Deviations from NIRS-derived optimal blood pressure are associated with worse outcomes after pediatric cardiac arrest

Abstract

Aim: Evaluate cerebrovascular autoregulation (CAR) using near-infrared spectroscopy (NIRS) after pediatric cardiac arrest and determine if deviations from CAR-derived optimal mean arterial pressure (MAPopt) are associated with outcomes. Methods: CAR was quantified by a moving, linear correlation between time-synchronized mean arterial pressure (MAP) and regional cerebral oxygenation, called cerebral oximetry index (COx). MAPopt was calculated using a multi-window weighted algorithm. We calculated burden (magnitude and duration) of MAP less than 5 mmHg below MAPopt (MAPopt - 5), as the area between MAP and MAPopt - 5 curves using numerical integration and normalized as percentage of monitoring duration. Unfavorable outcome was defined as death or pediatric cerebral performance category (PCPC) at hospital discharge ≥3 with ≥1 change from baseline. Univariate logistic regression tested association between burden of MAP less than MAPopt - 5 and outcome. Results: Thirty-four children (median age 2.9 [IQR 1.5,13.4] years) were evaluated. Median COx in the first 24 h post-cardiac arrest was 0.06 [0,0.20]; patients spent 27% [19,43] of monitored time with COx ≥ 0.3. Patients with an unfavorable outcome (n = 24) had a greater difference between MAP and MAPopt - 5 (13 [11,19] vs. 9 [8,10] mmHg, p = 0.01) and spent more time with MAP below MAPopt - 5 (38% [26,61] vs. 24% [14,28], p = 0.03). Patients with unfavorable outcome had a higher burden of MAP less than MAPopt - 5 than patients with favorable outcome in the first 24 h post-arrest (187 [107,316] vs. 62 [43,102] mmHg × Min/Hr; OR 4.93 [95% CI 1.16–51.78]). Conclusions: Greater burden of MAP below NIRS-derived MAPopt - 5 during the first 24 h after cardiac arrest was associated with unfavorable outcomes.