Cerebrovascular Pressure Reactivity Has a Strong and Independent Association With Outcome in Children With Severe Traumatic Brain Injury

Abstract

Objectives: To examine cerebrovascular pressure reactivity index (PRx) in a large cohort of children with severe traumatic brain injury (sTBI) in association with physiologic variables and outcome.

Design: Retrospective observational cohort study.

Setting: Red Cross War Memorial Children's Hospital in Cape Town, South Africa.

Patients: Pediatric (≤ 14 yr old) sTBI patients with intracranial pressure (ICP) monitoring (postresuscitation Glasgow Coma Score [Glasgow Coma Scale (GCS)] of ≤ 8).

Measurements and main results: Data were analyzed from ICM+ files sampled at 100Hz. PRx (a mathematical indicator of pressure reactivity) was calculated as a moving correlation coefficient between ICP and mean arterial pressure (MAP) as previously described. Associations between PRx, age, GCS, ICP, MAP, and cerebral perfusion pressure (CPP) were examined with summary measures and correlation analysis using high-frequency data. Associations between PRx and mortality/outcome were examined with multivariable logistic regression analysis and the prognostic ability of PRx with receiver operating characteristic (ROCs) curves. The dataset included over 1.7 million minutes (28,634 hr) of MAP and ICP data in 196 children. The series mortality was 10.7% (21/196), and unfavorable outcome 29.6% (58/196). PRx had a moderate positive correlation with ICP ( r = 0.44; p < 0.001), a moderate negative correlation with CPP ( r = -0.43; p < 0.001), and a weak negative correlation with MAP ( r = -0.21; p = 0.004). PRx was consistently higher in patients with poor outcome and had a strong, independent association with mortality (ROC area under the curve = 0.91). A PRx threshold of 0.25 showed the best predictive ability for mortality.

Conclusions: This is the largest cohort of children with PRx analysis of cerebrovascular reactivity to date. PRx had a strong association with outcome that was independent of ICP, CPP, GCS, and age. The data suggest that impaired autoregulation is an independent factor associated with poor outcome and may be useful in directing clinical care.