Early Blood Pressure Changes in Neonatal Sepsis and the Risk of Mortality.

Abstract

To compare blood pressures (BP) between neonates with culture-proven sepsis and clinical sepsis in the first 120h of sepsis onset and to examine association between BP and in-hospital mortality. In this cohort study, consecutively enrolled neonates with 'culture-proven' sepsis [growth in blood/ cerebrospinal fluid (CSF) within 48h] and clinical sepsis (sepsis workup negative, cultures sterile) were analyzed. Their BP was recorded every 3-hourly for initial 120h and averaged in 20 time-epochs of 6h each (0-6h to 115-120h). BP Z-scores were compared between neonates with culture-proven vs. clinical sepsis and survivors vs. non-survivors. Two hundred twenty eight neonates (102-culture-proven and 126-clinical sepsis) were enrolled. Both groups had comparable BP Z-scores except significantly lower diastolic BP (DBP) and mean BP (MBP) in 0-6 and 13-18 time-epochs in culture-proven sepsis group. Fifty-four neonates (24%) died during their hospital stay. BP Z-scores in the initial 54h of sepsis were independently associated with mortality [systolic BP (SBP) Z-scores in first 54h, DBP Z-scores in first 24h, and MBP Z-scores in first 24h] after adjusting for gestational age, birth weight, cesarean delivery, and 5-min Apgar score. On receiver operating characteristic curves, SBP Z-scores showed better discriminative ability than DBP and MBP to identify non-survivors. Neonates with culture-proven and clinical sepsis had comparable BP Z-scores except low DBP and MBP in the initial few hours in culture-proven sepsis. BP in initial 54h of sepsis was significantly associated with in-hospital mortality. SBP discriminated non-survivors better than DBP and MBP.