Concordance Between Bedside and Electronic Detection of Intracranial Pressure Crises: Insights From the Brain Tissue Oxygen Monitoring and Management in Severe Traumatic Brain Injury II Trial.

Abstract

Traumatic brain injury (TBI) research and intracranial pressure (ICP) management depends on bedside ICP (B-ICP) crisis identification. We analyzed background-collected electronic ICP (E-ICP) data to study the concordance of this identification. Post hoc comparison of background-collected continuous E-ICP data to routine B-ICP information during severe TBI management in the Brain Tissue Oxygen Monitoring and Management in Severe TBI (BOOST II) randomized trial. Ten U.S. ICUs. Seventy of 110 randomized BOOST II severe TBI patients with complete datasets for this study. None. We studied "minimalist" B-ICP episodes lasting less than or equal to 60 minutes and requiring only tier 1 treatments (83% of total BOOST II I isolated ICP episodes). Bedside clinicians (BCs) identified and treated 509 minimalist B-ICP episodes, defined by 5 minutes of B-ICP greater than or equal to 20 mm Hg (B-ICP 5 ). Corresponding E-ICP during this defining period (E-ICP 5 ) confirmed only 47% of these. The 241 "concordant" B-ICP episodes (both B-ICP 5 and E-ICP 5 ≥ 20 mm Hg) had average E-ICP values for the entire B-ICP episodes (E-ICP AVG ) less than 20 mm Hg in 38%. The 286 "discordant" B-ICP episodes (B-ICP 5 ≥ 20 mm Hg but E-ICP 5 < 20 mm Hg) had E-ICP AVG values less than 20 mm Hg in 76% and both E-ICP AVG and maximal E-ICP's (E-ICP MAX ) values less than 20 mm Hg in 31%. Testing for confounding (e.g., brief, first, or easily controlled B-ICP episodes) did not provide explanations. Insufficient data were available to evaluate temporal asynchrony between BC and electronic datapoints. Study limitations were lacking a rigid end-of-B-ICP episode definition and inability to fully control for temporal synchrony confounding. Our findings suggest that current methods of clinical identification of B-ICP episodes may not reliably distinguish true episodes of sustained ICP for the most common B-ICP episode type. We suggest developing open-source, real-time, temporally synchronized electronic B-ICP episode definition methods to direct future treatment and research.