In this study, available medical literature were reviewed to determine whether brain hypoxia as measured by brain tissue oxygen (Bto2) levels is associated with increased risk of poor outcome after traumatic brain injury (TBI). A secondary objective was to examine the safety profile of a direct BtO2 probe.
Clinical studies published between 1993 and 2008 were identified from electronic databases, Index Medicus, bibliographies of pertinent articles, and expert consultation. The following inclusion criteria were applied for outcome analysis: 1) more than 10 patients described, 2) use of a direct Bto2 monitor, 3) brain hypoxia defined as Bto2 <10 mm Hg for >15 or 30 minutes, 4) 6-month outcome data, and 5) clear reporting of patient outcome associated with Bto2. For the analysis, each selected article had to have adequate data to determine odds ratios (ORs) and confidence intervals (CIs). Thirteen studies met the initial inclusion criteria and three were included in the final outcome analysis. Safety data were abstracted from any report where it was mentioned.
The three studies included 150 evaluable patients with severe TBI (Glasgow Coma Scale ≤8). Brain hypoxia was identified in 71 (47%) of these patients. Among the patients with brain hypoxia, 52 (73%) had unfavorable outcome including 39 (55%) who died. In the absence of brain hypoxia, 34 (43%) patients had an unfavorable outcome, including 17 (22%) who died. Overall brain hypoxia (Bto2 <10 mm Hg >15 minutes) was associated with worse outcome (OR 4.0; 95% CI 1.9–8.2) and increased mortality (OR 4.6; 95% CI 2.2–9.6). We reviewed published safety data; in 292 patients monitored with a Bto2 probe, only two adverse events were reported.
Summary results indicate that brain hypoxia (<10 mm Hg) is associated with worse outcome after severe TBI and that Bto2 probes are safe. These results imply that treating patients to increase Bto2 may improve outcome after severe TBI. This question will require further study.
From the Department of Neurosurgery (EM-W, SB, PDL), Division of Trauma Surgery and Surgical Critical Care (VG), and Department of Biostatistics and Epidemiology and Center for Clinical Epidemiology and Biostatistics (WY), University of Pennsylvania School of Medicine, Philadelphia, PA; and Stat-Trade, Inc. (AI, KH, SC), Morrisville, PA.
The authors have not disclosed any potential conflicts of interest.
For information regarding this article, E-mail: firstname.lastname@example.org