Severe traumatic brain injury case series

537 patients (mean age, 40 years ± 17 [SD]; 422 men) from one institution from November 2002 to December 2018. Transfer learning and curriculum learning were applied to a convolutional neural network using admission head CT to predict mortality and unfavorable outcomes (Glasgow Outcomes Scale scores 1-3) at 6 months. This was combined with clinical input for a holistic fusion model. The models were evaluated using an independent internal test set and an external cohort of 220 patients with sTBI (mean age, 39 years ± 17; 166 men) from 18 institutions in the Transforming Research and Clinical Knowledge in Traumatic Brain Injury (TRACK-TBI) study from February 2014 to April 2018. The models were compared with the International Mission on Prognosis and Analysis of Clinical Trials in TBI (IMPACT) model and the predictions of three neurosurgeons. Area under the receiver operating characteristic curve (AUC) was used as the main model performance metric. Results The fusion model had higher AUCs than did the IMPACT model in the prediction of mortality (AUC, 0.92 [95% CI: 0.86, 0.97] vs 0.80 [95% CI: 0.71, 0.88]; P < .001) and unfavorable outcomes (AUC, 0.88 [95% CI: 0.82, 0.94] vs 0.82 [95% CI: 0.75, 0.90]; P = .04) on the internal data set. For external TRACK-TBI testing, there was no evidence of a significant difference in the performance of any models compared with the IMPACT model (AUC, 0.83; 95% CI: 0.77, 0.90) in the prediction of mortality. The Imaging model (AUC, 0.73; 95% CI: 0.66-0.81; P = .02) and the fusion model (AUC, 0.68; 95% CI: 0.60, 0.76; P = .02) underperformed as compared with the IMPACT model (AUC, 0.83; 95% CI: 0.77, 0.89) in the prediction of unfavorable outcomes. The fusion model outperformed the predictions of the neurosurgeons.

A deep learning model of head computed tomography and clinical information can be used to predict 6-month severe traumatic brain injury outcome ¹⁾.

Murray et al. in a retrospective study, analyzed a prospectively enrolled cohort of patients with a sTBI at an academic level 1 trauma center. Inclusion criteria were nonpenetrating TBI, age ≥16 years, Glasgow Coma Scale (GCS) score ≤8, and presence of an ICP monitor. Two independent reviewers manually evaluated 30 prespecified features on serial head computed tomography (CTs). Patient characteristics and radiologic features were correlated with elevated ICP. The primary outcome was clinically relevant ICP elevation, defined as ICP ≥ 20 mm Hg on at least 5 or more hourly recordings during postinjury days 0-7 with concurrent administration of an ICP-lowering treatment.

Among 111 sTBI patients, the median GCS was 6 (interquartile range 3-8), and 45% had elevated ICP. Features associated with elevated ICP were younger age (every 10-year decrease, odds ratio [OR] 1.4), modified Fisher scale (mFS) score at 0-4 hours postinjury (every 1 point, OR 1.8), and combined volume of contusional hemorrhage and peri-hematoma edema (10 ml, OR 1.2) at 4-18 hours postinjury.

Younger age, modified Fisher scale (mFS) score, and Intracerebral hemorrhage volume are associated with Intracranial pressure elevation in patients with a severe traumatic brain injury. Imaging features may stratify patients by their risk of subsequent ICP elevation ²⁾.

Thirty-three severe traumatic brain injury patients from a single center who developed severe refractory intracranial hypertension (ICP > 40 mm Hg for longer than 1 h) with ICP, arterial blood pressure, and brain tissue oxygen tension (PBTO2) monitoring (subcohort, n = 9) were selected for retrospective review. Secondary parameters reflecting autoregulation (including pressure reactivity index-PRx, which was available in 24 cases), cerebrospinal compensatory reserve (RAP), and ICP pulse amplitude were calculated.

PRx deteriorated from 0.06 ± 0.26 a.u. at baseline levels of ICP to 0.57 ± 0.24 a.u. (p < 0.0001) at high levels of ICP (> 50 mm Hg). In 4 cases, PRx was impaired (> 0.25 a.u.) before ICP was raised above 25 mm Hg. Concurrently, PBTO2 decreased from 27.3 ± 7.32 mm Hg at baseline ICP to 12.68 ± 7.09 mm Hg at high levels of ICP (p < 0.001). The pulse amplitude of the ICP waveform increased with increasing ICP but showed an 'upper breakpoint'-whereby further increases in ICP lead to decreases in pulse amplitude-in 6 out of the 33 patients.

Severe intracranial hypertension after TBI leads to decreased brain oxygenation, impaired pressure reactivity, and changes in the pulse amplitude of ICP. Impaired pressure reactivity may denote increased risk of developing refractory intracranial hypertension in some patients ³⁾.

Kim et al., enrolled 322 patients with severe trauma and TBI from January 2015 to December 2016. Clinical factors, indexes, and outcomes were compared before and after trauma center establishment (September 2015). The outcome was the Glasgow outcome scale classification at 3 months post-trauma.

Of the 322 patients, 120 (37.3%) and 202 (62.7%) were admitted before and after trauma center establishment, respectively. The two groups were significantly different in age (p=0.038), the trauma location within the city (p=0.010), the proportion of intensive care unit (ICU) admissions (p=0.001), and the emergency room stay time (p<0.001). Mortality occurred in 37 patients (11.5%). Although the preventable death rate decreased from before to after center establishment (23.1% vs. 12.5%), the difference was not significant. None of the clinical factors, indexes, or outcomes were different from before to after center establishment for patients with severe TBI (Glasgow coma scale score ≤8). However, the proportion of inter-hospital transfers increased and the time to emergency room arrival was longer in both the entire cohort and patients with severe TBI after versus before trauma center establishment.

They confirmed that for patients with severe trauma and TBI, establishing a trauma center increased the proportion of ICU admissions and decreased the emergency room stay time and preventable death rate. However, management strategies for handling the high proportion of inter-hospital transfers and long times to emergency room arrival will be necessary ⁴⁾.

Chesnut et al. prospectively studied the outcome from severe head injury (GCS score < or = 8) in 717 cases in the Traumatic Coma Data Bank. They investigated the impact on outcome of hypotension (SBP < 90 mm Hg) and hypoxia (Pao2 < or = 60 mm Hg or apnea or cyanosis in the field) as secondary brain insults, occurring from injury through resuscitation. Hypoxia and hypotension were independently associated with significant increases in morbidity and mortality from severe head injury. Hypotension was profoundly detrimental, occurring in 34.6% of these patients and associated with a 150% increase in mortality. The increased morbidity and mortality related to severe trauma to an extracranial organ system appeared primarily attributable to associated hypotension. Improvements in trauma care delivery over the past decade have not markedly altered the adverse influence of hypotension. Hypoxia and hypotension are common and detrimental secondary brain insults. Hypotension, particularly, is a major determinant of outcome from severe head injury. Resuscitation protocols for brain injured patients should assiduously avoid hypovolemic shock on an absolute basis ⁵⁾.

During 1977-1978, 127 patients with severe head injury were admitted and underwent intracranial pressure monitoring. All patients had Glasgow Coma Scale (GCS) scores of 7 or less. All received identical initial treatment according to a standardized protocol. The patients' average age was 29 years; 60% had multiple trauma, and 35% needed emergency intracranial operations. Treatment for elevations of ICP was begun when ICP rose to 20 to 25 mm Hg, and included mannitol therapy and drainage of cerebrospinal fluid (CSF) when possible. Forty-three patients (34%) had ICP greater than or equal to 25 mm Hg; of these, 36 (84%) died. The mortality rate of the entire group was 46%. During 1979-1980, 106 patients with severe head injury were admitted and underwent ICP monitoring. Their average ager was 29 years; 51% had multiple trauma, and 31% underwent emergency intracranial surgery. All patients received the same standardized protocol as the previous series, with the exception of the treatment of ICP. In this present series: if ICP was 15 mm Hg or less (normal ICP), patients were continued on hyperventilation, steroids, and intensive care; if ICP was 16 to 24 mm Hg, mannitol was administered and CSF was drained; if ICP was 25 mm Hg or greater, the patients were randomized into a controlled barbiturate therapy study. Twenty-six patients (25%) had ICP's of 25 mm Hg or greater, compared to 34% in the previous series (p less than 0.05), and 18 of these 26 patients (69%) died. The overall mortality for this current series was 28% compared to 46% in the previous series (p less than 0.0005). This study reconfirms the high mortality rate if ICP is 25 mm Hg or greater; however, the data also document that early aggressive treatment based on ICP monitoring significantly lessens the incidence of ICP of 25 mm Hg or greater and reduces the overall mortality rate of severe head injury ⁶⁾.