Abstract
BACKGROUND AND OBJECTIVES: Cerebrovascular pressure autoregulation (CAR) maintains adequate cerebral blood flow (CBF) despite changes in cerebral perfusion pressure. CAR is disrupted after cardiac arrest, making the brain vulnerable to inadequate perfusion and oxygen delivery at population-derived guideline-recommended blood pressures. A metric of CAR can be used to determine the mean arterial pressure (MAP) at which CAR is most preserved, termed the optimal MAP (MAP(opt)). Our objective was to determine whether deviations above or below personalized CAR-derived MAP(opt) after pediatric cardiac arrest are associated with outcomes. METHODS: We conducted a retrospective analysis of prospectively collected data of patients aged 18 years or younger admitted to the pediatric intensive care unit at Children's Hospital of Philadelphia between October 2018 and December 2023 for post-cardiac arrest care. We computed cerebral oximetry index (COx), a metric of CAR, using a moving, linear correlation between time-synchronized brain tissue oxygenation (StO(2)) from near-infrared spectroscopy and MAP. A multiwindow weighted algorithm determined each patient's MAP(opt) over time. We compared each patient's MAP with their CAR-derived MAP(opt) during the first 72 hours after arrest. Unfavorable outcome was defined as a Pediatric Cerebral Performance Category score of 4-6 at hospital discharge with change of ≥1 from baseline. We tested association between burden (combining magnitude and duration) of MAP < MAP(opt)-5 mm Hg and unfavorable outcome, and between duration of MAP > MAP(opt) and favorable outcomes using logistic regression models adjusted for age, prearrest developmental disability, and measures of arrest severity. RESULTS: Among 147 patients included (median age 4.5 years, interquartile range [IQR] 1.1-11.7 years, 59% male), 52% had unfavorable outcomes. The median time from return of circulation to data collection was 4 (IQR 2.1-8.3) hours. The median burden of MAP < MAP(opt)-5 mm Hg was greater for the unfavorable outcome group compared with the favorable outcome group (192 [IQR 114-310] vs 147 [IQR 86-199] mm Hg·min/h, p = 0.002). A 1 SD higher burden of MAP < MAP(opt)-5 was associated with 2.4 times increased odds of unfavorable outcomes (95% CI 1.24-4.51). Patients with favorable outcomes had a longer duration of MAP > MAP(opt) than patients with unfavorable outcomes (48% [IQR 38-56] vs 40% [IQR 28-52], p = 0.011). One SD higher duration of MAP > MAP(opt) was associated with 2.5 times increased odds of favorable outcomes (95% CI 1.20-5.13). DISCUSSION: Greater burden of MAP < MAP(opt)-5 mm Hg in the first 72 hours after pediatric cardiac arrest was associated with increased odds of unfavorable outcomes after controlling for potential confounders, and longer duration of MAP > MAP(opt) was associated with increased odds of favorable outcomes.