Evaluation of ChatGPT in Predicting 6-Month Outcomes After Traumatic Brain Injury.

OBJECTIVES: To evaluate the capacity of ChatGPT, a widely accessible and uniquely popular artificial intelligence-based chatbot, in predicting the 6-month outcome following moderate-to-severe traumatic brain injury (TBI). DESIGN: Single-center observational retrospective study. SETTING: Data are from a neuro-ICU from a level 1 trauma center. PATIENTS: All TBI patients admitted to ICU between September 2021 and October 2022 were included in a prospective database. INTERVENTIONS: None. MEASUREMENTS AND MAIN RESULTS: Based on anonymized clinical, imaging, and biological information available at the patients' hospital admission and extracted from the database, clinical vignettes were retrospectively submitted to ChatGPT for prediction of patients' outcomes. The predictions of two intensivists (one neurointensivist and one non-neurointensivist) both from another level 1 trauma center (Beaujon Hospital), were also collected as was the International Mission on Prognosis and Analysis of Clinical Trials in Traumatic Brain Injury (IMPACT) scoring. Each intensivist, as well as ChatGPT, made their prognostic evaluations independently, without knowledge of the others' predictions and of the patients' actual management and outcome. Both the intensivists and ChatGPT were given access to the exact same set of information. The main outcome was a 6-month-functional status dichotomized into favorable (Glasgow Outcome Scale Extended [GOSE] ≥ 5) versus poor (GOSE < 5). Prediction of intracranial hypertension management, pulmonary infectious risk, and removal of life-sustaining therapies was also investigated as secondary outcomes. Eighty consecutive moderate-to-severe TBI patients were included. For the 6-month outcome prognosis, area under the receiver operating characteristic curve (AUC-ROC) for ChatGPT, the neurointensivist, the non-neurointensivist, and IMPACT were, respectively, 0.62 (0.50-0.74), 0.70 (0.59-0.82), 0.71 (0.59-0.82), and 0.81 (0.72-0.91). ChatGPT had the highest sensitivity (100%), but the lowest specificity (26%). For secondary outcomes, ChatGPT's prognoses were generally less accurate than clinicians' prognoses, with lower AUC values for most outcomes. CONCLUSIONS: This study does not support the use of ChatGPT for prediction of outcomes after TBI.

Pilot deployment of a machine-learning enhanced prediction of need for hemorrhage resuscitation after trauma - the ShockMatrix pilot study.

IMPORTANCE: Decision-making in trauma patients remains challenging and often results in deviation from guidelines. Machine-Learning (ML) enhanced decision-support could improve hemorrhage resuscitation. AIM: To develop a ML enhanced decision support tool to predict Need for Hemorrhage Resuscitation (NHR) (part I) and test the collection of the predictor variables in real time in a smartphone app (part II). DESIGN, SETTING, AND PARTICIPANTS: Development of a ML model from a registry to predict NHR relying exclusively on prehospital predictors. Several models and imputation techniques were tested. Assess the feasibility to collect the predictors of the model in a customized smartphone app during prealert and generate a prediction in four level-1 trauma centers to compare the predictions to the gestalt of the trauma leader. MAIN OUTCOMES AND MEASURES: Part 1: Model output was NHR defined by 1) at least one RBC transfusion in resuscitation, 2) transfusion ≥ 4 RBC within 6 h, 3) any hemorrhage control procedure within 6 h or 4) death from hemorrhage within 24 h. The performance metric was the F4-score and compared to reference scores (RED FLAG, ABC). In part 2, the model and clinician prediction were compared with Likelihood Ratios (LR). RESULTS: From 36,325 eligible patients in the registry (Nov 2010-May 2022), 28,614 were included in the model development (Part 1). Median age was 36 [25-52], median ISS 13 [5-22], 3249/28614 (11%) corresponded to the definition of NHR. A XGBoost model with nine prehospital variables generated the best predictive performance for NHR according to the F4-score with a score of 0.76 [0.73-0.78]. Over a 3-month period (Aug-Oct 2022), 139 of 391 eligible patients were included in part II (38.5%), 22/139 with NHR. Clinician satisfaction was high, no workflow disruption observed and LRs comparable between the model and the clinicians. CONCLUSIONS AND RELEVANCE: The ShockMatrix pilot study developed a simple ML-enhanced NHR prediction tool demonstrating a comparable performance to clinical reference scores and clinicians. Collecting the predictor variables in real-time on prealert was feasible and caused no workflow disruption.

Impact of continuous hypertonic (NaCl 20%) saline solution on renal outcomes after traumatic brain injury (TBI): a post hoc analysis of the COBI trial.

BACKGROUND: To evaluate if the increase in chloride intake during a continuous infusion of 20% hypertonic saline solution (HSS) is associated with an increase in the incidence of acute kidney injury (AKI) compared to standard of care in traumatic brain injury patients. METHODS: In this post hoc analysis of the COBI trial, 370 patients admitted for a moderate-to-severe TBI in the 9 participating ICUs were enrolled. The intervention consisted in a continuous infusion of HSS to maintain a blood sodium level between 150 and 155 mmol/L for at least 48 h. Patients enrolled in the control arm were treated as recommended by the latest Brain Trauma foundation guidelines. The primary outcome of this study was the occurrence of AKI within 28 days after enrollment. AKI was defined by stages 2 or 3 according to KDIGO criteria. RESULTS: After exclusion of missing data, 322 patients were included in this post hoc analysis. The patients randomized in the intervention arm received a significantly higher amount of chloride during the first 4 days (intervention group: 97.3 ± 31.6 g vs. control group: 61.3 ± 38.1 g; p < 0.001) and had higher blood chloride levels at day 4 (117.9 ± 10.7 mmol/L vs. 111.6 ± 9 mmol/L, respectively, p < 0.001). The incidence of AKI was not statistically different between the intervention and the control group (24.5% vs. 28.9%, respectively; p = 0.45). CONCLUSIONS: Despite a significant increase in chloride intake, a continuous infusion of HSS was not associated with AKI in moderate-to-severe TBI patients. Our study does not confirm the potentially detrimental effect of chloride load on kidney function in ICU patients. TRIAL REGISTRATION: The COBI trial was registered on clinicaltrial.gov (Trial registration number: NCT03143751, date of registration: 8 May 2017).

Effect of Continuous Infusion of Hypertonic Saline vs Standard Care on 6-Month Neurological Outcomes in Patients With Traumatic Brain Injury: The COBI Randomized Clinical Trial.

Importance: Fluid therapy is an important component of care for patients with traumatic brain injury, but whether it modulates clinical outcomes remains unclear. Objective: To determine whether continuous infusion of hypertonic saline solution improves neurological outcome at 6 months in patients with traumatic brain injury. Design, Setting, and Participants: Multicenter randomized clinical trial conducted in 9 intensive care units in France, including 370 patients with moderate to severe traumatic brain injury who were recruited from October 2017 to August 2019. Follow-up was completed in February 2020. Interventions: Adult patients with moderate to severe traumatic brain injury were randomly assigned to receive continuous infusion of 20% hypertonic saline solution plus standard care (n = 185) or standard care alone (controls; n = 185). The 20% hypertonic saline solution was administered for 48 hours or longer if patients remained at risk of intracranial hypertension. Main Outcomes and Measures: The primary outcome was Extended Glasgow Outcome Scale (GOS-E) score (range, 1-8, with lower scores indicating worse functional outcome) at 6 months, obtained centrally by blinded assessors and analyzed with ordinal logistic regression adjusted for prespecified prognostic factors (with a common odds ratio [OR] >1.0 favoring intervention). There were 12 secondary outcomes measured at multiple time points, including development of intracranial hypertension and 6-month mortality. Results: Among 370 patients who were randomized (median age, 44 [interquartile range, 27-59] years; 77 [20.2%] women), 359 (97%) completed the trial. The adjusted common OR for the GOS-E score at 6 months was 1.02 (95% CI, 0.71-1.47; P = .92). Of the 12 secondary outcomes, 10 were not significantly different. Intracranial hypertension developed in 62 (33.7%) patients in the intervention group and 66 (36.3%) patients in the control group (absolute difference, -2.6% [95% CI, -12.3% to 7.2%]; OR, 0.80 [95% CI, 0.51-1.26]). There was no significant difference in 6-month mortality (29 [15.9%] in the intervention group vs 37 [20.8%] in the control group; absolute difference, -4.9% [95% CI, -12.8% to 3.1%]; hazard ratio, 0.79 [95% CI, 0.48-1.28]). Conclusions and Relevance: Among patients with moderate to severe traumatic brain injury, treatment with continuous infusion of 20% hypertonic saline compared with standard care did not result in a significantly better neurological status at 6 months. However, confidence intervals for the findings were wide, and the study may have had limited power to detect a clinically important difference. Trial Registration: ClinicalTrials.gov Identifier: NCT03143751.

Early hyperoxemia is associated with lower adjusted mortality after severe trauma: results from a French registry.

BACKGROUND: Hyperoxemia has been associated with increased mortality in critically ill patients, but little is known about its effect in trauma patients. The objective of this study was to assess the association between early hyperoxemia and in-hospital mortality after severe trauma. We hypothesized that a PaO2 ≥ 150 mmHg on admission was associated with increased in-hospital mortality. METHODS: Using data issued from a multicenter prospective trauma registry in France, we included trauma patients managed by the emergency medical services between May 2016 and March 2019 and admitted to a level I trauma center. Early hyperoxemia was defined as an arterial oxygen tension (PaO2) above 150 mmHg measured on hospital admission. In-hospital mortality was compared between normoxemic (150 > PaO2 ≥ 60 mmHg) and hyperoxemic patients using a propensity-score model with predetermined variables (gender, age, prehospital heart rate and systolic blood pressure, temperature, hemoglobin and arterial lactate, use of mechanical ventilation, presence of traumatic brain injury (TBI), initial Glasgow Coma Scale score, Injury Severity Score (ISS), American Society of Anesthesiologists physical health class > I, and presence of hemorrhagic shock). RESULTS: A total of 5912 patients were analyzed. The median age was 39 [26-55] years and 78% were male. More than half (53%) of the patients had an ISS above 15, and 32% had traumatic brain injury. On univariate analysis, the in-hospital mortality was higher in hyperoxemic patients compared to normoxemic patients (12% versus 9%, p < 0.0001). However, after propensity score matching, we found a significantly lower in-hospital mortality in hyperoxemic patients compared to normoxemic patients (OR 0.59 [0.50-0.70], p < 0.0001). CONCLUSION: In this large observational study, early hyperoxemia in trauma patients was associated with reduced adjusted in-hospital mortality. This result contrasts the unadjusted in-hospital mortality as well as numerous other findings reported in acutely and critically ill patients. The study calls for a randomized clinical trial to further investigate this association.

Left double-lumen tube with or without a carinal hook: A randomised controlled trial.

BACKGROUND: Left double-lumen tracheal tubes (DLTs), with or without a hook to engage the carina, remain the standard device for lung isolation during anaesthesia. OBJECTIVE: The purpose of the study was to compare these DLTs with and without a hook. DESIGN: A randomised, controlled, single-blinded study. SETTING: University hospital. PARTICIPANTS: One hundred and eighty-four patients undergoing lung resection. MAIN OUTCOME MEASURE: Time required to position the tube from the introduction of the tube into the mouth to confirmation of correct placement in the supine position. RESULTS: Baseline characteristics were well balanced between the groups. Time to place DLTs was similar in both groups: median (interquartile range, IQR) 81.0 (50.0 to 146.2) s for DLTs without a hook and 67.5 s (45.0 to 138.7) for DLTs with a hook (P = 0.43). The incidence of adequate position at the first attempt was 68.5% in the No hook group and 69.6% in the Hook group (P = 0.95). Patients in both groups suffered similar incidences of sore throat at day 0 and day 1 (P = 0.80 and P = 0.20, respectively). No major lesion of the vocal cords or tracheobronchial tree was discovered and the incidence of minor lesions was similar in both groups. CONCLUSION: When a DLT is used, the presence of a carinal hook gives neither advantage nor added complications. TRIAL REGISTRATION: ClinicalTrials.gov, NCT00969683.

Comparison of mannitol and hypertonic saline solution for the treatment of suspected brain herniation during prehospital management of traumatic brain injury patients.

BACKGROUND AND IMPORTANCE: Occurrence of mydriasis during the prehospital management of traumatic brain injury (TBI) may suggest severe intracranial hypertension (ICH) subsequent to brain herniation. The initiation of hyperosmolar therapy to reduce ICH and brain herniation is recommended. Whether mannitol or hypertonic saline solution (HSS) should be preferred is unknown. OBJECTIVES: The objective of this study is to assess whether HSS, compared with mannitol, is associated with improved survival in adult trauma patients with TBI and mydriasis. DESIGN/SETTING AND PARTICIPANTS: A retrospective observational cohort study using the French Traumabase national registry to compare the ICU mortality of patients receiving either HSS or mannitol. Patients aged 16 years or older with moderate to severe TBI who presented with mydriasis during prehospital management were included. OUTCOME MEASURES AND ANALYSIS: We performed propensity score matching on a priori selected variables [i.e. age, sex and initial Coma Glasgow Scale (GCS)] with a ratio of 1 : 3 to ensure comparability between the two groups. The primary outcome was ICU mortality. The secondary outcomes were regression of pupillary abnormality during prehospital management, pulsatility index and diastolic velocity on transcranial Doppler within 24 h after TBI, early ICU mortality (within 48 h), ICU and hospital length of stay. RESULTS: Of 31 579 patients recorded in the registry between 2011 and 2021, 1417 presented with prehospital mydriasis and were included: 1172 (82.7%) received mannitol and 245 (17.3%) received HSS. After propensity score matching, 720 in the mannitol group matched 240 patients in the HSS group. Median age was 41 years [interquartile ranges (IQR) 26-60], 1058 were men (73%) and median GCS was 4 (IQR 3-6). No significant difference was observed in terms of characteristics and prehospital management between the two groups. ICU mortality was lower in the HSS group (45%) than in the mannitol group (54%) after matching [odds ratio (OR) 0.68 (0.5-0.9), P  = 0.014]. No differences were identified between the groups in terms of secondary outcomes. CONCLUSION: In this propensity-matched observational study, the prehospital osmotherapy with HSS in TBI patients with prehospital mydriasis was associated with a lower ICU mortality compared to osmotherapy with mannitol.

Impact of early external ventricular drainage on functional outcome after traumatic brain injury: a bicentric retrospective cohort analysis.

PURPOSE: Several studies have confirmed that external ventricular drain decreases intracranial pressure (ICP) after traumatic brain injury (TBI). Considering its impact on ICP control and cerebral waste metabolites clearance, timing of external ventricular drain (EVD) insertion could improve CSF drainage efficiency. The aim of the study was to evaluate the impact of early EVD versus a later one on the 3-month outcome. METHODS: For this retrospective cohort study conducted in two regional trauma-center (Caen CHU Côte de Nacre and Beaujon Hospital) between May 2011 and March 2019, all patients with intracranial hypertension following TBI and treated with EVD were included. We defined the early EVD by drainage within the 24 h of the hospital admission and the late EVD insertion by drainage beyond 24 h. A poor outcome was defined as a Glasgow Outcome Scale of one or two at 3 months. RESULTS: Among the cohort of 671 patients, we analyzed 127 patients. Sixty-one (48.0%) patients had an early insertion of EVD. In the early EVD group, the mean time to insertion was 10 h versus 55 h in the late EVD group. Among the analyzed patients, 69 (54.3%) had a poor outcome including 39 (63.9%) in the early group and 30 (45.5%) in the later one. After adjustment on prognostic factors, early EVD insertion was not associated with a decrease in a poor outcome at 3-months (OR = 1.80 [0.73-4.53]). CONCLUSION: Early insertion of EVD (<24 h) for intracranial hypertension after TBI was not associated with improved outcome at 3 months.

The Associative Pattern Between Segmental Arterial Damage and Complete Neurological Disorder After Spinal Cord Injury: A Case-Control Study.

INTRODUCTION: The prevalence of vascular trauma surrounding the thoracic spine following Spinal Cord Injury (SCI) is unknown. The potential for neurologic recovery is uncertain in many cases; in some cases, neurologic assessment is not possible, for example, in severe head injury or early intubation, and detection of segmental artery injury may help as a predictive factor. OBJECTIVE: To assess the prevalence of segmental vessel disruption in two groups, with and without neurologic deficit. MATERIAL AND METHODS: This is a retrospective cohort study, with a group SCI American Spinal Injury Association (ASIA) E and a group SCI ASIA A. All patients had a high-energy thoracic or thoracolumbar fracture from T1 to L1. Patients were matched 1:1 (one ASIA A matched with one ASIA E) according to the fracture type, age, and level. The primary variable was the assessment of the presence/disruption of the segmental arteries, bilaterally, around the fracture. Analysis was performed twice by two independent surgeons in a blinded fashion. RESULTS: Both groups had 2 type A, 8 type B, and 4 type C fractures. The right segmental artery was detected in 14/14 (100%) of the patients with ASIA E and in 3/14 (21%) or 2/14 (14%) of the patients with ASIA A, according to the observers, p=0.001. The left segmental artery was detectable in 13/14 (93%) or 14/14 (100%) of the patients ASIA E and in 3/14 (21%) of the patients ASIA A for both observers. All in all, 13/14 of the patients with ASIA A had at least one segmental artery undetectable. The sensibility varied between 78%to 92%, and the specificity from 82% to 100%. The Kappa Score varied between 0.55 and 0.78. CONCLUSION: Segmental arteries disruption was common in the group ASIA A. This may help to predict the neurological status of patients with no complete neurological assessment or potential for recovery post-injury.

Risk factors of second ventilator-associated pneumonia in trauma patients: a retrospective cohort study.

BACKGROUND: Ventilator acquired pneumonia (VAP) is a frequent and serious complication in ICU. Second episodes of VAP are common in trauma patients and may be related to severity of underlying conditions, treatment or bacterial factors of the first VAP. The aim of this study was to identify risk factors of second VAP episodes in trauma injured patients (defined as the development of a new pulmonary infection during or remotely following the first episode). DESIGN: This is a single-center, retrospective cohort study of trauma injured patients who underwent a first episode of VAP between January 1, 2013 and December 31, 2020 at Beaujon Hospital. RESULTS: A total of 533 patients with a first episode of VAP were analyzed, mostly with head and/or thoracic traumatic injury. A second episode of VAP occurred in one hundred sixty-seven patients (31.3%). The main risk factors found was the degree of hypoxemia at the time of the first episode [PaO2/FiO2 ratio 100-200, OR 3.12 (1.77-5.69); < 100, OR 5.80 (2.70-12.8)] and severe traumatic brain injury characterized by an initial GCS ≤ 8 [OR 1.65 (1.01-2.74)]. CONCLUSION: Depth of hypoxemia during the first VAP episode and severity of the initial brain injury are the main risk factors for VAP second episode in trauma injured patients.

External ventricular drainage for intracranial hypertension after traumatic brain injury: is it really useful?

PURPOSES: External ventricular drainage (EVD) is frequently used to control raised intracranial pressure after traumatic brain injury. However, the available evidence about its effectiveness in this context is limited. The aim of this study is to evaluate the effectiveness of EVD to control intracranial pressure and to identify the clinical and radiological factors associated with its success. METHODS: For this retrospective cohort study conducted in a Level 1 traumacenter in Paris area between May 2011 and March 2019, all patients with intracranial hypertension and treated with EVD were included. EVD success was defined as an efficient and continuous control of intracranial hypertension avoiding the use of third tier therapies (therapeutic hypothermia, decompressive craniectomy, and barbiturate coma) or avoiding a decision to withdraw life sustaining treatment due to both refractory intracranial hypertension and severity of brain injury lesions. RESULTS: 83 patients with EVD were included. EVD was successful in 33 patients (40%). Thirty-two patients (39%) required a decompressive craniectomy, eight patients (9%) received barbiturate coma. In ten cases (12%) refractory intracranial hypertension prompted a protocolized withdrawal of care. Complications occurred in nine patients (11%) (three cases of ventriculitis, six cases of catheter occlusion). Multivariate analysis identified no independent factors associated with EVD success. CONCLUSION: In a protocol-based management for traumatic brain injury, EVD allowed intracranial pressure control and avoided third tier therapeutic measures in 40% of cases with a favorable risk-benefit ratio.

Comparison of Injuries Associated With Electric Scooters, Motorbikes, and Bicycles in France, 2019-2022.

Importance: Electric scooter (e-scooter) use is increasing in France and in many urban environments worldwide. Yet little is known about injuries associated with use of e-scooters. Objective: To describe characteristics and outcomes of major trauma involving e-scooters. Design, Setting, and Participants: A multicenter cohort study was conducted in France using the national major trauma registry between January 1, 2019, and December 20, 2022. All patients admitted to a participating major trauma center following a road traffic crash (RTC) involving an e-scooter, a bicycle, or a motorbike were included. Exposure: Included patients were compared according to the 3 mechanisms. Main Outcomes and Measures: The primary outcome was trauma severity as defined by the Injury Severity Score (ISS). Secondary outcomes included the trends of the number of patients per year, a comparison of the RTC epidemiologic factors, injury severity, resources used, and in-hospital outcomes. Results: A total of 5233 patients involved in RTCs were admitted (median age, 33 [IQR, 24-48] years; 4629 [88.5%] men; median ISS, 13 [IQR, 8-22]). The population included 229 e-scooter RTCs (4.4%), 4094 motorbike RTCs (78.2%), and 910 bicycle RTCs (17.4%). The number of patients treated following e-scooter RTCs increased by 2.8-fold in 4 years (from 31 in 2019 to 88 in 2022), while bicycle RTCs increased by 1.2-fold and motorbike RTCs decreased by 0.9-fold. At admission, 36.7% of e-scooter users had a blood alcohol content higher than the legal threshold (n = 84) and 22.5% wore a protective helmet (n = 32). Among e-scooter RTCs, 102 patients (45.5%) had an ISS of 16 or higher. This proportion was similar for patients with motorbike RTCs (1557 [39.7%]; P = .10) and bicycle RTCs (411 [47.3%]; P = .69). With a proportion of 25.9% (n = 50), patients with e-scooter RTCs had twice as many severe traumatic brain injuries (Glasgow Coma Scale ≤8) as motorbike RTCs (445 [11.8%]) and a proportion comparable to bicycle RTCs (174 [22.1%]). The mortality of e-scooter RTCs was 9.2% (n = 20), compared with 5.2% (n = 196) (P = .02) for motorbikes and 10.0% (n = 84) (P = .82) for bicycles. Conclusions and Relevance: The findings of this study suggest that trauma involving e-scooters in France has significantly increased over the past 4 years. These patients presented with injury profiles as severe as those of individuals who experienced bicycle or motorbike RTCs, with a higher proportion of severe traumatic brain injury.

Anxiety and depression symptoms in relatives of moderate-to-severe traumatic brain injury survivors - A multicentre cohort.

INTRODUCTION: The prevalence and risk factors of anxiety and depression symptoms in relatives of moderate to severe traumatic brain injury (TBI) survivors have not been thoroughly investigated. METHODS: Ancillary study of a multicentric prospective randomized-controlled trial in nine university hospitals in 370 moderate-to-severe TBI patients. TBI survivor-relative dyads were included in the 6th month of follow-up. Relatives responded to the Hospital Anxiety and Depression Scale (HADS). The primary endpoints were the prevalence of severe symptoms of anxiety (HADS-Anxiety ≥ 11) and depression (HADS-Depression ≥ 11) in relatives. We explored the risk factors of severe anxiety and depression symptoms. RESULTS: Relatives were predominantly women (80.7%), spouse-husband (47.7%), or parents (39%). Out of the 171 dyads included, 83 (50.6%) and 59 (34.9%) relatives displayed severe symptoms of anxiety and depression, respectively. Severe anxiety symptoms in relatives were independently associated with the patient's discharge at home (OR 2.57, 95%CI [1.04-6.37]) and the patient's higher SF-36 Mental Health domain scores (OR 1.03 95%CI [1.01-1.05]). Severe depression symptoms were independently associated with a lower SF-36 Mental Health domain score (OR = 0.98 95%CI [0.96-1.00]). No ICU organization characteristics were associated with psychological symptoms in relatives. DISCUSSION: There is a high prevalence of anxiety and depression symptoms among relatives of moderate-to-severe TBI survivors at 6 months. Anxiety and depression were inversely correlated with the patient's mental health status at 6 months. CONCLUSIONS: Long-term follow-up must provide psychological care to relatives after TBI.

Massive transfusion in trauma: an evolving paradigm.

A considerable amount of literature has nurtured the idea that massive transfusion is an independent trauma disease and therapeutic tool. In this opinion paper, the authors expose the evolution and challenge the classic paradigm and historic definition of massive transfusion. Based on current evidence the elements of an evolving strategy in transfusion management and bleeding control are exposed to use of tranexamic acid, combination and ratios of blood products, use of fluids and viscoelastic testing, etc. The synergy of these elements provides the basis to develop updated strategies and perspectives for transfusion management after trauma and to consider a classic definition of massive transfusion as outdated or the need for massive transfusion as failure. An alternative concept, time critical transfusion may be better placed to take into account modern transfusion management after trauma.

Empiric antimicrobial therapy for early-onset pneumonia in severe trauma patients.

PURPOSE: The bacterial ecology involved in early pneumonia of severe trauma patients is mostly commensal and would allow wide use of narrow-spectrum antibiotics. We describe risk factors for treatment failure of severe trauma patients' pneumonia with the use of narrow-spectrum antimicrobial therapy in order to develop a score that could help clinicians to determine which patients might be treated with narrow-spectrum antibiotics. METHODS: A retrospective, observational, monocentric cohort study was conducted of severe trauma patients requiring mechanical ventilation for > 48 h and developing a first episode of microbiologically confirmed pneumonia occurring within the first 10 days after admission. RESULTS: Overall, 370 patients were included. The resistance rate against narrow-spectrum antibiotics (amoxicillin/clavulanic acid) was 22.7% (84 pneumonia). In a multivariate analysis, two independent risk factors were associated with this resistance: prior antimicrobial therapy ≥ 48 h (OR 4.00; 95 CI [2.39; 6.75]) and age ≥ 30y (OR 2.10; 95 CI [1.21; 3.78]). We created a prediction score that defined patient with one or two risk factors at high risk of resistance. This score presented a sensitivity of 0.92 [0.88; 0.94], a specificity of 0.33 [0.28; 0.38], a positive predictive value of 0.29 [0.24; 0.33] and a negative predictive value of 0.93 [0.90; 0.95]. CONCLUSION: Simple risk factors may help clinicians to identify severe trauma patients at high risk of pneumonia treatment failure with the use of narrow-spectrum antimicrobial therapy and, thus, use better tailored empiric therapy and limit the use of unnecessary broad-spectrum antimicrobial therapy.

Machine learning-based prediction of emergency neurosurgery within 24 h after moderate to severe traumatic brain injury.

BACKGROUND: Rapid referral of traumatic brain injury (TBI) patients requiring emergency neurosurgery to a specialized trauma center can significantly reduce morbidity and mortality. Currently, no model has been reported to predict the need for acute neurosurgery in severe to moderate TBI patients. This study aims to evaluate the performance of Machine Learning-based models to establish to predict the need for neurosurgery procedure within 24 h after moderate to severe TBI. METHODS: Retrospective multicenter cohort study using data from a national trauma registry (Traumabase®) from November 2011 to December 2020. Inclusion criteria correspond to patients over 18 years old with moderate or severe TBI (Glasgow coma score ≤ 12) during prehospital assessment. Patients who died within the first 24 h after hospital admission and secondary transfers were excluded. The population was divided into a train set (80% of patients) and a test set (20% of patients). Several approaches were used to define the best prognostic model (linear nearest neighbor or ensemble model). The Shapley Value was used to identify the most relevant pre-hospital variables for prediction. RESULTS: 2159 patients were included in the study. 914 patients (42%) required neurosurgical intervention within 24 h. The population was predominantly male (77%), young (median age 35 years [IQR 24-52]) with severe head injury (median GCS 6 [3-9]). Based on the evaluation of the predictive model on the test set, the logistic regression model had an AUC of 0.76. The best predictive model was obtained with the CatBoost technique (AUC 0.81). According to the Shapley values method, the most predictive variables in the CatBoost were a low initial Glasgow coma score, the regression of pupillary abnormality after osmotherapy, a high blood pressure and a low heart rate. CONCLUSION: Machine learning-based models could predict the need for emergency neurosurgery within 24 h after moderate and severe head injury. Potential clinical benefits of such models as a decision-making tool deserve further assessment. The performance in real-life setting and the impact on clinical decision-making of the model requires workflow integration and prospective assessment.

Long-term outcome following blunt cerebrovascular injuries: occurrence of ischemic complications, treatment, and outcome.

PURPOSE: Blunt cerebrovascular injuries (BCVI) are a rare but serious complication after trauma. Among patients with BCVI, neurological status is altered in 30% of cases and the prognosis seems to be associated with ischemic complications. The aim of this study was to assess the long-term outcome of BCVI-associated ischemic events. METHODS: This retrospective cohort study (2011-2017) included all patients admitted for severe trauma with identified BCVI in two level-1 trauma centers. Patients were considered to have a poor neurological outcome with a GOS-E between 2 and 5 and a good neurological outcome with GOS-E between 6 and 8. A multivariate logistic regression identified risk factors for poor neurological outcome at 1 year. RESULTS: Of the 6,294 patients admitted in both trauma centers between 2011 and 2017, 81 patients presenting BCVI were identified (incidence of 1.3%). The median age was 35 years (24-44) with a median Injury Severity Score of 28 (17-41). 29 patients (50%) had a good neurological prognosis, while 25 patients (43%) had a poor neurological prognosis at 1 year. Ischemic stroke occurred in 11 patients (13.6%) within a median time of 2 days (2-2.5). No ischemic stroke occurred in the first year after ICU discharge in both groups. In our study, good prognosis at 1 year was not associated with ischemic complications [3 (10) vs 3 (12) p = 1]. CONCLUSION: Ischemic complications after BCVI are rare, occur within the first week and do not seem to impact independently the 1-year neurological prognosis.

ISS is not an appropriate tool to estimate overtriage.

PURPOSE: The aim of this work is to study a cohort of patients of ISS < 15 admitted to a TC, and to determine the number of patients that ultimately benefited from the skills and resources specific of a level 1 trauma center. METHODS: Retrospective study from a prospective cohort of patients admitted to TC (Beaujon Hospital, APHP) for suspected severe trauma from January 2011 to December 2017. The main outcome criterion was the use of surgery or interventional radiology within the first 24 h after admission of patients with ISS < 15. The secondary outcomes were stratified into severe (mortality, resuscitation care, length of stay in intensive care units) and non-severe criteria (mild head injury, hospital discharge or transfer within 24 h). RESULTS: Of 3035 patients admitted during the study period, 1409 with an ISS < 15 were included, corresponding to a theoretical overtriage rate of 46.4%. Among these, 611 patients (43.4%) underwent emergency intervention within the first 24 h (586 surgical interventions, 19 direct transfers to the operating theater and 6 acts of interventional radiology), 238 (16.9%) of patients presented with severe and 531 (38%) with non-severe outcome criteria. CONCLUSION: This work demonstrates that in a cohort of patients classified as ISS < 15 admitted to a TC, a considerable amount of TC-specific resources are required, and patients present with severe outcome criteria despite being classified as overtriaged. These results suggest that triage of trauma patients should be based on resource use and clinical outcome rather than anatomic criteria.