(1201) Predictive Ability of Mechanical Power for Mortality in Ventilated Children: A Prospective Study

Introduction: Mechanical power (MP) of ventilation is the energy dissipated into the respiratory system by the ventilator per minute. We aimed to assess the predictive ability of MP at 12 and 24 hours of ventilation for mortality in children during PICU stay. The secondary outcomes were to study association between MP with 28-day ventilator free days (VFD), durations of PICU and hospital stay.

Methods: This prospective observational study was conducted over 18 months at 12-bedded PICU of a tertiary care referral hospital of North India after obtaining IRB clearance. 157 new orotracheally ventilated children receiving decelerating flow were enrolled after parental consent. MP [= 0.098×RR×VT(PEEP+∆Pinsp) J/min] at 12 and 24 hours of ventilation was calculated. Children were followed for survival and duration of stay in PICU.

Results: Of the 157 participants with a median age of 64 months (IQR-18,127), 107 were males. The mean(±SD) MP 12 and 24 hours were 6.16±4.0 and 6.26±4.85 respectively. The mean(±SD) MP 12 values were 5.7±3.2 vs 8.4±6.3 and for MP 24 hours 5.5±3.1 vs 9.9±8.9 for survivor (n=131) and non-survivor (n=26) group respectively (all p< 0.05). Area under ROC to predict in-hospital mortality was higher for MP 24 (AU-ROC 0.656) when compared to MP 12 hours (AU-ROC 0.609) (p >0.05). MP to predicted body weight (MP/PBW 24) had a better AU-ROC (0.745, 95% CI=0.63-0.85) compared to MP 24 and driving pressure 24 (0.692). For MP/PBW 12 hours, a value of 0.41 had the sensitivity 57%, specificity 77, positive predictive value (PPV) 33%, negative predictive value (NPV) 90 % and diagnostic accuracy of 73%, while for MP/PBW 24 hours a value of 0.42 had a sensitivity 65%, specificity 80%, PPV 39%, and NPV 92% and diagnostic accuracy of 77% to predict mortality. For every one unit increase in MP/PBW 24, the log-odds of mortality increase by 4.43 (OR=84, CI=8.7-811.4, P< 0.05). In subgroup analysis of survivors, MP 12 and 24 had negative correlation (r= -0.49 and -0.46, p< 0.05) with 28 day-VFD in children aged 12-18 years.

Conclusions: MP/PBW 24 hr was a better predictor of mortality compared to MP 24 in ventilated children. In survivor group, higher MP was associated with fewer 28 day-VFD in children ≥ 12 years.