Predicting septic shock and mortality in pediatric sepsis using CBC indices with and without immature granulocyte: a prospective study

Abstract Background Pediatric sepsis is a major cause of mortality in low- and middle-income countries (LMICs), where affordable prognostic tools are needed. This study aimed to evaluate whether a panel of complete blood count (CBC) indices - hemoglobin (Hb), lymphocyte count, and platelet count - used jointly in logistic regression, with or without immature granulocyte (IG) percentage, predicts septic shock and in-hospital mortality in children with sepsis. Methods In this prospective study at a Tertiary Pediatric Centre (July 2023–July 2025), 276 children with sepsis (1 month–16 years) were enrolled. CBC indices were measured in all patients; immature granulocytes counts were available in a subgroup ( n = 52). Logistic regression with propensity score weighting assessed prediction of septic shock and mortality using area under the receiver operating characteristic curve (AUC), net reclassification improvement (NRI), integrated discrimination improvement (IDI), and decision curve analysis. Results The CBC panel model (Hb, lymphocyte count, platelet count) predicted septic shock (AUC 0.725) and mortality (AUC 0.746). Adding immature granulocytes improved AUCs in the subgroup (0.908 and 0.807), but NRI, IDI, and decision curves indicated limited incremental benefit. Conclusions CBC indices provide moderate-to-good prognostic accuracy in pediatric sepsis and remain practical in LMICs. IG offers slight improvement but limited feasibility.