history.

Results: Final outcomes were available in 1804 subjects (88.78%); 203 (11.25%) did not survive to hospital discharge. The median age of the cohort was calculated as 25.26 months (IQR 13.14-54.35 months) and 56.99% (n=1158) of subjects were male. Of the essential pediatric sepsis protocol interventions, antibiotics, at least one fluid bolus and supplemental oxygen were administered to 44.49% (n=904), 32.32% (n=656), and 18.60% (n=378) of subjects, respectively. The most common reason that antibiotics were not administered was because they were not ordered (N=1161, 96.73%). Pediatric sepsis patients who received antibiotics had an unadjusted and adjusted odds ratio (OR) for mortality of 2.88 (95% CI, 2.12-3.96) and 2.56 (95% CI, 1.13-6.23), respectively, compared to children who did not receive antibiotics.

Conclusion: In this cohort, antibiotic administration was associated with increased mortality in children with sepsis, and the most common reason for antibiotics not being given was that they were not ordered. A limitation of this study is that we do not have direct data on why physicians did not order antibiotics. There may also be residual confounding effects, such as the duration of illness, treatment prior to arrival to MNH, timing of antibiotic administration, and sepsis etiology (bacterial, viral, fungal, or parasitic), which were unaccounted for in this study. Future research should explore potential mechanisms underlying this observed association and the association with sepsis protocol adherence and patient outcomes.