Distinguishing Sepsis from Systemic Inflammation induced by Prolonged PGE1 infusion in Neonates with Duct-dependent Congenital Heart Disease: A Prospective Observational Study

Objective : Prolonged PGE1 infusion (PPG _E1 I) is common in low- and middle-income countries (LMICs). This prospective observational study was planned to assess role of biomarkers in differentiating sepsis from PGE1-induced systemic inflammation. Study Design: Neonates (gestation age, > 34 weeks) with duct-dependent congenital heart disease having received PGE1 for 5 days were enrolled in Pediatric Cardiac Intensive Care Unit of a tertiary care hospital in an LMIC. Culture-positive sepsis was excluded. Complete blood counts, CRP, procalcitonin and presepsin were obtained at enrolment; every 5 days thereafter; and whenever sepsis was suspected until end point (preoperative death/surgery). Correlations between biomarkers and duration of PGE1 infusion were assessed at non-sepsis observation points (N-SOPs). Comparisons were made between biomarkers at N-SOPs and sepsis observation points (SOPs). ROC analysis was performed to find cut-offs. Results: Eighty-nine observations (N-SOPs, 78; SOPs, 11) were made from 30 patients. At N-SOPs, CRP, presepsisn and platelet count increased linearly with duration of PGE1 (p=0.068, p=0.055 and p<0.0001 respectively). CRP (p<0.001), procalcitonin (p<0.001) and presepsin (p<0.001) were higher, while platelet count was lower (p=0.003) at SOPs compared to at N-SOPs. Cut-offs to detect sepsis were— CRP: 19.6mg/dL (AUC, 0.97; sensitivity, 100%; specificity, 88%), procalcitonin: 0.55ng/mL (0.91; 90%; 82%) and presepsin: 2.23ng/mL (0.98; 100%; 94%). Conclusions: PPG _E1 I led to rise in CRP, presepsin and platelet count in dose-dependent manner. CRP >19.6mg/dL seems to detect sepsis with high sensitivity but lower specificity. Presepsin (>2.23ng/mL) seems to outperform CRP, procalcitonin and their combination. Lower platelet count may play a supporting role.