P0654 / #2072: PREDICTORS OF MORTALITY IN CRITICAL ILL SEPSIS CHILDREN IN 5 YEARS: INDONESIAN EXPERIENCE

Abstract

Aims & Objectives: Pediatric sepsis in Indonesia had different outcomes with small sample size. Unfortunately, many hospitals had limitations to perform PELOD-2 score at admission because limited resources and lack of facilities. These problems make patients referred late to the hospital with severe condition. The objective of this study is to determine the risk factors of mortality by clinical manifestation and sepsis biomarkers. Methods: This study is retrospective study with consecutive sampling, by aged 1 month-18 years admitted to ER or PICU Cipto Mangunkusumo Hospital Jakarta during January 2015-May 2020. We divided subjects into two groups survivors and non-surviviors. Clinical manifestations (consciousness, respiratory rate, blood pressure, capillary refill time, and temperature), sepsis biomarker (leukocytes, neutrophils count, C-reactive protein/CRP, and procalcitonin) was documented. We performed external validation of PELOD-2 score and analyze using multivariable logistic regression to determine predictors of mortality. Results: We identified 225 septics patients with 65% of mortality. In this study 56% had respiratory infection, 31% on mechanical ventilation, 59% with hypotension and 51% was used inotropes. The non-survivors had cut-off 8 for PELOD-2 score with sensitivity 73,42%, specificity 82,19% and AUC 66%. Bivariate analysis showed decrease of consciousness (p=0,001), tachypnea (p=0,030), hypotension (p=0,001), CRT>2’’ (p=0,080) and biomarkers showed neutrophils (p=0,153) and CRP ≥ 5 mg/L (p=0,155). Multivariable logistic regression analysis determined predictors of mortality in sepsis children were hypotension (OR=4,294), neutrophilia (OR=3,102) and CRP ≥ 5 mg/L (OR= 2,789). Conclusions: PELOD-2 score ≥ 8 was found correlated with high mortality. However in limited resource setting, the predictors of mortality were hypotension, neutrophilia and CRP ≥ 5.