See the corresponding editorial in this issue, p 643. J Neurosurg 111: , 2009

Transcription

1 See the corresponding editorial in this issue, p 643. J Neurosurg 111: , 2009 Management guided by brain tissue oxygen monitoring and outcome following severe traumatic brain injury Clinical article Ro s s P. Mar t i n i, B.S., 1,2 St e v e n De e m, M.D., 2,3 N. Dav i d Ya n e z, Ph.D., 4 Ra n da l l M. Ch e s n u t, M.D., 5 No e l S. We i s s, M.D., Dr.P.H., 6 St e p h e n Da n i e l, Ph.D., 4 Mi c h a e l So u t e r, M.B., Ch.B., F.R.C.A., 2,5 a n d Mir i a m M. Tr e g g i a r i, M.D., Ph.D., M.P.H. 2,5 1 The Warren Alpert Medical School, Brown University, Providence, Rhode Island; Departments of 2 Anesthesiology and Pain Medicine and 3 Medicine, University of Washington School of Medicine; 4 Department of Biostatistics, University of Washington School of Public Health and Community Medicine; 5 Department of Neurological Surgery, University of Washington School of Medicine; and 6 Department of Epidemiology, University of Washington School of Public Health and Community Medicine, Seattle, Washington Object. The authors sought to describe changes in clinical management associated with brain tissue oxygen (PbO 2 ) monitoring and how these changes affected outcomes and resource utilization. Methods. The cohort study comprised 629 patients admitted to a Level I trauma center with a diagnosis of severe traumatic brain injury over a period of 3 years. Hospital mortality rate, neurological outcome, and resource utilization of 123 patients who underwent both PbO 2 and intracranial pressure (ICP) monitoring were compared with the same measures in 506 patients who underwent ICP monitoring only. The main outcomes were hospital mortality rate, functional independence at hospital discharge, duration of mechanical ventilation, hospital length of stay, and hospital cost. Multivariable regression with robust variance was used to estimate the adjusted differences in the main outcome measures between patient groups. The models were adjusted for patient age, severity of injury, and pathological features seen on head CT scan at admission. Results. On average, patients who underwent ICP/PbO 2 monitoring were younger and had more severe injuries than patients who received ICP monitoring alone. Relatively more patients treated with PbO 2 monitoring received osmotic therapy, vasopressors, and prolonged sedation. After adjustment for baseline characteristics, the hospital mortality rate was, if anything, slightly higher in patients undergoing PbO 2 -guided management than in patients monitored with ICP only (adjusted mortality difference 4.4%, 95% CI 3.9 to 13%). Patients who underwent PbO 2 - guided management also had lower adjusted functional independence scores at hospital discharge (adjusted score difference 0.75, 95% CI 1.41 to 0.09). There was a 27% relative increase (95% CI 6 53%) in the median hospital length of stay when the PbO 2 group was compared with the ICP-only group. Conclusions. The mortality rate in patients with traumatic brain injury whose clinical management was guided by PbO 2 monitoring was not reduced in comparison with that in patients who received ICP monitoring alone. Brain tissue oxygen monitoring was associated with worse neurological outcome and increased hospital resource utilization. (DOI: / JNS08998) Ke y Wo r d s intracranial pressure human trauma resuscitation neurological outcomes cerebral ischemia Abbreviations used in this paper: AIS = Abbreviated Injury Scale; CPP = cerebral perfusion pressure; FIM = Functional Independence Measure; GCS = Glasgow Coma Scale; FiO 2 = fraction of inspired O 2 ; ICP = intracranial pressure; ICU = intensive care unit; ISS = Injury Severity Scale; LOS = length of stay; MABP = mean arterial blood pressure; PbO 2 = brain tissue O 2 ; SAH = subarachnoid hemorrhage; SAPS II = Simplified Acute Physiology Score II; TBI = traumatic brain injury. An important focus in the management of patients with TBI is the minimization and prevention of secondary injury. 6 Therapies to optimize cerebral perfusion and prevent ischemic injury are currently guided by ICP monitoring. 3 Refractory elevated ICP has been associated with poor outcome, 16,22 but several studies have shown that ischemic injury can occur in the absence of elevated ICP. 9,23,26 Recently, PbO 2 monitoring has been introduced in clinical practice to guide management of patients with severe TBI and to complement the data acquired by the ICP monitor. 2 Brain tissue oxygenation is measured by an intraparenchymal electrode that can be inserted through the same bur hole as an ICP monitor. Both the depth and duration of episodes of early brain tissue hypoxia have been associated with increased mortality rate and worse long-term neurological outcome. 30 A higher mortality rate has been described in relation to 644 J Neurosurg / Volume 111 / October 2009

2 Brain tissue oxygen monitoring guided management increasing duration of PbO 2 values < 15 mm Hg and instances of PbO 2 < 6 mm Hg. 29 No randomized trials have been conducted to evaluate PbO 2 -guided management in patients with TBI. Two studies compared outcomes in patients with TBI before and after the introduction of PbO 2 - guided management. Meixensberger et al. 19 observed that while PbO 2 -guided management of cerebral perfusion was associated with relative fewer episodes of cerebral hypoxia, there was no difference in 6-month neurological outcome. In a study of 53 patients, Stiefel et al. 28 reported that in patients who underwent PbO 2 -guided interventions the mortality rate was lower (25 and 44%, respectively) and neurological outcome was improved compared with patients without PbO 2 -guided interventions. Although the existing data suggest that episodes of cerebral hypoxia identified by PbO 2 monitoring are predictive of an adverse outcome in patients with severe TBI, it remains unclear whether PbO 2 -guided management of these patients is beneficial. The aim of the present study was to ascertain changes in clinical management associated with the use of PbO 2 monitoring and to estimate the association between the use of this monitoring modality and mortality rate, neurological outcome, and resource utilization in a large population of severe TBI patients. Methods Patient Population The present study is based on the experience obtained in all patients with a diagnosis of severe TBI (admission GCS score 8) admitted to Harborview Medical Center (Seattle, Washington) between July 1, 2004, and October 15, Harborview Medical Center is a 413-bed municipal medical center affiliated with the University of Washington and is the only Level 1 trauma center in a 5-state area (Washington, Wyoming, Alaska, Montana, and Idaho). All data were available from the hospital database originating from computerized medical and billing records and from a prospectively collected registry of trauma-related admissions. 21 The study was approved by the University of Washington institutional review board, which waived the need for informed consent. Monitoring Patients were monitored using an ICP monitor (Camino, Integra NeuroSciences) alone or a PbO 2 monitor (Licox Brain Tissue Oxygenation Probe, Integra NeuroSciences) in addition to the ICP monitor. Indication for PbO 2 monitoring was based on the discretion of the attending neurosurgeon, without the use of specific guidelines with respect to the indication for placement. Both monitors were inserted through a triple-lumen bolt into the frontal region, using the same bur hole. The PbO 2 and/or ICP monitor was inserted within the first 48 hours of ICU admission. Preferred placement was into the healthy tissue on the side displaying maximal pathological features or swelling on CT scanning. Correct placement was confirmed by examining daily follow-up head CT scans. Hemodynamic and neurological variables were continuously monitored and recorded hourly using a bedside monitor. J Neurosurg / Volume 111 / October 2009 Management All patients with isolated TBI were admitted to the critical care unit and comanaged by neurological surgery and neurocritical care services, according to head-injury treatment guidelines issued by the Brain Trauma Foundation. 1 Trauma intensivists, with neurological surgery consultation, managed the cases of patients with multiple traumatic injuries. Briefly, the threshold for treatment of brain tissue hypoxia was a PbO 2 < 20 mm Hg. A 100% FiO 2 challenge was performed to assess the reliability of PbO 2 values. 11 If the monitor was functioning properly, the FiO 2 challenge was followed by treatment corresponding to the severity of the patient s constellation of signs and symptoms. Patients with concomitant intracranial hypertension received intervention that was primarily directed at reducing ICP, according to CPP management guidelines. Patients with hypoxic PbO 2 and without intracranial hypertension were treated with therapy directed to increase PbO 2, according to 3 tiers of therapy escalation. Tier 1 management included adjustment of the head of bed to > 30, temperature management to < 37.5 C, increasing CPP to > 60 mm Hg, 12 optimization of hemodynamic profile with volume replacement and cardiac output monitoring, and possible continuous jugular bulb O 2 saturation monitoring in the absence of a response to previous measures. Tier 2 management included increasing FiO 2 to 0.60 and transfusion of packed red blood cells to a target hemoglobin of 10 g/dl. 27 Tier 3 management included transient increase of FiO 2 to 1.0 to allow any benefit of increasing positive endexpiratory pressure 20 or aggressively reducing ICP with increased sedation, CSF drainage, or osmotic therapy. Data Collection and Definitions In addition to demographic data, the ISS, head AIS, GCS score, SAPS II, MABP, PaO 2 to FiO 2 ratio, and body temperature were recorded at ICU admission. The first noncontrast head CT scan was reviewed, and data were abstracted for pathology and features known to be associated with outcome. 14,17,18 The Marshall CT Classification Head Injury was used to assign degrees of brain injury severity based on the following characteristics of diffuse injury patterns: I, no intracranial disease; II, 0 5 mm midline shift, no mass lesion 25 ml; III, 0 5 mm midline shift, basal cisterns compressed or absent, no mass lesion 25 ml; and IV, > 5 mm midline shift, no mass lesion 25 ml. Mass lesions were identified as traumatic intracranial abnormalities with cumulative mass 25 ml and as to whether they were surgically evacuated. Midline shift > 5 mm and basilar cistern compression were also noted. Main intracranial diagnoses included epidural hematoma, subdural hematoma, traumatic SAH, intraparenchymal hemorrhage, features of diffuse axonal injury, and depressed skull fractures. Intraventricular blood was classified as SAH. Both ICP and PbO 2 values were reported as the daily average value and the daily worst value recorded over the entire monitoring course. The type and frequency of therapeutic interventions, as described in the therapeutic injury level score, were also recorded. 15 Patients with 645

3 R. P. Martini et al. > 2 consecutive days of serum sodium above 145 meq/l were considered to have received hypertonic saline solution therapy. Hyperventilation was defined as PaCO 2 values 35 mm Hg for > 30 minutes. Mortality during hospitalization was evaluated. Scores of the FIM, on a scale from 1 to 4, were assessed at hospital discharge and included independence with feeding, locomotion, and expression. Scores for each category were added for a total maximum score of 12: 1, total dependence; 2, partial dependence; 3, independent with assistance; and 4, complete independence. Neurological outcome was assessed only in survivors. The LOS and duration of mechanical ventilation were also recorded. Cost was analyzed from the institutional perspective. Hospital charges for each patient were obtained from hospital administrative records. Charges were converted to cost by applying the institution specific cost-to-charge ratio (0.668). Dollar values for cost have been adjusted for inflation and are reported in 2007 US dollars. Statistical Analysis Continuous variables are presented as mean ± SD. Means were compared using 2-sample t-tests with unequal variances between patients in the PbO 2 /ICP monitoring group and patients in the ICP-only monitoring group. Categorical variables are presented as frequency distribution. Comparisons between groups were made using chi-square test. All p values are 2 sided. Regression analysis with robust variance estimates was used to investigate mean differences of the outcome measures. Unadjusted models that included only the main predictor of interest (that is, the type of neuromonitoring) were fitted to estimate mean differences in both monitoring groups. The estimates were then adjusted for a priori selected potential confounding factors, including age, admission GCS score, head AIS score, and Marshall CT classification. The adjusted models included type of neuromonitoring, significant predictors of death, and the confounders. The ISS and head AIS scores were included separately in the model to estimate the respective contribution of each component. Because the scores were highly collinear, we only included the head AIS score in the model to fully account for brain injury severity. Multivariable linear regression was also used to model LOS and hospital charges after appropriate transformation. Because LOS and charges were skewed, they were log transformed before fitting the regression model. To allow for sufficient time for the treatment encouraged by monitoring modality to have an effect, sensitivity analysis was used to explore associations between monitoring modality and hospital mortality rate, restricting the analysis to patients who survived beyond 48 hours after ICU admission. A commercially available statistical program was used for all analyses (STATA for Macintosh, version 10.0; StataCorp). Results Patient Demographics During the study period, 629 patients were admitted to Harborview Medical Center with a diagnosis of severe traumatic brain injury. Of these, 506 patients underwent insertion of an ICP monitor, and 123 patients underwent placement of both ICP and PbO 2 monitors. Demographic characteristics of the study population are shown in Table 1. No important differences were noted with respect to sex, race/ethnicity, body mass index, SAPS II, MABP, PaO 2 / FiO 2 ratio, or body temperature at ICU admission. On average, patients in the ICP/PbO 2 group were younger and had more severe brain injuries than patients in the ICP-only group. The median admission GCS, mean ISS, and mean head AIS scores were worse in patients in whom a PbO 2 monitor was placed. Patient Management Neuromonitoring data and therapeutic interventions received during the ICU stay stratified by study group are shown in Table 2. Patients in the PbO 2 /ICP-monitored group had the same average daily ICP values, but a greater number of episodes of intracranial hypertension, than patients in the ICP-only group. Patients who underwent PbO 2 /ICP monitoring received more aggressive management in many components of the therapeutic intensity level score, including administration of hypertonic saline, mannitol, use of vasopressors, and use of sedatives. There was no significant difference between the groups in the numbers of patients undergoing hyperventilation or surgical decompression. During the ICU stay, PbO 2 -monitored patients had a significantly lower worst PaO 2 /FiO 2 ratio compared with patients in the ICP-only group. Mortality Rate The cumulative hospital mortality rate was 29% in patients in the ICP/PbO 2 group and 23% in ICP-only group (Table 3). The crude difference in mortality rate was 6.7% (95% CI 2.1 to 15.6%). After adjustment for head AIS score, ICU admission GCS score, Marshall classification, and age, the excess mortality rate in the PbO 2 group persisted but to a smaller degree (4.4% [95% CI 3.9 to 12.7%]; Table 4). Neurological Outcome Relatively fewer patients in the ICP/PbO 2 -monitored group were discharged with a FIM score indicating some level of independence (FIM score 7) compared with the ICP-only group. The mean FIM score at discharge was 7.6 ± 3.0 in ICP/PbO 2 -monitored patients and 8.6 ± 2.8 in the ICP group (p < 0.01, Table 3). After adjustment, there was a 0.75 point difference (95% CI 1.41 to 0.09) in the mean FIM score in patients with PbO 2 monitoring compared with ICP-only monitoring. Resource Utilization Monitoring of PbO 2 was associated with increases in resource utilization compared with ICP monitoring alone (Table 3). The unadjusted difference in the mean duration of mechanical ventilation was 3.5 days (95% CI ). The adjusted mean difference in the duration of mechanical ventilation was a 2.96 days (95% CI ). The crude and adjusted ratios of median LOS and 646 J Neurosurg / Volume 111 / October 2009

4 Brain tissue oxygen monitoring guided management TABLE 1: Clinical characteristics at ICU admission in patients managed with either ICP monitoring only or combined ICP and PbO 2 monitoring* Clinical Characteristic ICP-Only median hospital cost are shown in Table 4. Following adjustment, there was a 27% relative increase in the median LOS and a 29% relative increase in the median hospital cost in the ICP/PbO 2 -monitored group compared with the ICP-only group (Table 4). J Neurosurg / Volume 111 / October 2009 PbO 2 Monitored p Value no. of patients <0.01 mean age (yrs) 40.7 ± ± 16.9 no. of males 373 (73.7) 92 (74.8) 0.81 race/ethnic group 0.65 American Indian 8 (1.6) 2 (1.6) Asian 28 (5.5) 5 (4.1) black 30 (5.9) 7 (5.7) Caucasian 382 (75.5) 101 (82.1) Hispanic 36 (7.1) 5 (4.1) unknown 22 (4.4) 3 (2.4) mean ISS score 35.2 ± ± 12.6 <0.01 mean head AIS score 4.62 ± ± 0.43 <0.01 mean GCS 5.6 ± ± mean SAPS II 51.2 ± ± mean MABP (mm Hg) 85.0 ± ± mean PaO 2 /FiO 2 ratio (mm Hg) ± ± mean temperature (ºC) 36.3 ± ± BMI (kg/m 2 ) 25.6 ± ± no. w/ initial head CT findings traumatic SAH 354 (70.0) 91 (74.0) 0.38 subdural hematoma 298 (58.9) 81 (65.9) 0.16 epidural hematoma 50 (9.9) 16 (13.0) 0.31 intraparenchymal hematoma 255 (50.4) 65 (52.9) 0.63 diffuse axonal injury 119 (23.5) 30 (24.4) 0.84 midline shift >5mm 115 (22.7) 32 (26.0) 0.44 basilar cistern compression 148 (29.3) 44 (25.8) 0.16 depressed skull fracture 65 (12.9) 14 (11.4) 0.66 Marshall classification 0.52 diffuse injury Type I 5 (1.0) 0 (0.0) diffuse injury Type II 268 (53.0) 58 (47.2) diffuse injury Type III 36 (7.1) 12 (9.8) diffuse injury Type IV 7 (1.4) 1 (0.8) evacuated mass lesion 155 (30.6) 46 (37.4) nonevacuated mass lesion 34 (6.7) 6 (4.9) * Mean values are presented ± SD. Abbreviations: BMI = body mass index; IQR = interquartile range. The Marshall CT classification was used to assign degrees of injury severity based on the following characteristics of diffuse injury patterns. See text for definition of classes. Mass lesions were identified as traumatic intracranial abnormalities with cumulative mass 25 ml and as whether they were surgically evacuated. TABLE 2: Physiological values and therapeutic interventions during the ICU stay* ICU Variable ICP-Only PbO 2 Monitored p Value physiological values mean daily ICP (mm Hg) 16.2 ± ± no. of patients w/ ICP >20 mm Hg 353 (69.8) 106 (86.2) 0.40 mean daily CPP (mm Hg) 74.8 ± ± mean daily PbO 2 (mm Hg) 24.7 ± 10.2 no. of patients w/ PbO 2 <15 mm Hg 74 (60.2) lowest mean PaO 2 /FiO 2 ratio (mm Hg) 185 ± ± 95 <0.01 mean daily serum sodium 143 ± ± 5.6 <0.01 therapeutic interventions hyperventilation 443 (87.6) 113 (92.7) 0.11 hypertonic saline 296 (58.5) 96 (78.1) <0.01 mannitol 194 (38.3) 65 (52.8) <0.01 vasopressors 152 (30.0) 62 (50.4) <0.01 midazolam 154 (30.4) 53 (43.1) <0.01 fentanyl 325 (64.2) 99 (80.5) <0.01 propofol 418 (82.6) 116 (94.3) <0.01 surgical decompression 155 (30.6) 46 (37.4) 0.15 * Mean values are presented ± SD. Hyperventilation was defined as a PaCO 2 < 35 mm Hg for > 30 minutes. Analyses that excluded patients who did not survive beyond the first 48 hours after trauma yielded results similar to those presented above. Discussion We observed no reduction in the hospital mortality rate in patients with severe TBI who were managed with ICP and PbO 2 monitoring compared with those who were managed with ICP monitoring only. In patients who survived, the likelihood of a good neurological outcome was smaller in the PbO 2 group compared with ICP-only group. The use of PbO 2 monitoring was associated with an increase in the number of therapeutic interventions, a longer course of mechanical ventilation, and an increased LOS, and thus with increased hospital costs. The observational nature of our study design allows for the possibility that the patients with and without PbO 2 monitoring had an inherently different prognosis. Indeed, the patients who underwent PbO 2 monitoring were younger and had severer brain injuries as measured by the admission GCS score and the head AIS score. However, the decision to insert a PbO 2 monitor was made at the discretion of the admitting neurosurgeon, with variability in practice among physicians. While we attempted to adjust for these imbalances in baseline characteristics predictive of prognosis in our analyses, it is possible that residual confounding is present, and our results should be interpreted with this in mind. 647

5 R. P. Martini et al. TABLE 3: Outcome at hospital discharge after ICP monitoring only or combined ICP and PbO 2 monitoring* Outcome Variable ICP-Only PbO 2 Monitored p Value no. of deaths 114 (22.5) 36 (29.3) 0.12 mean FIM score in survivors 8.6 ± ± 3.0 <0.01 functional independence 303 (77.3) 56 (64.4) 0.01 length of stay, days, mean ± SD 19.1 ± ± hospital disposition 0.46 home, no assistance 59 (11.7) 9 (7.3) home, healthcare assistance 8 (1.6) 2 (1.6) skilled nursing facility 154 (30.4) 44 (35.8) rehabilitation 159 (31.4) 31 (25.2) dead 114 (22.5) 36 (29.3) other 12 (2.4) 1 (0.81) mean hospital cost ± ± 78.1 <0.01 * Mean values are presented ± SD. Functional independence was assessed only in survivors (392 in the ICP-only group and 87 in the PbO 2 group). Survivors were considered independent at discharge if they had an FIM score 7. Hospital cost is presented in this table as ratios of the medians. In contrast to our results, Stiefel and colleagues 28 found a reduced risk of hospital death in patients with severe TBI after the initiation of PbO 2 -guided management. However, the patients who did and did not receive PbO 2 monitoring in that study were not comparable in a number of respects. Furthermore, the risk of in-hospital death after the introduction of PbO 2 monitoring (25%) was quite similar to that seen in our study of patients with TBI (irrespective of monitoring status) and in US trauma centers in general; 5 only the mortality rate (44%) in their earlier patient group differed. TABLE 4: Unadjusted and adjusted mean differences and 95% CIs for main outcome Outcome Variable Crude Difference (95% CI) Adjusted Difference (95% CI)* hospital death 6.7 ( 0.21 to 16) 4.4 ( 3.9 to 13) FIMs in survivors 1.03 ( 1.70 to 0.35) 0.75 ( 1.41 to 0.09) duration of mechan ( ) 2.96 ( ) ical ventilation LOS 1.28 ( ) 1.27 ( ) hospital cost 1.35 ( ) 1.29 ( ) * Adjusted for age, head AIS score, ICU admission GCS score, and Marshall classification of head CT scan. Functional independence was assessed only in survivors (392 in the ICP-only group and 87 in the PbO 2 group). The LOS and hospital cost are presented as unadjusted and adjusted ratios of the medians. We found a significant reduction in the odds of a better neurological outcome in patients who received PbO 2 - directed management. Although it has been previously reported that poor neurological outcome was associated with hypoxic PbO 2 values, 29,30 few studies have investigated whether interventions directed at improving PbO 2 lead to better neurological outcome. In a small study, Meixensberger et al. 19 reported no significant difference in the proportion of patients with good neurological outcome (Glasgow Outcome Scale scores of 4 5) following a PbO 2 - guided treatment protocol to improve cerebral perfusion compared with patients in whom PbO 2 monitoring was not used to guide therapy. The outcomes were not adjusted for potential confounders, and it is unknown if the results would have been different after accounting for patient characteristics. In the present study, we attempted to control for confounding variables, and we used an assessment of FIMs at hospital discharge to quantify neurological outcome. It is difficult to make a direct comparison because neurological outcome was not reported in detail in previous studies. We found that patients were more likely to have a lower level of functional independence at hospital discharge if they had undergone ICP/PbO 2 monitoring rather than ICP monitoring only. It is possible that the means of monitoring PbO 2 in the period during which our study took place, or of acting on the results, may have been less than ideal. Technical issues in human clinical use, improper positioning in a given injured brain, a very small monitored tissue volume, and device-related damage to the monitored tissue are all established complications associated with insertion of a PbO 2 monitor. 10 The treatment threshold could also be incorrect. In addition, recent work suggests that PbO 2 is more reflective of O 2 diffusion than O 2 delivery, and that PbO 2 may not be an effective ischemia monitor. 25 Nonetheless, it could also be true that the therapeutic approach to desaturations may be ineffective or even have unfavorable risk-benefit ratios with respect to overall outcomes. Many more of the patients in the PbO 2 group received therapeutic interventions to manage hypoxic PbO 2 values, each with possible adverse systemic consequences. Aggressive treatment to improve CPP is associated with increased systemic complications, 24 and many of the same treatments are indicated for the management of hypoxic PbO 2 values. For example, the use of vasopressors to raise CPP is a risk factor for pulmonary complications, 7 and pulmonary complications such as acute lung injury have been associated with worse neurological outcome following TBI. 4 A greater proportion of patients with PbO 2 -guided management received vasopressors than the ICP-only group. The use of sedatives also was more frequent in the PbO 2 group. It is well documented that the use of heavy sedation causes increased duration of mechanical ventilation and LOS. 8,13 Conclusions The data presented in this study suggest that there are reasons to question the efficacy of PbO 2 -guided management of patients with severe TBI. However, because of the potential for residual confounding in our observa- 648 J Neurosurg / Volume 111 / October 2009

6 Brain tissue oxygen monitoring guided management tional study, we recommend that a randomized trial be conducted to better address the question. Disclosure This study was partially supported by a medical student research fellowship from the Foundation for Anesthesia Education and Research and by the Harborview Anesthesiology Research Center. References 1. Brain Trauma Foundation, American Association of Neurological Surgeons Joint Section on Neurotrauma and Critical Care: Guidelines for the management of severe head injury. J Neurotrauma 13: , Bratton SL, Chestnut RM, Ghajar J, McConnell Hammond FF, Harris OA, Hartl R, et al: Guidelines for the management of severe traumatic brain injury. X. Brain oxygen monitoring and thresholds. J Neurotrauma 24 Suppl 1:S65 S70, Bratton SL, Chestnut RM, Ghajar J, McConnell Hammond FF, Harris OA, Hartl R, et al: Guidelines for the management of severe traumatic brain injury. VI. Indications for intracranial pressure monitoring. J Neurotrauma 24 Suppl 1:S37 S44, Bratton SL, Davis RL: Acute lung injury in isolated traumatic brain injury. Neurosurgery 40: , Bulger EM, Nathens AB, Rivara FP, Moore M, MacKenzie EJ, Jurkovich GJ: Management of severe head injury: institutional variations in care and effect on outcome. Crit Care Med 30: , Chesnut RM, Marshall LF, Klauber MR, Blunt BA, Baldwin N, Eisenberg HM, et al: The role of secondary brain injury in determining outcome from severe head injury. J Trauma 34: , Contant CF, Valadka AB, Gopinath SP, Hannay HJ, Robertson CS: Adult respiratory distress syndrome: a complication of induced hypertension after severe head injury. J Neurosurg 95: , Girard TD, Kress JP, Fuchs BD, Thomason JW, Schweickert WD, Pun BT, et al: Efficacy and safety of a paired sedation and ventilator weaning protocol for mechanically ventilated patients in intensive care (Awakening and Breathing Controlled trial): a randomised controlled trial. Lancet 371: , Gopinath SP, Robertson CS, Contant CF, Hayes C, Feldman Z, Narayan RK, et al: Jugular venous desaturation and outcome after head injury. J Neurol Neurosurg Psychiatry 57: , Gopinath SP, Valadka AB, Uzura M, Robertson CS: Comparison of jugular venous oxygen saturation and brain tissue PO 2 as monitors of cerebral ischemia after head injury. Crit Care Med 27: , Gupta AK, Hutchinson PJ, Fryer T, Al-Rawi PG, Parry DA, Minhas PS, et al: Measurement of brain tissue oxygenation performed using positron emission tomography scanning to validate a novel monitoring method. J Neurosurg 96: , Johnston AJ, Steiner LA, Coles JP, Chatfield DA, Fryer TD, Smielewski P, et al: Effect of cerebral perfusion pressure augmentation on regional oxygenation and metabolism after head injury. Crit Care Med 33: , Kress JP, Pohlman AS, O Connor MF, Hall JB: Daily interruption of sedative infusions in critically ill patients undergoing mechanical ventilation. N Engl J Med 342: , Maas AI, Hukkelhoven CW, Marshall LF, Steyerberg EW: Prediction of outcome in traumatic brain injury with computed tomographic characteristics: a comparison between the computed tomographic classification and combinations of computed tomographic predictors. Neurosurgery 57: , Marion DW, Penrod LE, Kelsey SF, Obrist WD, Kochanek PM, Palmer AM, et al: Treatment of traumatic brain injury with moderate hypothermia. N Engl J Med 336: , Marmarou A, Ward JD, Young HF, Anderson RL, Choi SC, Eisenberg HM, et al: Impact of ICP instability and hypotension on outcome in patients with severe head trauma. J Neurosurg 75:S59 S66, Marshall LF, Marshall SB, Klauber MR, Van Berkum Clark M, Eisenberg H, Jane JA, et al: The diagnosis of head injury requires a classification based on computed axial tomography. J Neurotrauma 9 (1 Suppl):S287 S292, Marshall LF, Marshall SB, Klauber MR, Van Berkum Clark M, Eisenberg HM, Jane JA, et al: A new classification of head injury based on computerized tomography. J Neurosurg 75: S14 S20, Meixensberger J, Jaeger M, Vath A, Dings J, Kunze E, Roosen K: Brain tissue oxygen guided treatment supplementing ICP/ CPP therapy after traumatic brain injury. J Neurol Neurosurg Psychiatry 74: , Menzel M, Doppenberg EM, Zauner A, Soukup J, Reinert MM, Bullock R: Increased inspired oxygen concentration as a factor in improved brain tissue oxygenation and tissue lactate levels after severe human head injury. J Neurosurg 91:1 10, Mueller BA, Rivara FP, Lii SM, Weiss NS: Environmental factors and the risk for childhood pedestrian-motor vehicle collision occurrence. Am J Epidemiol 132: , Narayan RK, Greenberg RP, Miller JD, Enas GG, Choi SC, Kishore PR, et al: Improved confidence of outcome prediction in severe head injury. A comparative analysis of the clinical examination, multimodality evoked potentials, CT scanning, and intracranial pressure. J Neurosurg 54: , Narayan RK, Kishore PR, Becker DP, Ward JD, Enas GG, Greenberg RP, et al: Intracranial pressure: to monitor or not to monitor? A review of our experience with severe head injury. J Neurosurg 56: , Robertson CS, Valadka AB, Hannay HJ, Contant CF, Gopinath SP, Cormio M, et al: Prevention of secondary ischemic insults after severe head injury. Crit Care Med 27: , Rosenthal G, Hemphill JC III, Sorani M, Martin C, Morabito D, Obrist WD, et al: Brain tissue oxygen tension is more indicative of oxygen diffusion than oxygen delivery and metabolism in patients with traumatic brain injury. Crit Care Med 36: , Servadei F, Antonelli V, Giuliani G, Fainardi E, Chieregato A, Targa L: Evolving lesions in traumatic subarachnoid hemorrhage: prospective study of 110 patients with emphasis on the role of ICP monitoring. Acta Neurochir Suppl 81:81 82, Smith MJ, Stiefel MF, Magge S, Frangos S, Bloom S, Gracias V, et al: Packed red blood cell transfusion increases local cerebral oxygenation. Crit Care Med 33: , Stiefel MF, Spiotta A, Gracias VH, Garuffe AM, Guillamondegui O, Maloney-Wilensky E, et al: Reduced mortality rate in patients with severe traumatic brain injury treated with brain tissue oxygen monitoring. J Neurosurg 103: , Valadka AB, Gopinath SP, Contant CF, Uzura M, Robertson CS: Relationship of brain tissue PO 2 to outcome after severe head injury. Crit Care Med 26: , van den Brink WA, van Santbrink H, Steyerberg EW, Avezaat CJ, Suazo JA, Hogesteeger C, et al: Brain oxygen tension in severe head injury. Neurosurgery 46: , 2000 Manuscript submitted August 8, Accepted February 16, Please include this information when citing this paper: published online April 24, 2009; DOI: / JNS Address correspondence to: Ross Martini, B.S., Brown University, Box G-8269, Providence, Rhode Island ross_ martini@brown.edu. J Neurosurg / Volume 111 / October