Abstract
OBJECTIVES: Targets for treatment of raised intracranial pressure or decreased cerebral perfusion pressure in pediatric neurocritical care are not well defined. Current pediatric guidelines, based on traumatic brain injury, suggest an intracranial pressure target of less than 20 mm Hg and cerebral perfusion pressure minimum of 40-50 mm Hg, with possible age dependence of cerebral perfusion pressure. We sought to define intracranial pressure and cerebral perfusion pressure thresholds associated with inhospital mortality across a large single-center pediatric neurocritical care cohort. DESIGN: Retrospective chart review. SETTING: PICU, single quaternary-care center. PATIENTS: Individuals receiving intracranial pressure monitoring from January 2012 to December 2016. INTERVENTIONS: None. MEASUREMENTS AND MAIN RESULTS: Intracranial pressure and cerebral perfusion pressure measurements from 262 neurocritical care patients (87 traumatic brain injury and 175 nontraumatic brain injury; 63% male; 8.3 ± 5.8 yr; mortality 11.1%). Mean intracranial pressure and cerebral perfusion pressure had area under the receiver operating characteristic curves of 0.75 and 0.64, respectively, for association of inhospital mortality. Cerebral perfusion pressure cut points increased with age (< 2 yr = 47, 2 to < 8 yr = 58 mm Hg, ≥ 8 yr = 73 mm Hg). In the traumatic brain injury subset, mean intracranial pressure and cerebral perfusion pressure had area under the receiver operating characteristic curves of 0.70 and 0.78, respectively, for association of inhospital mortality. Traumatic brain injury cerebral perfusion pressure cut points increased with age (< 2 yr = 45, 2 to < 8 yr = 57, ≥ 8 yr = 68 mm Hg). Mean intracranial pressure greater than 15 mm Hg, male sex, and traumatic brain injury status were independently associated with inhospital mortality (odds ratio, 14.23 [5.55-36.46], 2.77 [1.04-7.39], and 2.57 [1.03-6.38], respectively; all p < 0.05). Mean cerebral perfusion pressure less than 67 mm Hg and traumatic brain injury status were independently associated with inhospital mortality (odds ratio, 5.16 [2.05-12.98] and 3.71 [1.55-8.91], respectively; both p < 0.01). In the nontraumatic brain injury subset, mean intracranial pressure had an area under the receiver operating characteristic curve 0.77 with an intracranial pressure cut point of 15 mm Hg, whereas mean cerebral perfusion pressure was not predictive of inhospital mortality. CONCLUSIONS: We identified mean intracranial pressure thresholds, utilizing receiver operating characteristic and regression analyses, associated with inhospital mortality that is below current guidelines-based treatment targets in both traumatic brain injury and nontraumatic brain injury patients, and age-dependent cerebral perfusion pressure thresholds associated with inhospital mortality that were above current guidelines-based targets in traumatic brain injury patients. Further study is warranted to identify data-driven intracranial pressure and cerebral perfusion pressure targets in children undergoing intracranial pressure monitoring, whether for traumatic brain injury or other indications.