Abstract
BACKGROUND: Mechanical power (MP) of ventilation is the energy dissipated into the respiratory system by the ventilator per minute. Adult studies have demonstrated positive correlation between MP and mortality. This study aims to find out the predictive ability of MP for mortality in ventilated children. OBJECTIVES: We aimed to assess the predictive ability of Mechanical power (MP) at 12 and 24 hours of ventilation for mortality in children during Paediatric Intensive Care Unit (PICU) stay. The secondary outcomes were to find-out the correlation between MP vs 28-day ventilator free days (VFD), duration of ICU stay and hospital stay. MATERIAL AND METHODS: This prospective observational study was conducted over 1 year at 12-bedded PICU of a tertiary care referral and teaching hospital of North India after obtaining IRB clearance. 75 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. AU-ROC analysis of measured variables was performed to evaluate diagnostic accuracy. Spearman's Rank correlation coefficient used to find-out the correlation between 28-day VFD, hospital stay, ICU stay with MP. RESULTS: Of the 75 participants, 52 were male with a median age of 48 months (IQR-16.5,108). The mean MP(±SD) 12 and 24 hours were 8.11±5.44 and 8.29±6.51 respectively. The mean MP(±SD) 12 values were 7.27±4.0 vs 12.51±9.17 and for MP 24 hours 7.26±4.46 vs 13.7±11.60 for survivor (n=63) and non-survivor (n=12) group respectively (p<0.05). Area under ROC to predict in-hospital mortality was higher for MP 24 (AU-ROC 0.706) when compared to MP 12 hours (AU-ROC 0.687) (p >0.05). MP to predicted body weight (MP/PBW) had a better AU-ROC compared to MP or driving pressure (AU-ROC for MP/PBW 24 = 0.73). For MP 12 hours, a value of 10.5 had the sensitivity 42%, specificity 84%, positive predictive value (PPV) 33%, negative predictive value (NPV) 88 % and diagnostic accuracy of 77%, while for MP 24 hours a value of 8.95 had a sensitivity 67%, specificity 78%, PPV 36%, and NPV 92% and diagnostic accuracy of 72% to predict mortality. At cut off value of < 8.95 J/min of MP 24 had survival benefit (Kaplan-Meier survival analysis Log rank 8.1, p<0.05). In subgroup analysis of survivors, MP had negative correlation (r=-0.6, p<0.05) with 28 day-VFD in children aged 12-18 years. CONCLUSION: MP and MP/PBW at 24 hours of ventilation is a better predictor of mortality than MP at 12 hours in children. Patients with higher MP should be considered for early escalation to extracorporeal support. In survivor group, higher MP was associated with fewer 28 day-VFD in children ≥ 12 years.