Association of Postarrest Hypotension Burden With Unfavorable Neurologic Outcome After Pediatric Cardiac Arrest

Abstract

OBJECTIVE: Quantify hypotension burden using high-resolution continuous arterial blood pressure (ABP) data and determine its association with outcome after pediatric cardiac arrest. DESIGN: Retrospective observational study. SETTING: Academic PICU. PATIENTS: Children 18 years old or younger admitted with in-of-hospital or out-of-hospital cardiac arrest who had invasive ABP monitoring during postcardiac arrest care. INTERVENTIONS: None. MEASUREMENTS AND MAIN RESULTS: High-resolution continuous ABP was analyzed up to 24 hours after the return of circulation (ROC). Hypotension burden was the time-normalized integral area between mean arterial pressure (MAP) and fifth percentile MAP for age. The primary outcome was unfavorable neurologic status (pediatric cerebral performance category ≥ 3 with change from baseline) at hospital discharge. Mann-Whitney U tests compared hypotension burden, duration, and magnitude between favorable and unfavorable patients. Multivariable logistic regression determined the association of unfavorable outcomes with hypotension burden, duration, and magnitude at various percentile thresholds from the 5th through 50th percentile for age. Of 140 patients (median age 53 [interquartile range 11–146] mo, 61% male); 63% had unfavorable outcomes. Monitoring duration was 21 (7–24) hours. Using a MAP threshold at the fifth percentile for age, the median hypotension burden was 0.01 (0–0.11) mm Hg-hours per hour, greater for patients with unfavorable compared with favorable outcomes (0 [0–0.02] vs. 0.02 [0–0.27] mm Hg-hr per hour, p < 0.001). Hypotension duration and magnitude were greater for unfavorable compared with favorable patients (0.03 [0–0.77] vs. 0.71 [0–5.01]%, p = 0.003; and 0.16 [0–1.99] vs. 2 [0–4.02] mm Hg, p = 0.001). On logistic regression, a 1-point increase in hypotension burden below the fifth percentile for age (equivalent to 1 mm Hg-hr of burden per hour of recording) was associated with increased odds of unfavorable outcome (adjusted odds ratio [aOR] 14.8; 95% CI, 1.1–200; p = 0.040). At MAP thresholds of 10th–50th percentiles for age, MAP burden below the threshold was greater in unfavorable compared with favorable patients in a dose-dependent manner. CONCLUSIONS: High-resolution continuous ABP data can be used to quantify hypotension burden after pediatric cardiac arrest. The burden, duration, and magnitude of hypotension are associated with unfavorable neurologic outcomes.