Stroke and Disseminated Intravascular Coagulation in a Patient With Metastatic Prostate Cancer.

Cancer and stroke comprise two of the most common causes of death worldwide. Despite a significantly increased risk of stroke among patients with cancer, most stroke trials have excluded patients with malignancy. There is thus limited evidence to help guide management decisions in this complex population. We present the case of a 78-year-old man with recurrent strokes - both ischemic and hemorrhagic - in the setting of newly-identified metastatic prostate cancer. An atypical cause of cancer-associated stroke is reviewed and the management is discussed.

A Full-Stack Application for Detecting Seizures and Reducing Data During Continuous Electroencephalogram Monitoring.

Continuous electroencephalogram monitoring is associated with lower mortality in critically ill patients; however, it is underused due to the resource-intensive nature of manually interpreting prolonged streams of continuous electroencephalogram data. Here, we present a novel real-time, machine learning-based alerting and monitoring system for epilepsy and seizures that dramatically reduces the amount of manual electroencephalogram review. METHODS: We developed a custom data reduction algorithm using a random forest and deployed it within an online cloud-based platform, which streams data and communicates interactively with caregivers via a web interface to display algorithm results. We developed real-time, machine learning-based alerting and monitoring system for epilepsy and seizures on continuous electroencephalogram recordings from 77 patients undergoing routine scalp ICU electroencephalogram monitoring and tested it on an additional 20 patients. RESULTS: We achieved a mean seizure sensitivity of 84% in cross-validation and 85% in testing, as well as a mean specificity of 83% in cross-validation and 86% in testing, corresponding to a high level of data reduction. This study validates a platform for machine learning-assisted continuous electroencephalogram analysis and represents a meaningful step toward improving utility and decreasing cost of continuous electroencephalogram monitoring. We also make our high-quality annotated dataset of 97 ICU continuous electroencephalogram recordings public for others to validate and improve upon our methods.

Serial Neurologic Assessment in Pediatrics (SNAP): A New Tool for Bedside Neurologic Assessment of Critically Ill Children.

OBJECTIVES: We developed a tool, Serial Neurologic Assessment in Pediatrics, to screen for neurologic changes in patients, including those who are intubated, are sedated, and/or have developmental disabilities. Our aims were to: 1) determine protocol adherence when performing Serial Neurologic Assessment in Pediatrics, 2) determine the interrater reliability between nurses, and 3) assess the feasibility and acceptability of using Serial Neurologic Assessment in Pediatrics compared with the Glasgow Coma Scale. DESIGN: Mixed-methods, observational cohort. SETTING: Pediatric and neonatal ICUs. SUBJECTS: Critical care nurses and patients. INTERVENTIONS: None. MEASUREMENTS AND MAIN RESULTS: Serial Neurologic Assessment in Pediatrics assesses Mental Status, Cranial Nerves, Communication, and Motor Function, with scales for children less than 6 months, greater than or equal to 6 months to less than 2 years, and greater than or equal to 2 years old. We assessed protocol adherence with standardized observations. We assessed the interrater reliability of independent Serial Neurologic Assessment in Pediatrics assessments between pairs of trained nurses by percent- and bias- adjusted kappa and percent agreement. Semistructured interviews with nurses evaluated acceptability and feasibility after nurses used Serial Neurologic Assessment in Pediatrics concurrently with Glasgow Coma Scale during routine care. Ninety-eight percent of nurses (43/44) had 100% protocol adherence on the standardized checklist. Forty-three nurses performed 387 paired Serial Neurologic Assessment in Pediatrics assessments (149 < 6 mo; 91 ≥ 6 mo to < 2 yr, and 147 ≥ 2 yr) on 299 patients. Interrater reliability was substantial to near-perfect across all components for each age-based Serial Neurologic Assessment in Pediatrics scale. Percent agreement was independent of developmental disabilities for all Serial Neurologic Assessment in Pediatrics components except Mental Status and lower extremity Motor Function for patients deemed "Able to Participate" with the assessment. Nurses reported that they felt Serial Neurologic Assessment in Pediatrics, compared with Glasgow Coma Scale, was easier to use and clearer in describing the neurologic status of patients who were intubated, were sedated, and/or had developmental disabilities. About 92% of nurses preferred to use Serial Neurologic Assessment in Pediatrics over Glasgow Coma Scale. CONCLUSIONS: When used by critical care nurses, Serial Neurologic Assessment in Pediatrics has excellent protocol adherence, substantial to near-perfect interrater reliability, and is feasible to implement. Further work will determine the sensitivity and specificity for detecting clinically meaningful neurologic decline.

Coronavirus Disease (COVID-19)-Related Disseminated Leukoencephalopathy: A Retrospective Study of Findings on Brain MRI.

Among 2820 inpatients with coronavirus disease (COVID-19), 59 (2.1%) underwent brain MRI. Of them, six (10.2%) had MRI findings suspicious for COVID-19-related disseminated leukoencephalopathy (CRDL), which is characterized by extensive confluent or multifocal white matter lesions (with characteristics and locations atypical for other causes), microhemorrhages, diffusion restriction, and enhancement. CRDL is an uncommon but important differential consideration in patients with neurologic manifestations of COVID-19.

The Safety of Multimodality Monitoring Using a Triple-Lumen Bolt in Severe Acute Brain Injury.

BACKGROUND: Multimodality monitoring is used frequently to guide care of patients with severe acute brain injury. The aim of this study was to examine the safety and reliability of multimodality monitoring. METHODS: From a prospective observational database at a Level I trauma center, 501 patients, including 300 men and 201 women (mean age 58 ± 39 years) were identified retrospectively. Each patient received a triple-lumen bolt and 3 monitors: intracranial pressure, brain temperature, and brain oxygen. Intensive care unit and hospital records were examined to identify complications, reasons for device replacement, malfunction and infection. Head computed tomography (CT) scans performed before and after the monitors were inserted were examined for evidence of monitor-related adverse effects. RESULTS: A total of 696 triple-lumen bolts were placed. Median duration of monitoring was 78.88 hours (interquartile range, 33.0-133.2 hours). Bilateral monitors were inserted in 22 (3.16%) patients. Ten (1.43%) monitors were replaced to allow magnetic resonance imaging, and 40 (5.74%) monitors were replaced to facilitate additional cranial surgery. Of 35 (5.02%) monitors that were replaced because they were thought to not be functioning properly, 19 (54.29%) were subsequently found to be functioning normally. Follow-up CT scans were compared with CT scans obtained before insertion of monitors; 9 (2.13%) small contusions and 10 (2.36%) extra-axial hematomas associated with the devices were identified. Based on the CT findings, the hematomas were thought to be associated with the insertion technique rather than the device; 4 hematomas required treatment. Twenty-two (3.16%) devices were incorrectly placed (e.g., the probe was in an infarct or an already existing contusion). Only 1 associated infection was identified. CONCLUSIONS: Placement of intracranial monitors for multimodality neuromonitoring using a triple-lumen bolt appears to be safe. The complication rate is similar to published complication rates for single-lumen bolts and single monitors.

Race and stroke in an atrial fibrillation inception cohort: Findings from the Penn Atrial Fibrillation Free study.

BACKGROUND: Stroke may be the initial manifestation of atrial fibrillation (AF). Limited studies, however, have evaluated racial differences in stroke before the diagnosis of AF. OBJECTIVE: We assessed racial differences in strokes that occurred before and after AF diagnosis in the Penn Atrial Fibrillation Free study. METHODS: The Penn Atrial Fibrillation Free study consists of 56,835 patients from the University of Pennsylvania Health System who were free of AF at the index visit. We developed an inception cohort of 3507 patients with incident AF and without any remote history of stroke. RESULTS: Among the AF inception cohort, there were 538 patients with ischemic strokes and 54 with hemorrhagic strokes. Nearly half (n = 254; 47%) of the ischemic strokes occurred within a 6-month period before the diagnosis of AF. Of these, the majority of strokes occurred either on the day of (n = 158) or within a 7-day period before (n = 30) the diagnosis of incident AF. The remaining 284 (53%) ischemic strokes occurred a median of 3.6 years (interquartile range 1.9-5.4 years) after AF diagnosis. Compared with whites, blacks had an independently higher risk of having an ischemic stroke either before (adjusted odds ratio 1.37; 95% confidence interval 1.03-1.81) or after (adjusted hazard ratio 1.67; 95% confidence interval 1.30-2.14) AF diagnosis. CONCLUSION: In the population with incident AF, nearly half of the ischemic strokes occurred before the diagnosis of AF. Compared with whites, blacks had a higher risk of developing an ischemic stroke that persisted whether the stroke occurred in the period either before or after AF diagnosis.

Plasma ADAMTS13 activity and von Willebrand factor antigen and activity in patients with subarachnoid haemorrhage.

Increased von Willebrand factor (VWF) and reduced ADAMTS13 activity are associated with arterial thrombosis. This may also be the culprit mechanism implicated in delayed cerebral ischaemia after aneurysmal subarachnoid haemorrhage (SAH). It was our objective to determine plasma VWF and ADAMTS13 in patients with SAH and healthy subjects; and to explore the levels of those markers and outcome after SAH. Forty consecutive patients were enrolled between September 2007 and April 2014 in a pilot study. Plasma samples were collected from SAH patients on post-bleed day (PBD) 0, 1, 3, 5, 7 and 10 and healthy controls. VWF antigen (VWFAg) and VWF activity (VWFAc) were determined by enzyme-linked immunoassay and collagen binding assay, respectively. ADAMTS13 activity was determined by the cleavage of a fluorescent substrate. Univariate descriptive statistics and cluster analyses were performed based on outcomes in the group with SAH only. Mean age of SAH patients was 52.4 years (26-84 years) and 30 (75 %) were women. 12/40 (30 %) had a high Hunt and Hess grade (IV-V) and 25 (62.5 %) were treated with coil embolisation. Plasma VWFAg and VWFAc were significantly higher in SAH patients than those in healthy subjects on each PBD (p<0.0001). Concurrently, plasma ADAMTS13 activity in SAH patients was significantly lower than that in healthy subjects (p<0.0001). Among those with SAH, cluster analysis demonstrated that patients with higher VWFAg and VWFAc and/or lower ADAMTS13 activity might be at risk of increased mortality. In conclusion, the relative deficiency of plasma ADAMTS13 activity in SAH patients may associate with worse outcome.

Elevated Red Cell Distribution Width is Associated with Cerebral Infarction in Aneurysmal Subarachnoid Hemorrhage.

BACKGROUND: Elevated red blood cell distribution width (RDW) has been associated with thrombotic disorders including myocardial infarction, venous thromboembolism, and ischemic stroke, independent of other inflammatory and coagulation biomarkers. The purpose of this study was to determine whether elevated RDW is associated with cerebral infarction and poor outcome after aneurysmal subarachnoid hemorrhage (aSAH). METHODS: In this retrospective single-center cohort of aSAH patients (October 2009-September 2014), elevated RDW was defined as a mean RDW >14.5 % during the first 14 days after aSAH. Outcomes included cerebral infarction (CI) by any mechanism and poor functional outcome, defined as discharge modified Rankin Scale (mRS) >4, indicating severe disability or death. RESULTS: Of 179 patients, 27 % had a high Hunt-Hess grade (IV-V), and 76 % were women. Twenty-four patients (13.4 %) underwent red blood cell (RBC) transfusion and compared to patients with normal RDW, patients with an elevated RDW were at greater odds of RBC transfusion (OR 2.56 [95 % CI, 1.07-6.11], p = 0.035). In univariate analysis, more patients with elevated RDW experienced CI (30.8 vs. 13.7 %, p = 0.017). In the multivariable model, elevated RDW was significantly associated with CI (OR 3.08 [95 % CI, 1.30-7.32], p = 0.011), independent of known confounders including but not limited to age, sex, race, high Hunt-Hess grade, and RBC transfusion. In multivariable analysis, RDW elevation was also associated with poor functional outcome (mRS > 4) at discharge (OR 2.59 [95 % CI, 1.04-629], p = 0.040). CONCLUSIONS: RDW elevation is associated with cerebral infarction and poor outcome after aSAH. Further evaluation of this association is warranted as it may shed light on mechanistic relations between anemia, inflammation, and thrombosis after aSAH.

Thromboelastography Parameter Predicts Outcome After Subarachnoid Hemorrhage: An Exploratory Analysis.

OBJECTIVE: Hypercoagulability after subarachnoid hemorrhage (SAH) is well described and may be platelet mediated. Thromboelastography (TEG) provides a global assessment of coagulation. We sought to determine whether the maximum amplitude (MA) parameter of TEG, a measure of platelet strength and function, is associated with outcome after SAH. METHODS: One hundred ten TEG analyses were performed for patients with moderate-to-severe SAH and compared with 6 healthy age- and sex-matched controls. TEG indices included MA, G value (G), alpha angle, and thrombus generation and were correlated to functional outcomes and laboratory tests including complete blood count, erythrocyte sedimentation rate, high sensitivity C-reactive protein, fibrinogen, and d-dimer, obtained on post-bleed days (PBDs) 1, 3, 5, 7, and 10. RESULTS: MA was significantly elevated compared with controls on PBD 3 (70.0 mm ± 4.5 mm vs. 64.1 mm ± 6.5 mm; P = 0.02), PBD 5 (72.6 mm ± 5.3 mm vs. 64.1 mm ± 6.5 mm; P = 0.003), PBD 7 (73.0 mm ± 5.4 mm vs. 64.1 mm ± 6.5 mm; P = 0.003), and PBD 10 (73.4 mm ± 6.0 mm vs. 64.1 mm ± 6.5 mm; P = 0.005). G was significantly elevated compared with controls on PBD 3 (P = 0.03), PBD 5 (P = 0.01), PBD 7 (P = 0.01), and PBD 10 (P = 0.02). The only biomarker associated with poor outcome was CRP. Multivariate logistic regression demonstrated an association between elevated MA and outcome (odds ratio 39.1, P = 0.006) independent of CRP, age, Hunt Hess grade, and transfusion. CONCLUSIONS: TEG indices are associated with poor outcome after SAH and may identify a platelet-mediated hypercoagulable state. The association between MA and outcome was stronger than that between traditional biomarkers and was independent of age and Hunt Hess grade.

Concussion and football: a review and editorial.

The issue of concussion in football is of substantial interest to players, coaches, fans, and physicians. In this article, we review specific cultural hindrances to diagnosis and treatment of concussion in football. We review current trends in management and identify areas for improvement. We also discuss the obligations that physicians, particularly neurosurgeons and neurologists, have toward brain-injured football players and the larger societal role they may play in helping to minimize football-associated brain injury.

Fever and therapeutic normothermia in severe brain injury: an update.

PURPOSE OF REVIEW: Fever is common in the ICU among patients with severe brain injury. Fever has been consistently shown to exacerbate brain injuries in animal models and has been consistently associated with poor outcome in human studies. However, whether fever control improves outcome and the ideal means of fever control remain unknown. This review will address recent literature on the impact of fever on severe brain injury and on interventions to maintain normothermia. RECENT FINDINGS: Current guidelines generally recommend maintenance of normothermia after brain injury but have scant recommendations on methods to do this. Observational trials have continued to demonstrate the association between fever and poor outcome after severe brain injury. Recent trials have shown the efficacy of more aggressive approaches to fever reduction, whereas a large randomized trial showed the relative ineffectiveness of acetaminophen alone for fever control. Several studies have also described the impact of fever and of fever control on brain physiology. SUMMARY: The value of therapeutic normothermia in the neurocritical care unit (NCCU) is increasingly accepted, yet prospective trials that demonstrate a functional benefit to patients are lacking.

Continuous optical monitoring of cerebral hemodynamics during head-of-bed manipulation in brain-injured adults.

INTRODUCTION: Head-of-bed manipulation is commonly performed in the neurocritical care unit to optimize cerebral blood flow (CBF), but its effects on CBF are rarely measured. This pilot study employs a novel, non-invasive instrument combining two techniques, diffuse correlation spectroscopy (DCS) for measurement of CBF and near-infrared spectroscopy (NIRS) for measurement of cerebral oxy- and deoxy-hemoglobin concentrations, to monitor patients during head-of-bed lowering. METHODS: Ten brain-injured patients and ten control subjects were monitored continuously with DCS and NIRS while the head-of-bed was positioned first at 30° and then at 0°. Relative CBF (rCBF) and concurrent changes in oxy- (ΔHbO2), deoxy- (ΔHb), and total-hemoglobin concentrations (ΔTHC) from left/right frontal cortices were monitored for 5 min at each position. Patient and control response differences were assessed. RESULTS: rCBF, ΔHbO2, and ΔTHC responses to head lowering differed significantly between brain-injured patients and healthy controls (P < 0.02). For patients, rCBF changes were heterogeneous, with no net change observed in the group average (0.3 ± 28.2 %, P = 0.938). rCBF increased in controls (18.6 ± 9.4 %, P < 0.001). ΔHbO2, ΔHb, and ΔTHC increased with head lowering in both groups, but to a larger degree in brain-injured patients. rCBF correlated moderately with changes in cerebral perfusion pressure (R = 0.40, P < 0.001), but not intracranial pressure. CONCLUSION: DCS/NIRS detected differences in CBF and oxygenation responses of brain-injured patients versus controls during head-of-bed manipulation. This pilot study supports the feasibility of continuous bedside measurement of cerebrovascular hemodynamics with DCS/NIRS and provides the rationale for further investigation in larger cohorts.

Response of brain oxygen to therapy correlates with long-term outcome after subarachnoid hemorrhage.

BACKGROUND: Brain oxygen (PbtO2) monitoring can help guide care of poor-grade aneurysmal subarachnoid hemorrhage (aSAH) patients. The relationship between PbtO2-directed therapy and long-term outcome is unclear. We hypothesized that responsiveness to PbtO2-directed interventions is associated with outcome. METHODS: Seventy-six aSAH patients who underwent PbtO2 monitoring were included. Long-term outcome [Glasgow Outcome Score-Extended (GOS-E) and modified Rankin Scale (mRS)] was ascertained using the social security death database and structured telephone interviews. Univariate and multivariate regression were used to identify variables that correlated with outcome. RESULTS: Data from 64 patients were analyzed (12 were lost to follow-up). There were 530 episodes of compromised PbtO2 (<20 mmHg) during a total of 7,174 h of monitor time treated with 1,052 interventions. Forty-two patients (66 %) survived to discharge. Median follow-up was 8.5 months (range 0.1-87). At most recent follow-up 35 (55 %) patients were alive, and 28 (44 %) had a favorable outcome (mRS ≤3). In multivariate ordinal regression analysis, only age and response to PbtO2-directed intervention correlated significantly with outcome. Increased age was associated with worse outcome (coeff. 0.8, 95 % CI 0.3-1.3, p = 0.003), and response to PbtO2-directed intervention was associated with improved outcome (coeff. -2.12, 95 % CI -4.0 to -0.26, p = 0.03). Patients with favorable outcomes had a 70 % mean rate of response to PbtO2-directed interventions whereas patients with poor outcomes had a 45 % response rate (p = 0.005). CONCLUSIONS: Response to PbtO2-directed intervention is associated with improved long-term functional outcome in aSAH patients.

Postarrest targeted temperature management immediately following craniotomy-a case report.

Therapeutic hypothermia or targeted temperature management (TTM) has been shown to improve survival and neurological outcome after cardiac arrest. TTM is not frequently utilized in the postoperative setting because of the concern for exacerbation of bleeding. We present the case of a 65-year-old man who had a cardiac arrest during craniotomy for a brain tumor resection. He was successfully resuscitated from pulseless electrical activity and remained unresponsive. After assessment for postoperative brain hemorrhage, the neurocritical care team initiated TTM. Repeat imaging revealed no additional bleeding. The patient was discharged with a cerebral performance category of 1 to an acute rehabilitation center 11 days following his cardiac arrest. This case highlights the need for further consideration of TTM in the postoperative cardiac arrest population.

Temporal dynamics of microparticle elevation following subarachnoid hemorrhage.

OBJECT: Microparticles (MPs), small membrane fragments shed from various cell types, have been implicated in thrombosis, inflammation, and endothelial dysfunction. Their involvement in subarachnoid hemorrhage (SAH) and the development of cerebral infarction and clinical deterioration caused by delayed cerebral ischemia (DCI) remain ill defined. The authors sought to quantify the magnitude of elevations in MPs, delineate the temporal dynamics of elevation, and analyze the correlation between MPs and DCI in patients with SAH. METHODS: On the day of hemorrhage and on Days 1, 3, 5, 7, and 10 after hemorrhage, peripheral blood samples were drawn from 22 patients with SAH. Plasma samples were labeled with Annexin V and CD142, CD41a, CD235a, CD146, CD66b, or von Willebrand factor (vWF) and were quantified by flow cytometry. Clinical data, including the 3-month extended Glasgow Outcome Scale (GOS-E) scores, infarction as measured on MRI at 14 days after SAH, and vasospasm as measured by transcranial Doppler ultrasonography and angiography, were collected and compared with the MP burden. RESULTS: When averaged over time, all MP subtypes were elevated relative to controls. The CD235a+(erythrocyte)-, CD66b+(neutrophil)-, and vWF-associated MPs peaked on the day of hemorrhage and quickly declined. The CD142+(tissue factor [TF])-associated MPs and CD146+(endothelial cell)-associated MPs were significantly elevated throughout the study period. There was a strong negative correlation between TF-expressing and endothelial-derived MPs at Day 1 after SAH and the risk of infarction at Day 14 after SAH. CONCLUSIONS: Microparticles of various subtypes are elevated following SAH; however, the temporal profile of this elevation varies by subtype. Those subtypes closely associated with thrombosis and endothelial dysfunction, for example, CD145+(TF)-associated MPs and CD146+(endothelial cell)-associated MPs, had the most durable response and demonstrated a significant negative correlation with radiographic infarction at 14 days after SAH. Levels of these MPs predict infarction as early as Day 1 post-SAH.

Anemia and brain oxygen after severe traumatic brain injury.

PURPOSE: To investigate the relationship between hemoglobin (Hgb) and brain tissue oxygen tension (PbtO(2)) after severe traumatic brain injury (TBI) and to examine its impact on outcome. METHODS: This was a retrospective analysis of a prospective cohort of severe TBI patients whose PbtO(2) was monitored. The relationship between Hgb-categorized into four quartiles (≤9; 9-10; 10.1-11; >11 g/dl)-and PbtO(2) was analyzed using mixed-effects models. Anemia with compromised PbtO(2) was defined as episodes of Hgb ≤ 9 g/dl with simultaneous PbtO(2) < 20 mmHg. Outcome was assessed at 30 days using the Glasgow outcome score (GOS), dichotomized as favorable (GOS 4-5) vs. unfavorable (GOS 1-3). RESULTS: We analyzed 474 simultaneous Hgb and PbtO(2) samples from 80 patients (mean age 44 ± 20 years, median GCS 4 (3-7)). Using Hgb > 11 g/dl as the reference level, and controlling for important physiologic covariates (CPP, PaO(2), PaCO(2)), Hgb ≤ 9 g/dl was the only Hgb level that was associated with lower PbtO(2) (coefficient -6.53 (95 % CI -9.13; -3.94), p < 0.001). Anemia with simultaneous PbtO(2) < 20 mmHg, but not anemia alone, increased the risk of unfavorable outcome (odds ratio 6.24 (95 % CI 1.61; 24.22), p = 0.008), controlling for age, GCS, Marshall CT grade, and APACHE II score. CONCLUSIONS: In this cohort of severe TBI patients whose PbtO(2) was monitored, a Hgb level no greater than 9 g/dl was associated with compromised PbtO(2). Anemia with simultaneous compromised PbtO(2), but not anemia alone, was a risk factor for unfavorable outcome, irrespective of injury severity.

Is aggressive treatment of traumatic brain injury cost-effective?

OBJECT: The object of this study was to determine whether aggressive treatment of severe traumatic brain injury (TBI), including invasive intracranial monitoring and decompressive craniectomy, is cost-effective. METHODS: A decision-analytical model was created to compare costs, outcomes, and cost-effectiveness of 3 strategies for treating a patient with severe TBI. The aggressive-care approach is compared with "routine care," in which Brain Trauma Foundation guidelines are not followed. A "comfort care" category, in which a single day in the ICU is followed by routine floor care, is included for comparison only. Probabilities of each treatment resulting in various Glasgow Outcome Scale (GOS) scores were obtained from the literature. The GOS scores were converted to quality-adjusted life years (QALYs), based on expected longevity and calculated quality of life associated with each GOS category. Estimated direct (acute and long-term medical care) and indirect (loss of productivity) costs were calculated from the perspective of society. Sensitivity analyses employed a 2D Monte Carlo simulation of 1000 trials, each with 1000 patients. The model was also used to estimate these values for patients 40, 60, and 80 years of age. RESULTS: For the average 20-year-old, aggressive care yields 11.7 (± 1.6 [SD]) QALYs, compared with routine care (10.0 ± 1.5 QALYs). This difference is highly significant (p < 0.0001). Although the differences in effectiveness between the 2 strategies diminish with advancing age, aggressive care remains significantly better at all ages. When all costs are considered, aggressive care is also significantly less costly than routine care ($1,264,000 ± $118,000 vs $1,361,000 ± $107,000) for the average 20-year-old. Aggressive care remains significantly less costly until age 80, at which age it costs more than routine care. However, even in the 80-year-old, aggressive care is likely the more cost-effective approach. Comfort care is associated with poorer outcomes at all ages and with higher costs for all groups except 80-year-olds. CONCLUSIONS: When all the costs of severe TBI are considered, aggressive treatment is a cost-effective option, even for older patients. Comfort care for severe TBI is associated with poor outcomes and high costs, and should be reserved for situations in which aggressive approaches have failed or testing suggests such treatment is futile.