The American Journal of Surgery (2014) 208, 1071-1077
Southwestern Surgical Congress
The impact of early flow and brain oxygen crisis on the outcome of patients with severe traumatic brain injury Corrado P. Marini, M.D., F.A.C.S.a,*, Christy Stoller, M.D.a, Omar Shah, M.D.b, Antoni Policastro, M.D.a, Gary Lombardo, M.D.a, Juan A. Asensio, M.D.a, Yin C. Hu, M.D.c, Michael F. Stiefel, M.D., Ph.D.c a
Division of Trauma Surgery, Surgical Critical Care and Acute Care Surgery, Department of Surgery, Division of Neurology, Department of Medicine, cDepartment of Neurosurgery, New York Medical College, Westchester Medical Center University Hospital, 100 Woods Rd Taylor Pavilion E 138, Valhalla, NY 10595, USA
b
KEYWORDS: Brain flow; Oxygen crisis; CRASH model; TBI; Multimodality monitoring
Abstract BACKGROUND: Multimodality monitoring and goal-directed therapy may not prevent blood flow and brain oxygen (Flow/BrOx) crisis. We sought to determine the impact of these events on outcome in patients with severe traumatic brain injury (sTBI). METHODS: Twenty-four patients with sTBI were treated to maintain intracranial pressure (ICP) less than or equal to 20 mm Hg, cerebral perfusion pressure (CPP) greater than or equal to 60 mm Hg, brain oxygen greater than or equal to 20 mm Hg, and near infrared spectroscopy greater than or equal to 60%. Flow/BrOx crisis events were recorded. The 14-day predicted mortality was compared with actual mortality. RESULTS: Nonsurvivors had a significantly higher number of crisis events nonresponsive to treatment (P , .05). Mortality was 87.5% in patients with greater than or equal to 20 events versus 6.3% in patients with less than 20 events. The predicted mortality was 58%, whereas actual mortality was 33.3% (8/24), yielding a 42% reduction in mortality. CONCLUSIONS: A multimodality monitoring and goal-directed therapy may decrease mortality in sTBI. However, Flow/BrOx crisis events still occur and predict a poor outcome. Ó 2014 Elsevier Inc. All rights reserved.
Each year approximately 1.5 million people sustain traumatic brain injury (TBI), 270,000 require hospitalization, and 52,000 die.1,2 Patients with severe TBI (sTBI) are often treated using the Brain Trauma Foundation guidelines that include a monitoring protocol with described methods of interventions.3 The authors declare no conflicts of interest. * Corresponding author. Tel.: 11-914-493-5213; fax: 914-493-5271. E-mail address: [email protected] Manuscript received March 27, 2014; revised manuscript July 29, 2014 0002-9610/$ - see front matter Ó 2014 Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.amjsurg.2014.08.003
There is growing evidence that during the first 48 hours following sTBI, the recommended systolic blood pressure (SBP) threshold of greater than 90 mm Hg is associated with episodes of brain hypoperfusion. The cumulative effect of brain hypoperfusion may lead to increased mortality and a worse functional outcome. As a result, a higher SBP may be necessary to prevent secondary brain injury.4 However, in the absence of secondary monitoring techniques such as brain oxygen tension (PbtO2), near infrared spectroscopy (NIRS) oxygen saturation, or cerebral microdialysis (CMD), it is unclear whether a higher SBP is beneficial. Theoretically, an
1072
The American Journal of Surgery, Vol 208, No 6, December 2014
Figure 1
Multimodality monitoring and goal-directed therapy protocol.
increase in CPP may provide the brain with increased oxygen tension and as a result may prevent secondary brain ischemia. To date, the correlation between cerebral blood flow (CBF) obtained by surrogate variables such as SBP, CPP, and NIRS oxygen saturation and the critical threshold of PbtO2 remains unclear. It is within this context that we sought to determine the relationship between early flow and brain oxygen (Flow/BrOx) crisis events on the outcome of patients with sTBI.
Methods From July 2011 to September 2012, 24 patients with sTBI were treated with a multimodality monitoring and
goal-directed therapy (MM&GDTP) for 5 days (Fig. 1). Multimodality monitoring included monitoring of ICP, CPP, PbtO2 (Licox; Integra Life Science, Plainsboro, NJ), NIRS (Covidien, Mansfield, MA), and minute ventilation (Ve). The 5-day targeted therapy protocol included normothermia (37 C) with dry water immersion (Arctic Sun, Medivance, Inc. Luoisville, CO), PbtO2 greater than or equal to 20 mm Hg, ICP less than or equal to 20 mm Hg, CPP greater than or equal to 60 mm Hg, NIRS greater than or equal to 60%, and nutritional support targeted to a respiratory quotient, measured by indirect calorimetry, of .83 by day 3 and positive nitrogen balance (NB) by day 7 with initiation of peptide-based enteral nutrition on completion of the resuscitation phase. All patients were sedated to achieve synchrony with the ventilator, avoidance
C.P. Marini et al.
Early flow and brain oxygen crisis events in sTBI
of cough and a modified Ramsey score of 2 with a combination of midazolam infusion and propofol. Osmotherapy included 3% saline and the addition of mannitol when appropriate. Burst suppression (BS) was initiated when ICP, low PbtO2, or low NIRS was not responsive to therapy. BS was achieved with an infusion of midazolam at a dose up to 15 mg/hour and propofol at a dose up to 100 mg/kg/ minute titrated to 1 to 2 burst by continuous electro-encephalogram monitoring. Decompressive hemicraniectomy (DC) was employed in patients with surgical lesions causing mass effect and increased ICP. The 14-day mortality was predicted by the web-based computed tomography (CT) corticosteroid randomization after significant head injury (CrasH) model and by the Marshall CT classification. In brief, CrasH uses a combination of 4 demographic and clinical variables, namely age, Glasgow Outcome Scale (GCS), pupil reactivity and the presence of major extracranial injuries and the CT presence of petechial hemorrhages, obliteration of the third ventricle or basal cisterns, subarachnoid bleeding, midline shift and nonevacuated hematoma to predict mortality.5
Statistical analysis Data acquired included demographics, Injury Severity Score, AIS (H), Marshall CT classification, Flow/BrOx crisis events, and mortality. Parametric data were analyzed with Student t test, whereas nonparametric data were analyzed with chi-square or Fisher’s exact tests. The predicted mortality (PM) by the CrasH model and by the Marshall CT classifcation was compared with the actual mortality of the group. Flow/BrOx crisis was defined by the presence of any one of the following for a period of 5 minutes or longer: (1) CPP less than 60 mm Hg and NIRS less than 60% (Flow/BrOx1); (2) PbtO2 less than 20 mm Hg and NIRS less than 60% (Flow/BrOx-2); (3) Flow/BrOx-3 (Flow/BrOx-1 1 Flow/ BrOx-2); and (4) simultaneous CPP less than 60 mm Hg and PbtO2 less than 20 mm Hg and NIRS less than 60% (Flow/BrOx-4). Patients were analyzed with respect to survival status to identify differences predictive of outcome. Functional outcome at discharge was assessed with the Table 1
1073
Glasgow Outcome Scale (GOS), whereas functional outcome at 6 and 12 months was assessed with extended GOS (GOSE). The GOSE of survivors (SURV) was obtained by telephone interviews by one of the authors. GOSE score of 1 to 4 defined a poor functional outcome, whereas a score of 5 to 8 defined a good functional outcome. Binomial logistic regression analysis with mortality as the dependent variable was used to identify variables predictive of outcome. Data are presented as means 6 standard deviation, median and interquartile range (IQR), and as proportions. Statistical significance was accepted to correspond to a P value of less than .05. The study was approved by the institutional review board.
Results Twenty-four patients with sTBI were treated with the MM&GDTP for 5 days (Table 1). Two patients with ICP who remained refractory to osmotherapy and developed surgical lesions underwent DC within 24 hours of initiation of the treatment protocol and both the patients survived. In 22 of the 24 patients, enteral feeding was started within 24 hours of injury on completion of the resuscitation phase that included restoration of normovolemia, acid–base balance, and appropriate Ve. The 2 patients who underwent DC had initiation of enteral feeding following the procedure. All 24 patients had IC on day 3 and measurement of NB by day 7. The PM for the group was 58%, whereas actual mortality was 33.3% (8/24), yielding a 42% reduction in mortality. The cause of death of the 8 patients included the following: 2 from progression of sTBI to brain death on day 5, 1 each from a brainstem infarct, myocardial infarction, pulmonary embolism, sepsis-induced multiple organ dysfunction syndrome, and two from withdrawal of care. There was no difference between SURV and nonsurvivors (NS) with respect to GCS, Abbreviated Injury Score (Head) (AIS [H]), Injury Severity Score, and Marshall scores; however, they differed with respect to age and PM (Table 1). Shown in Table 2 is the timeline of ICP, CCP, and PbtO2 stratified by survival status for the 24 patients. NS had a trend toward higher ICP and lower CPP and
Demographic injury severity and Flow/BrOx crisis events
Variable
SURV (n 5 16)
NS (n 5 8)
P value
Age GCS (median with IQR) ISS AIS (H) CrasH 14-day mortality Marshall PM score Flow/BrOx-1 Flow/BrOx-2 Flow/BrOx-3 Flow/BrOx-4
41 6 20 3.5; 3 45 6 18 4.4 6 .9 45 6 28 31 6 20 567 12 6 15 17 6 14 365
62 6 21 3.5; 3.25 37 6 8 5.0 6 .9 83 6 17 37 6 14 19 6 14 33 6 24 54 6 30 45 6 29
.02 1 .18 .1 .005 .45 .002 .01 .001 .001
AIS (H) 5 Abbreviated Injury Score (Head); CrasH 5 corticosteroid randomization after significant head injury; Flow/BrOx 5 blood flow and brain oxygen; GCS 5 Glasgow Outcome Scale; IQR 5 interquartile range; ISS 5 Injury Severity Score; NS 5 nonsurvivors; SURV 5 survivors.
1074 Table 2
The American Journal of Surgery, Vol 208, No 6, December 2014 Timeline of flow and oxygenation variables during the 5-day MM&GDTP CPP
Day
SURV
1 2 3 4 5
71 72 77 77 74
6 6 6 6 6
ICP NS 11 8 8 6 8
63 66 67 76 74
PbtO2
SURV 6 6 6 6 6
14 15 9 15 16
13 14 11 11 10
6 6 6 6 6
NS 8 7 4 4 5
18 21 18 10 12
SURV 6 6 6 6 6
12 15 12 6 9
29 33 32 31 30
6 6 6 6 6
NS 16 11 10 10 10
19 25 27 25 30
6 6 6 6 6
18 14 12 14 9
CPP 5 Cerebral perfusion pressure; ICP 5 intracranial pressure; MM&GDTP 5 multimodality monitoring and goal-directed therapy; NS 5 nonsurvivors; PbtO2 5 brain oxygen tension; SURV 5 survivors.
PbtO2 during the first 2 days following their injury; however, this did not reach statistical significance when compared with SURV.
Flow/BrOx crisis events NS had significantly more Flow/BrOx crisis events than SURV (Table 1). The number of Flow/BrOx-4 crisis events was statistically greater in NS vs SURV (45 6 29 vs 3 6 5, respectively; P , .05). There was a significantly greater number of Flow/BrOx-4 crisis events during the first 3 days of treatment in NS (Fig. 2); no significant difference was observed on days 4 and 5. The analysis of Flow/BrOx-4 showed that 7 of the 8 (87.5%) patients with greater than or equal to 20 episodes died as opposed to only 1 of the 16 (6.3%) patients in the group with less than 20 episodes. Only 1 patient without a single Flow/BrOx-4 crisis event died as opposed to only 1 of the 16 patients having 22 episodes of Flow/BrOx-4 crisis events surviving. There was a statistically significant difference in the mean number of minutes of Flow/BrOx-4 crisis events between SURV and NS (15 6 25 vs 225 6 145 minutes, respectively; P , .05). Despite the MM&GDTP approach to the treatment of the patients, Flow/BrOx events occurred more commonly during the night shifts. The presence of Flow/BrOx-4 crisis events was the only variable predictive of poor outcome by binary stepwise logistic regression with backward and forward elimination. The median Glasgow Outcome Score on hospital discharge was 3.0 with 1.5 IQR for the 16 surviving patients. Functional outcome data at 6 and 12 months was available for 6 patients. Four of the 6 patients had good functional outcome, and median GOSE was 5.0 with 1.5 IQR and 5.5 with 4 IQR at the 6- and 12-month follow-up evaluations, respectively. However, the 2 patients who were discharged with a GOS of 2 showed little improvement and had a GOSE of 3 at 6- and 12-month follow-up evaluations.
recent literature review suggests that the use of an ICP monitor and adherence to guidelines for the management of patients with sTBI are associated with improved outcomes.9 However, to date there is no Level I evidence demonstrating that ICP monitoring is more effective in managing sTBI. Furthermore, CMD data suggest that cellular injury from hypoxia and/or metabolic mitochondrial dysfunction may occur even when CBF is normal as a result of normal ICP and CPP.10 Brain oxygen monitoring was added into the management guidelines for sTBI in 2007.11 Several studies have shown that compromised PbtO2 (,20 mm Hg) is associated with worse outcome in patients with sTBI. As a result, ICP– CPP with PbtO2-targeted therapy has become a common treatment paradigm.12–15 The purpose of this study was to determine the frequency and the impact of early Flow/ BrOx crisis events on the outcome of patients with sTBI when an MM&GDTP was used to guide and maintain hemodynamic and intracranial resuscitation.
MM&GDTP rationale The selection of the end points of our MM&GDTP was based on physiologic and pathophysiologic principles applicable to patients with sTBI. Current opinion is that hyperthermia after TBI is associated with a worse outcome.16 We elected to target temperature control with normothermia rather than hypothermia. Normothermia
Comments Current treatment guidelines for patients with sTBI emphasize the prevention of secondary brain injury for at least the first 48 hours after injury. Therapy is focused on maintaining adequate SBP, ICP, CPP, and oxygenation.6-8 A
Figure 2
Time line of Flow/BrOx-4 crisis events.
C.P. Marini et al.
Early flow and brain oxygen crisis events in sTBI
control with the dry water immersion technique permits a higher water temperature than a hypothermic end point and in theory would be associated with fewer episodes of shivering and less detrimental effect on brain oxygen consumption.17 The target CPP greater than or equal to 60 mm Hg with a PbtO2 greater than 20 mm Hg with an NIRS greater than or equal to 60% incorporates all aspects of the individual monitor end points into one goal. Incorporation of NIRS assumed that a saturation of less than 60% may indicate suboptimal CBF because increased oxygen extraction typically is a compensatory response required to maintain consumption constant in the setting of decreased oxygen delivery.18 We postulated that an increased oxygen extraction is associated with an imbalance between oxygen supply and demand in the injured brain despite adequate CPP and PbtO2. The targeted nutritional approach was based on recent evidence showing that aggressive nutritional support in patients with sTBI provides a survival advantage.19 The incorporation of IC at day 3 provided guidance to the achievement of a positive NB by day 7 of treatment. The decision to treat patients with the sTBI 5-day protocol was based on the physical evaluation of the patient following the initial resuscitation and review of the CT scan findings. The duration of the treatment was based on evidence suggesting that CMD L/P ratios typically do not normalize for at least 3 days following sTBI as well as the 5-day US Food and Drug Administration-approved use of many of the intracranial monitoring devices.20 One consistent observation in our patients was the fact that a SBP targeted to 120 mm Hg could not assure the necessary CPP required to maintain the PbtO2 and NIRS above our established critical thresholds. Our findings on the relationship between SBP CPP and brain oxygenation contrast with the results of a recent paper suggesting that a SBP of 120 mm Hg may be may efficacious in minimizing secondary brain injury in TBI patients.4 A fundamental difference between our study and that by Brenner et al4 is the use of invasive intracranial monitoring techniques. Most of our patients required a SBP between 140 and 160 mm Hg to reach the flow and oxygenation end points of our targeted therapy. The use of norepinephrine to raise SBP to 140 to 160 mm Hg was necessary in 82% of the patients. The increased SBP was rarely associated with a concomitant increase in ICP, indicating intact global autoregulation. Although CPP, ICP, and PbtO2 were within the therapeutic targets in both SURV and NS for the majority of the treatment protocol period, Flow/BrOx crisis events still occurred and were more common in NS as opposed to SURV. The most important finding of the analysis of the Flow/BrOx-4 crisis events pertains to the conclusion that sustained simultaneous decrease in CPP, PbtO2 with NIRS less than 60% for longer than 1 hour in aggregate over the first 3 days are detrimental to the outcome of patients with sTBI. However, in the absence of CMD data correlating Flow/BrOx-4 events with metabolism, we must rely on the binary logistic regression
1075
analysis with mortality as the dependent variable that isolated the number of Flow/BrOx-4 events as predictive of increased mortality. The implementation of our monitoring and resuscitation protocol suggests a lower mortality than that predicted by the CrasH model. The small sample size of this study, however, does not allow a concrete conclusion regarding the impact of an MM&GDTP on the mortality of patients with sTBI. Interestingly, the analysis of the causes of death showed that only 3 of the 8 (37.5%) patients died from progression on their brain injury with the remaining patients dying from non– sTBI-related etiologies. The patient who died from myocardial infarction had the event on day 12, several days after completion of MM&DGTP; therefore, it is unlikely to be related to the use of MM&GDTP. Age provides a significant contribution to the CrasH model. The higher PM for NS in our cohort can be accounted for by the older age of this group. This observation is further supported by the absence of a statistically significant difference in the Marshall score between SURV and NS.
Conclusions Based on the results of our study, we conclude that Flow/ BrOx crisis events predict a poor outcome in patients with sTBI. We believe that further studies are needed to identify physiologic parameters and therapeutic measures that can prevent and/or effectively treat these events. Moreover, a further understanding on the effect of these crisis events on cerebral metabolism may better define the most effective treatment strategies.
Limitations of the study Our study does have limitations. First, CPP and NIRS data were extracted from nursing intensive care unit flow sheets leaving room for transcription error. Second limitation is the temporal distribution of the crisis events, which were more likely to occur at night. We cannot ascertain whether the increased number of crisis events in NS was secondary to the pathophysiology of the brain injury or a less intense adherence to our treatment protocol. A third limitation is with respect to the functional outcome because data on GOSE at 6 and 12 months are limited to only 6 of the 16 patients because of loss to follow-up of the patients. The final limitation is the limited application of the stepwise logistic regression analysis to a model containing only 2 variables because of the number of patients in the study.
References 1. Thurman DJ, Alverson C, Dunn KA, et al. Traumatic brain injury in the United States: a public health perspective. J Head Trauma Rehabil 1999;14:602–15.
1076
The American Journal of Surgery, Vol 208, No 6, December 2014
2. Rutland-Brown W, Langlois JA, Thomas KE, et al. Incidence of traumatic brain injury in the United States, 2003. J Head Trauma Rehabil 2006;21:544–8. 3. Brain Trauma Foundation, American Association of Neurological Surgeons, Congress of Neurological Surgeons. Guidelines for the management of severe traumatic brain injury. J Neurotrauma 2007; 24(Suppl 1):S1–106. 4. Brenner M, Stein DM, Hu PF, et al. Traditional systolic blood pressure targets underestimate hypotension-induced secondary brain injury. J Trauma 2012;72:1135–9. 5. MRC CRASH Trial Collaborators. Predicting outcome after traumatic brain injury: practical prognostic models based on large cohort of international patients. BMJ 2008;336:425–9. 6. Stein SC, Georgoff P, Meghan S, et al. 150 years of treating severe traumatic brain injury: a systematic review of progress in mortality. J Neurotrauma 2010;27:1343–53. 7. Stiefel MF, Udouteuk JD, Spiotta AM, et al. Conventional neurocritical care and cerebral oxygenation after traumatic brain injury. J Neurosurg 2006;105:568–75. 8. Fletcher JJ, Bergman K, Bolstein PA, et al. Fluid balance, complications, and brain tissue oxygen tension monitoring following severe traumatic brain injury. Neurocrit Care 2012;17:49–57. 9. Stein SC, Georgoff P, Meghan S, et al. Relationship of aggressive monitoring and treatment to improved outcomes in severe traumatic brain injury. J Neurosurg 2010;112:1105–12. 10. Vespa P, Bergsneider M, Hattori N, et al. Metabolic crisis without brain ischemia is common after traumatic brain injury: a combined microdialysis, positron emission tomography study. J Cereb Blood Flow Metab 2005;25:763–74. 11. Bratton SL, Chestnut RM, Ghajar J, et al. Guidelines for the management of severe traumatic brain injury. X. Brain oxygen monitoring and thresholds. J Neurotrauma 2007;24(Suppl 1):S65–70. 12. Bardt TF, Unterberg AW, Hartl R, et al. Monitoring of brain tissue PO2 in traumatic brain injury: effect of cerebral hypoxia on outcome. Acta Neurochir Suppl 1998;71:153–6. 13. Chang JJ, Youn TS, Benson D, et al. Physiologic, functional outcome correlates of brain tissue hypoxia in traumatic brain injury. Crit Care Med 2009;37:283–90. 14. Valadka AB, Gopinath SP, Contant CF, et al. Relationship of brain tissue PO2 to outcome after severe head injury. Crit Care Med 1998;26: 1576–81. 15. van den Brink WA, van Santbrink H, Steyerberg EW, et al. Brain oxygen tension in severe head injury. Neurosurgery 2000;46:868–76. 16. Greer DM, Funk SE, Reaven NL, et al. Impact of fever on outcome in patients with stroke and neurological injury. Stroke 2008;39:3029–35. 17. Childs C, Wieloch T, Lecky F, et al. Report of a consensus meeting on human brain temperature after severe traumatic brain injury: its measurement and management during pyrexia. Front Neurol 2010;146: 1–8. 18. Siegal J, Kohli C. Correlation of noninvasive cerebral oximetry with cerebral perfusion in the severe head injured patient: a pilot study. J Trauma 2002;52:40–6. 19. Hartl R, Gerber LM, Ni Q, et al. Effect of early nutrition on deaths due to severe traumatic brain injury. J Neurosurg 2008;109:50–6. 20. Timofeev I, Carpenter KLH, Nortje J, et al. Cerebral extracellular chemistry and outcome following traumatic brain injury: a microdialysis study of 223 patients. Brain 2011;1:1–11.
Discussion Discussant: Dr Fredric Michael Pieracci (Denver, CO). The authors have reported a very nice study, documenting the incidence of crisis events in severe TBI patients and their association with mortality. I have three questions.
First, as a trauma surgeon in a county hospital, the first thing that occurred to me when I was reading the manuscript was how much does the multimodality therapy cost? In our ICU, we follow severe TBI patients with clinical evaluation, serial imaging, and selective ICP monitoring. Do you think that the routine use of the multimodality therapy is cost effective, or do you think that we should be targeting specific subgroups of patients to contain costs? Second, you compared your actual mortality to the predicted mortality based on a calculated score. I wonder if you could really invoke the multimodality therapy specifically as the reason for your decreased mortality, especially since, according to the manuscript, most of the deaths were not related to TBI. So possibly you had just gotten better at providing overall ICU care since you implemented the therapy. Have you thought about doing a pre/post analysis prior to when you implemented the multimodality therapy to see if there was a difference in mortality there? Finally, the nonsurvivors, if my calculations are correct, spent a lot of time in crisis, over the first three days, on average, about two to three hours. I just wanted to get a sense of why these patients who were obviously failing the goaldirected therapy weren’t considered for early decompressive craniotomy. Were their lesions not amenable on CAT scan, or do you not believe that that works, or were they so far gone that you just decided to pursue limited care? Dr Corrado Marini: Let’s start with costs. The daily cost varies between 1,000 and 1,200 dollars since we have added CMD to our multimodality and goal-directed therapy protocol. I cannot stand here and tell you at this time that our multimodality monitoring and goaldirected therapy protocol should be considered the standard of care for patients with severe traumatic brain injury. However, following this study, we have continued to treat over 100 patients with our multimodality monitoring and goal-directed therapy protocol, and we have a survival advantage at this point compared to that predicted by the Rotterdam and Marshall scores of about 35%. And that has been sustainable. However, I do not believe that what we are doing in our trauma intensive care unit, where we have full control of the neurotrauma patients, is generalizable at this time. So I will not conclude here that what we are doing is the standard of care. With respect to the pre and post implementation outcome analysis of our multimodality monitoring and goal-directed therapy protocol, I have conducted that analysis. I arrived at the institution three years ago, and I have had access to the outcome data of patients with severe TBI before my arrival. Before the implementation of our new protocol, patients with severe TBI had only ICP monitoring and craniotomy when indicated. The implementation of the multimodality monitoring and goaldirected therapy protocol has reduced mortality by 25% compared to our historic control. Whether this is the result of a Hawthorne effect, I cannot say.
C.P. Marini et al.
Early flow and brain oxygen crisis events in sTBI
With respect to the question regarding the timing and use of decompressive craniectomy, I can tell you that decompressive craniectomy is not done unless there is a mass effect that explains the increased ICP, which is not controllable with osmotherapy and BS therapy. So, in conclusion, I cannot claim that what we are doing at this time should be generalized. Another important point regards the accuracy of the predicted mortality by the Crash model; the model has good discrimination, but it does not have a great calibration. Therefore, we are now looking at our outcome based on Marshall and Rotterdam scores that tend to be more conservative with respect to predicted mortality. Dr Glenn Ihde (Red Oak, TX). I think it’s a great study in that it shows us an opportunity to improve mortality. Perhaps a window there. My question is, did you keep track of time to initiation of therapy? Because about 15 years ago, there was a study presented here showing that ICP monitors didn’t affect outcome. But having worked at the facility, I knew that the ICP monitors were not placed until maybe 24 or 36 hours into the patient’s stay.
1077
Dr Corrado Marini: First, I want to go back to the question by Dr. Pieracci regarding the number of minutes that patients spent in crisis that I forgot to answer previously. When we analyzed the timeline of the crisis events, we found that despite the use of a standard multimodality and goal-directed therapy protocol, the majority of the events occurred during the night shift. We are now trying to embed a neural network methodology to avoid the variability of the night shift versus the day shift. With respect to the question of the time to the implementation of the protocol, once the trauma surgeons and neurosurgeons have reviewed the CT scan and identified the morphology of the injury, typically within 1 to 2 hours of injury, a decision is made about whether or not the protocol should be started. So unless the patient has a radiological injury that suggests that we should avoid sedation and do a clinical reevaluation, we implement the protocol. The average time to implementation of our multimodality monitoring and goal-directed therapy protocol is less than three hours. Whether or not that makes a difference in preventing secondary ischemic injury, I cannot tell you. But our protocol is initiated very rapidly.
Southwestern Surgical Congress
The impact of early flow and brain oxygen crisis on the outcome of patients with severe traumatic brain injury Corrado P. Marini, M.D., F.A.C.S.a,*, Christy Stoller, M.D.a, Omar Shah, M.D.b, Antoni Policastro, M.D.a, Gary Lombardo, M.D.a, Juan A. Asensio, M.D.a, Yin C. Hu, M.D.c, Michael F. Stiefel, M.D., Ph.D.c a
Division of Trauma Surgery, Surgical Critical Care and Acute Care Surgery, Department of Surgery, Division of Neurology, Department of Medicine, cDepartment of Neurosurgery, New York Medical College, Westchester Medical Center University Hospital, 100 Woods Rd Taylor Pavilion E 138, Valhalla, NY 10595, USA
b
KEYWORDS: Brain flow; Oxygen crisis; CRASH model; TBI; Multimodality monitoring
Abstract BACKGROUND: Multimodality monitoring and goal-directed therapy may not prevent blood flow and brain oxygen (Flow/BrOx) crisis. We sought to determine the impact of these events on outcome in patients with severe traumatic brain injury (sTBI). METHODS: Twenty-four patients with sTBI were treated to maintain intracranial pressure (ICP) less than or equal to 20 mm Hg, cerebral perfusion pressure (CPP) greater than or equal to 60 mm Hg, brain oxygen greater than or equal to 20 mm Hg, and near infrared spectroscopy greater than or equal to 60%. Flow/BrOx crisis events were recorded. The 14-day predicted mortality was compared with actual mortality. RESULTS: Nonsurvivors had a significantly higher number of crisis events nonresponsive to treatment (P , .05). Mortality was 87.5% in patients with greater than or equal to 20 events versus 6.3% in patients with less than 20 events. The predicted mortality was 58%, whereas actual mortality was 33.3% (8/24), yielding a 42% reduction in mortality. CONCLUSIONS: A multimodality monitoring and goal-directed therapy may decrease mortality in sTBI. However, Flow/BrOx crisis events still occur and predict a poor outcome. Ó 2014 Elsevier Inc. All rights reserved.
Each year approximately 1.5 million people sustain traumatic brain injury (TBI), 270,000 require hospitalization, and 52,000 die.1,2 Patients with severe TBI (sTBI) are often treated using the Brain Trauma Foundation guidelines that include a monitoring protocol with described methods of interventions.3 The authors declare no conflicts of interest. * Corresponding author. Tel.: 11-914-493-5213; fax: 914-493-5271. E-mail address: [email protected] Manuscript received March 27, 2014; revised manuscript July 29, 2014 0002-9610/$ - see front matter Ó 2014 Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.amjsurg.2014.08.003
There is growing evidence that during the first 48 hours following sTBI, the recommended systolic blood pressure (SBP) threshold of greater than 90 mm Hg is associated with episodes of brain hypoperfusion. The cumulative effect of brain hypoperfusion may lead to increased mortality and a worse functional outcome. As a result, a higher SBP may be necessary to prevent secondary brain injury.4 However, in the absence of secondary monitoring techniques such as brain oxygen tension (PbtO2), near infrared spectroscopy (NIRS) oxygen saturation, or cerebral microdialysis (CMD), it is unclear whether a higher SBP is beneficial. Theoretically, an
1072
The American Journal of Surgery, Vol 208, No 6, December 2014
Figure 1
Multimodality monitoring and goal-directed therapy protocol.
increase in CPP may provide the brain with increased oxygen tension and as a result may prevent secondary brain ischemia. To date, the correlation between cerebral blood flow (CBF) obtained by surrogate variables such as SBP, CPP, and NIRS oxygen saturation and the critical threshold of PbtO2 remains unclear. It is within this context that we sought to determine the relationship between early flow and brain oxygen (Flow/BrOx) crisis events on the outcome of patients with sTBI.
Methods From July 2011 to September 2012, 24 patients with sTBI were treated with a multimodality monitoring and
goal-directed therapy (MM&GDTP) for 5 days (Fig. 1). Multimodality monitoring included monitoring of ICP, CPP, PbtO2 (Licox; Integra Life Science, Plainsboro, NJ), NIRS (Covidien, Mansfield, MA), and minute ventilation (Ve). The 5-day targeted therapy protocol included normothermia (37 C) with dry water immersion (Arctic Sun, Medivance, Inc. Luoisville, CO), PbtO2 greater than or equal to 20 mm Hg, ICP less than or equal to 20 mm Hg, CPP greater than or equal to 60 mm Hg, NIRS greater than or equal to 60%, and nutritional support targeted to a respiratory quotient, measured by indirect calorimetry, of .83 by day 3 and positive nitrogen balance (NB) by day 7 with initiation of peptide-based enteral nutrition on completion of the resuscitation phase. All patients were sedated to achieve synchrony with the ventilator, avoidance
C.P. Marini et al.
Early flow and brain oxygen crisis events in sTBI
of cough and a modified Ramsey score of 2 with a combination of midazolam infusion and propofol. Osmotherapy included 3% saline and the addition of mannitol when appropriate. Burst suppression (BS) was initiated when ICP, low PbtO2, or low NIRS was not responsive to therapy. BS was achieved with an infusion of midazolam at a dose up to 15 mg/hour and propofol at a dose up to 100 mg/kg/ minute titrated to 1 to 2 burst by continuous electro-encephalogram monitoring. Decompressive hemicraniectomy (DC) was employed in patients with surgical lesions causing mass effect and increased ICP. The 14-day mortality was predicted by the web-based computed tomography (CT) corticosteroid randomization after significant head injury (CrasH) model and by the Marshall CT classification. In brief, CrasH uses a combination of 4 demographic and clinical variables, namely age, Glasgow Outcome Scale (GCS), pupil reactivity and the presence of major extracranial injuries and the CT presence of petechial hemorrhages, obliteration of the third ventricle or basal cisterns, subarachnoid bleeding, midline shift and nonevacuated hematoma to predict mortality.5
Statistical analysis Data acquired included demographics, Injury Severity Score, AIS (H), Marshall CT classification, Flow/BrOx crisis events, and mortality. Parametric data were analyzed with Student t test, whereas nonparametric data were analyzed with chi-square or Fisher’s exact tests. The predicted mortality (PM) by the CrasH model and by the Marshall CT classifcation was compared with the actual mortality of the group. Flow/BrOx crisis was defined by the presence of any one of the following for a period of 5 minutes or longer: (1) CPP less than 60 mm Hg and NIRS less than 60% (Flow/BrOx1); (2) PbtO2 less than 20 mm Hg and NIRS less than 60% (Flow/BrOx-2); (3) Flow/BrOx-3 (Flow/BrOx-1 1 Flow/ BrOx-2); and (4) simultaneous CPP less than 60 mm Hg and PbtO2 less than 20 mm Hg and NIRS less than 60% (Flow/BrOx-4). Patients were analyzed with respect to survival status to identify differences predictive of outcome. Functional outcome at discharge was assessed with the Table 1
1073
Glasgow Outcome Scale (GOS), whereas functional outcome at 6 and 12 months was assessed with extended GOS (GOSE). The GOSE of survivors (SURV) was obtained by telephone interviews by one of the authors. GOSE score of 1 to 4 defined a poor functional outcome, whereas a score of 5 to 8 defined a good functional outcome. Binomial logistic regression analysis with mortality as the dependent variable was used to identify variables predictive of outcome. Data are presented as means 6 standard deviation, median and interquartile range (IQR), and as proportions. Statistical significance was accepted to correspond to a P value of less than .05. The study was approved by the institutional review board.
Results Twenty-four patients with sTBI were treated with the MM&GDTP for 5 days (Table 1). Two patients with ICP who remained refractory to osmotherapy and developed surgical lesions underwent DC within 24 hours of initiation of the treatment protocol and both the patients survived. In 22 of the 24 patients, enteral feeding was started within 24 hours of injury on completion of the resuscitation phase that included restoration of normovolemia, acid–base balance, and appropriate Ve. The 2 patients who underwent DC had initiation of enteral feeding following the procedure. All 24 patients had IC on day 3 and measurement of NB by day 7. The PM for the group was 58%, whereas actual mortality was 33.3% (8/24), yielding a 42% reduction in mortality. The cause of death of the 8 patients included the following: 2 from progression of sTBI to brain death on day 5, 1 each from a brainstem infarct, myocardial infarction, pulmonary embolism, sepsis-induced multiple organ dysfunction syndrome, and two from withdrawal of care. There was no difference between SURV and nonsurvivors (NS) with respect to GCS, Abbreviated Injury Score (Head) (AIS [H]), Injury Severity Score, and Marshall scores; however, they differed with respect to age and PM (Table 1). Shown in Table 2 is the timeline of ICP, CCP, and PbtO2 stratified by survival status for the 24 patients. NS had a trend toward higher ICP and lower CPP and
Demographic injury severity and Flow/BrOx crisis events
Variable
SURV (n 5 16)
NS (n 5 8)
P value
Age GCS (median with IQR) ISS AIS (H) CrasH 14-day mortality Marshall PM score Flow/BrOx-1 Flow/BrOx-2 Flow/BrOx-3 Flow/BrOx-4
41 6 20 3.5; 3 45 6 18 4.4 6 .9 45 6 28 31 6 20 567 12 6 15 17 6 14 365
62 6 21 3.5; 3.25 37 6 8 5.0 6 .9 83 6 17 37 6 14 19 6 14 33 6 24 54 6 30 45 6 29
.02 1 .18 .1 .005 .45 .002 .01 .001 .001
AIS (H) 5 Abbreviated Injury Score (Head); CrasH 5 corticosteroid randomization after significant head injury; Flow/BrOx 5 blood flow and brain oxygen; GCS 5 Glasgow Outcome Scale; IQR 5 interquartile range; ISS 5 Injury Severity Score; NS 5 nonsurvivors; SURV 5 survivors.
1074 Table 2
The American Journal of Surgery, Vol 208, No 6, December 2014 Timeline of flow and oxygenation variables during the 5-day MM&GDTP CPP
Day
SURV
1 2 3 4 5
71 72 77 77 74
6 6 6 6 6
ICP NS 11 8 8 6 8
63 66 67 76 74
PbtO2
SURV 6 6 6 6 6
14 15 9 15 16
13 14 11 11 10
6 6 6 6 6
NS 8 7 4 4 5
18 21 18 10 12
SURV 6 6 6 6 6
12 15 12 6 9
29 33 32 31 30
6 6 6 6 6
NS 16 11 10 10 10
19 25 27 25 30
6 6 6 6 6
18 14 12 14 9
CPP 5 Cerebral perfusion pressure; ICP 5 intracranial pressure; MM&GDTP 5 multimodality monitoring and goal-directed therapy; NS 5 nonsurvivors; PbtO2 5 brain oxygen tension; SURV 5 survivors.
PbtO2 during the first 2 days following their injury; however, this did not reach statistical significance when compared with SURV.
Flow/BrOx crisis events NS had significantly more Flow/BrOx crisis events than SURV (Table 1). The number of Flow/BrOx-4 crisis events was statistically greater in NS vs SURV (45 6 29 vs 3 6 5, respectively; P , .05). There was a significantly greater number of Flow/BrOx-4 crisis events during the first 3 days of treatment in NS (Fig. 2); no significant difference was observed on days 4 and 5. The analysis of Flow/BrOx-4 showed that 7 of the 8 (87.5%) patients with greater than or equal to 20 episodes died as opposed to only 1 of the 16 (6.3%) patients in the group with less than 20 episodes. Only 1 patient without a single Flow/BrOx-4 crisis event died as opposed to only 1 of the 16 patients having 22 episodes of Flow/BrOx-4 crisis events surviving. There was a statistically significant difference in the mean number of minutes of Flow/BrOx-4 crisis events between SURV and NS (15 6 25 vs 225 6 145 minutes, respectively; P , .05). Despite the MM&GDTP approach to the treatment of the patients, Flow/BrOx events occurred more commonly during the night shifts. The presence of Flow/BrOx-4 crisis events was the only variable predictive of poor outcome by binary stepwise logistic regression with backward and forward elimination. The median Glasgow Outcome Score on hospital discharge was 3.0 with 1.5 IQR for the 16 surviving patients. Functional outcome data at 6 and 12 months was available for 6 patients. Four of the 6 patients had good functional outcome, and median GOSE was 5.0 with 1.5 IQR and 5.5 with 4 IQR at the 6- and 12-month follow-up evaluations, respectively. However, the 2 patients who were discharged with a GOS of 2 showed little improvement and had a GOSE of 3 at 6- and 12-month follow-up evaluations.
recent literature review suggests that the use of an ICP monitor and adherence to guidelines for the management of patients with sTBI are associated with improved outcomes.9 However, to date there is no Level I evidence demonstrating that ICP monitoring is more effective in managing sTBI. Furthermore, CMD data suggest that cellular injury from hypoxia and/or metabolic mitochondrial dysfunction may occur even when CBF is normal as a result of normal ICP and CPP.10 Brain oxygen monitoring was added into the management guidelines for sTBI in 2007.11 Several studies have shown that compromised PbtO2 (,20 mm Hg) is associated with worse outcome in patients with sTBI. As a result, ICP– CPP with PbtO2-targeted therapy has become a common treatment paradigm.12–15 The purpose of this study was to determine the frequency and the impact of early Flow/ BrOx crisis events on the outcome of patients with sTBI when an MM&GDTP was used to guide and maintain hemodynamic and intracranial resuscitation.
MM&GDTP rationale The selection of the end points of our MM&GDTP was based on physiologic and pathophysiologic principles applicable to patients with sTBI. Current opinion is that hyperthermia after TBI is associated with a worse outcome.16 We elected to target temperature control with normothermia rather than hypothermia. Normothermia
Comments Current treatment guidelines for patients with sTBI emphasize the prevention of secondary brain injury for at least the first 48 hours after injury. Therapy is focused on maintaining adequate SBP, ICP, CPP, and oxygenation.6-8 A
Figure 2
Time line of Flow/BrOx-4 crisis events.
C.P. Marini et al.
Early flow and brain oxygen crisis events in sTBI
control with the dry water immersion technique permits a higher water temperature than a hypothermic end point and in theory would be associated with fewer episodes of shivering and less detrimental effect on brain oxygen consumption.17 The target CPP greater than or equal to 60 mm Hg with a PbtO2 greater than 20 mm Hg with an NIRS greater than or equal to 60% incorporates all aspects of the individual monitor end points into one goal. Incorporation of NIRS assumed that a saturation of less than 60% may indicate suboptimal CBF because increased oxygen extraction typically is a compensatory response required to maintain consumption constant in the setting of decreased oxygen delivery.18 We postulated that an increased oxygen extraction is associated with an imbalance between oxygen supply and demand in the injured brain despite adequate CPP and PbtO2. The targeted nutritional approach was based on recent evidence showing that aggressive nutritional support in patients with sTBI provides a survival advantage.19 The incorporation of IC at day 3 provided guidance to the achievement of a positive NB by day 7 of treatment. The decision to treat patients with the sTBI 5-day protocol was based on the physical evaluation of the patient following the initial resuscitation and review of the CT scan findings. The duration of the treatment was based on evidence suggesting that CMD L/P ratios typically do not normalize for at least 3 days following sTBI as well as the 5-day US Food and Drug Administration-approved use of many of the intracranial monitoring devices.20 One consistent observation in our patients was the fact that a SBP targeted to 120 mm Hg could not assure the necessary CPP required to maintain the PbtO2 and NIRS above our established critical thresholds. Our findings on the relationship between SBP CPP and brain oxygenation contrast with the results of a recent paper suggesting that a SBP of 120 mm Hg may be may efficacious in minimizing secondary brain injury in TBI patients.4 A fundamental difference between our study and that by Brenner et al4 is the use of invasive intracranial monitoring techniques. Most of our patients required a SBP between 140 and 160 mm Hg to reach the flow and oxygenation end points of our targeted therapy. The use of norepinephrine to raise SBP to 140 to 160 mm Hg was necessary in 82% of the patients. The increased SBP was rarely associated with a concomitant increase in ICP, indicating intact global autoregulation. Although CPP, ICP, and PbtO2 were within the therapeutic targets in both SURV and NS for the majority of the treatment protocol period, Flow/BrOx crisis events still occurred and were more common in NS as opposed to SURV. The most important finding of the analysis of the Flow/BrOx-4 crisis events pertains to the conclusion that sustained simultaneous decrease in CPP, PbtO2 with NIRS less than 60% for longer than 1 hour in aggregate over the first 3 days are detrimental to the outcome of patients with sTBI. However, in the absence of CMD data correlating Flow/BrOx-4 events with metabolism, we must rely on the binary logistic regression
1075
analysis with mortality as the dependent variable that isolated the number of Flow/BrOx-4 events as predictive of increased mortality. The implementation of our monitoring and resuscitation protocol suggests a lower mortality than that predicted by the CrasH model. The small sample size of this study, however, does not allow a concrete conclusion regarding the impact of an MM&GDTP on the mortality of patients with sTBI. Interestingly, the analysis of the causes of death showed that only 3 of the 8 (37.5%) patients died from progression on their brain injury with the remaining patients dying from non– sTBI-related etiologies. The patient who died from myocardial infarction had the event on day 12, several days after completion of MM&DGTP; therefore, it is unlikely to be related to the use of MM&GDTP. Age provides a significant contribution to the CrasH model. The higher PM for NS in our cohort can be accounted for by the older age of this group. This observation is further supported by the absence of a statistically significant difference in the Marshall score between SURV and NS.
Conclusions Based on the results of our study, we conclude that Flow/ BrOx crisis events predict a poor outcome in patients with sTBI. We believe that further studies are needed to identify physiologic parameters and therapeutic measures that can prevent and/or effectively treat these events. Moreover, a further understanding on the effect of these crisis events on cerebral metabolism may better define the most effective treatment strategies.
Limitations of the study Our study does have limitations. First, CPP and NIRS data were extracted from nursing intensive care unit flow sheets leaving room for transcription error. Second limitation is the temporal distribution of the crisis events, which were more likely to occur at night. We cannot ascertain whether the increased number of crisis events in NS was secondary to the pathophysiology of the brain injury or a less intense adherence to our treatment protocol. A third limitation is with respect to the functional outcome because data on GOSE at 6 and 12 months are limited to only 6 of the 16 patients because of loss to follow-up of the patients. The final limitation is the limited application of the stepwise logistic regression analysis to a model containing only 2 variables because of the number of patients in the study.
References 1. Thurman DJ, Alverson C, Dunn KA, et al. Traumatic brain injury in the United States: a public health perspective. J Head Trauma Rehabil 1999;14:602–15.
1076
The American Journal of Surgery, Vol 208, No 6, December 2014
2. Rutland-Brown W, Langlois JA, Thomas KE, et al. Incidence of traumatic brain injury in the United States, 2003. J Head Trauma Rehabil 2006;21:544–8. 3. Brain Trauma Foundation, American Association of Neurological Surgeons, Congress of Neurological Surgeons. Guidelines for the management of severe traumatic brain injury. J Neurotrauma 2007; 24(Suppl 1):S1–106. 4. Brenner M, Stein DM, Hu PF, et al. Traditional systolic blood pressure targets underestimate hypotension-induced secondary brain injury. J Trauma 2012;72:1135–9. 5. MRC CRASH Trial Collaborators. Predicting outcome after traumatic brain injury: practical prognostic models based on large cohort of international patients. BMJ 2008;336:425–9. 6. Stein SC, Georgoff P, Meghan S, et al. 150 years of treating severe traumatic brain injury: a systematic review of progress in mortality. J Neurotrauma 2010;27:1343–53. 7. Stiefel MF, Udouteuk JD, Spiotta AM, et al. Conventional neurocritical care and cerebral oxygenation after traumatic brain injury. J Neurosurg 2006;105:568–75. 8. Fletcher JJ, Bergman K, Bolstein PA, et al. Fluid balance, complications, and brain tissue oxygen tension monitoring following severe traumatic brain injury. Neurocrit Care 2012;17:49–57. 9. Stein SC, Georgoff P, Meghan S, et al. Relationship of aggressive monitoring and treatment to improved outcomes in severe traumatic brain injury. J Neurosurg 2010;112:1105–12. 10. Vespa P, Bergsneider M, Hattori N, et al. Metabolic crisis without brain ischemia is common after traumatic brain injury: a combined microdialysis, positron emission tomography study. J Cereb Blood Flow Metab 2005;25:763–74. 11. Bratton SL, Chestnut RM, Ghajar J, et al. Guidelines for the management of severe traumatic brain injury. X. Brain oxygen monitoring and thresholds. J Neurotrauma 2007;24(Suppl 1):S65–70. 12. Bardt TF, Unterberg AW, Hartl R, et al. Monitoring of brain tissue PO2 in traumatic brain injury: effect of cerebral hypoxia on outcome. Acta Neurochir Suppl 1998;71:153–6. 13. Chang JJ, Youn TS, Benson D, et al. Physiologic, functional outcome correlates of brain tissue hypoxia in traumatic brain injury. Crit Care Med 2009;37:283–90. 14. Valadka AB, Gopinath SP, Contant CF, et al. Relationship of brain tissue PO2 to outcome after severe head injury. Crit Care Med 1998;26: 1576–81. 15. van den Brink WA, van Santbrink H, Steyerberg EW, et al. Brain oxygen tension in severe head injury. Neurosurgery 2000;46:868–76. 16. Greer DM, Funk SE, Reaven NL, et al. Impact of fever on outcome in patients with stroke and neurological injury. Stroke 2008;39:3029–35. 17. Childs C, Wieloch T, Lecky F, et al. Report of a consensus meeting on human brain temperature after severe traumatic brain injury: its measurement and management during pyrexia. Front Neurol 2010;146: 1–8. 18. Siegal J, Kohli C. Correlation of noninvasive cerebral oximetry with cerebral perfusion in the severe head injured patient: a pilot study. J Trauma 2002;52:40–6. 19. Hartl R, Gerber LM, Ni Q, et al. Effect of early nutrition on deaths due to severe traumatic brain injury. J Neurosurg 2008;109:50–6. 20. Timofeev I, Carpenter KLH, Nortje J, et al. Cerebral extracellular chemistry and outcome following traumatic brain injury: a microdialysis study of 223 patients. Brain 2011;1:1–11.
Discussion Discussant: Dr Fredric Michael Pieracci (Denver, CO). The authors have reported a very nice study, documenting the incidence of crisis events in severe TBI patients and their association with mortality. I have three questions.
First, as a trauma surgeon in a county hospital, the first thing that occurred to me when I was reading the manuscript was how much does the multimodality therapy cost? In our ICU, we follow severe TBI patients with clinical evaluation, serial imaging, and selective ICP monitoring. Do you think that the routine use of the multimodality therapy is cost effective, or do you think that we should be targeting specific subgroups of patients to contain costs? Second, you compared your actual mortality to the predicted mortality based on a calculated score. I wonder if you could really invoke the multimodality therapy specifically as the reason for your decreased mortality, especially since, according to the manuscript, most of the deaths were not related to TBI. So possibly you had just gotten better at providing overall ICU care since you implemented the therapy. Have you thought about doing a pre/post analysis prior to when you implemented the multimodality therapy to see if there was a difference in mortality there? Finally, the nonsurvivors, if my calculations are correct, spent a lot of time in crisis, over the first three days, on average, about two to three hours. I just wanted to get a sense of why these patients who were obviously failing the goaldirected therapy weren’t considered for early decompressive craniotomy. Were their lesions not amenable on CAT scan, or do you not believe that that works, or were they so far gone that you just decided to pursue limited care? Dr Corrado Marini: Let’s start with costs. The daily cost varies between 1,000 and 1,200 dollars since we have added CMD to our multimodality and goal-directed therapy protocol. I cannot stand here and tell you at this time that our multimodality monitoring and goaldirected therapy protocol should be considered the standard of care for patients with severe traumatic brain injury. However, following this study, we have continued to treat over 100 patients with our multimodality monitoring and goal-directed therapy protocol, and we have a survival advantage at this point compared to that predicted by the Rotterdam and Marshall scores of about 35%. And that has been sustainable. However, I do not believe that what we are doing in our trauma intensive care unit, where we have full control of the neurotrauma patients, is generalizable at this time. So I will not conclude here that what we are doing is the standard of care. With respect to the pre and post implementation outcome analysis of our multimodality monitoring and goal-directed therapy protocol, I have conducted that analysis. I arrived at the institution three years ago, and I have had access to the outcome data of patients with severe TBI before my arrival. Before the implementation of our new protocol, patients with severe TBI had only ICP monitoring and craniotomy when indicated. The implementation of the multimodality monitoring and goaldirected therapy protocol has reduced mortality by 25% compared to our historic control. Whether this is the result of a Hawthorne effect, I cannot say.
C.P. Marini et al.
Early flow and brain oxygen crisis events in sTBI
With respect to the question regarding the timing and use of decompressive craniectomy, I can tell you that decompressive craniectomy is not done unless there is a mass effect that explains the increased ICP, which is not controllable with osmotherapy and BS therapy. So, in conclusion, I cannot claim that what we are doing at this time should be generalized. Another important point regards the accuracy of the predicted mortality by the Crash model; the model has good discrimination, but it does not have a great calibration. Therefore, we are now looking at our outcome based on Marshall and Rotterdam scores that tend to be more conservative with respect to predicted mortality. Dr Glenn Ihde (Red Oak, TX). I think it’s a great study in that it shows us an opportunity to improve mortality. Perhaps a window there. My question is, did you keep track of time to initiation of therapy? Because about 15 years ago, there was a study presented here showing that ICP monitors didn’t affect outcome. But having worked at the facility, I knew that the ICP monitors were not placed until maybe 24 or 36 hours into the patient’s stay.
1077
Dr Corrado Marini: First, I want to go back to the question by Dr. Pieracci regarding the number of minutes that patients spent in crisis that I forgot to answer previously. When we analyzed the timeline of the crisis events, we found that despite the use of a standard multimodality and goal-directed therapy protocol, the majority of the events occurred during the night shift. We are now trying to embed a neural network methodology to avoid the variability of the night shift versus the day shift. With respect to the question of the time to the implementation of the protocol, once the trauma surgeons and neurosurgeons have reviewed the CT scan and identified the morphology of the injury, typically within 1 to 2 hours of injury, a decision is made about whether or not the protocol should be started. So unless the patient has a radiological injury that suggests that we should avoid sedation and do a clinical reevaluation, we implement the protocol. The average time to implementation of our multimodality monitoring and goal-directed therapy protocol is less than three hours. Whether or not that makes a difference in preventing secondary ischemic injury, I cannot tell you. But our protocol is initiated very rapidly.
