Effects of PEEP on intracranial pressure in patients with acute brain injury: An observational, prospective and multicenter study.

OBJECTIVE: To analyze the impact of positive end-expiratory pressure (PEEP) changes on intracranial pressure (ICP) dynamics in patients with acute brain injury (ABI). DESIGN: Observational, prospective and multicenter study (PEEP-PIC study). SETTING: Seventeen intensive care units in Spain. PATIENTS: Neurocritically ill patients who underwent invasive neuromonitorization from November 2017 to June 2018. INTERVENTIONS: Baseline ventilatory, hemodynamic and neuromonitoring variables were collected immediately before PEEP changes and during the following 30 min. MAIN VARIABLES OF INTEREST: PEEP and ICP changes. RESULTS: One-hundred and nine patients were included. Mean age was 52.68 (15.34) years, male 71 (65.13%). Traumatic brain injury was the cause of ABI in 54 (49.54%) patients. Length of mechanical ventilation was 16.52 (9.23) days. In-hospital mortality was 21.1%. PEEP increases (mean 6.24-9.10 cmH2O) resulted in ICP increase from 10.4 to 11.39 mmHg, P < .001, without changes in cerebral perfusion pressure (CPP) (P = .548). PEEP decreases (mean 8.96 to 6.53 cmH2O) resulted in ICP decrease from 10.5 to 9.62 mmHg (P = .052), without changes in CPP (P = .762). Significant correlations were established between the increase of ICP and the delta PEEP (R = 0.28, P < .001), delta driving pressure (R = 0.15, P = .038) and delta compliance (R = -0.14, P = .052). ICP increment was higher in patients with lower baseline ICP. CONCLUSIONS: PEEP changes were not associated with clinically relevant modifications in ICP values in ABI patients. The magnitude of the change in ICP after PEEP increase was correlated with the delta of PEEP, the delta driving pressure and the delta compliance.

Current perspectives on the use of artificial intelligence in critical patient safety.

Intensive Care Units (ICUs) have undergone enhancements in patient safety, and artificial intelligence (AI) emerges as a disruptive technology offering novel opportunities. While the published evidence is limited and presents methodological issues, certain areas show promise, such as decision support systems, detection of adverse events, and prescription error identification. The application of AI in safety may pursue predictive or diagnostic objectives. Implementing AI-based systems necessitates procedures to ensure secure assistance, addressing challenges including trust in such systems, biases, data quality, scalability, and ethical and confidentiality considerations. The development and application of AI demand thorough testing, encompassing retrospective data assessments, real-time validation with prospective cohorts, and efficacy demonstration in clinical trials. Algorithmic transparency and explainability are essential, with active involvement of clinical professionals being crucial in the implementation process.

¿Podrían ayudarnos los algoritmos de machine learning en la predicción de hemorragia masiva a nivel prehospitalario? / Could machine learning algorithms help us in predicting massive hemorrhage at the prehospital level?

Objetivo: Comparación de la capacidad predictiva de diferentes algoritmos de machine learning (AML) respecto a escalas tradicionales de predicción de hemorragia masiva en pacientes con enfermedad traumática grave (ETG). Diseño: Sobre una base de datos de una cohorte retrospectiva con variables clínicas prehospitalarias y de resultado de hemorragia masiva se realizó un tratamiento de la base de datos para poder aplicar los AML, obteniéndose un conjunto total de 473 pacientes (80% entrenamiento, 20% validación). Para la modelización se realizó imputación proporcional y validación cruzada. El poder predictivo se evaluó con la métrica ROC y la importancia de las variables mediante los valores Shapley. Ámbito: Atención extrahospitalaria del paciente con ETG. Pacientes: Pacientes con ETG atendidos en el medio extrahospitalario por un servicio médico extrahospitalario desde enero de 2010 hasta diciembre de 2015 y trasladados a un centro de trauma en Madrid. Intervenciones: Ninguna. Variables de interés principales: Obtención y comparación de la métrica ROC de 4 AML: random forest, support vector machine, gradient boosting machine y neural network con los resultados obtenidos con escalas tradicionales de predicción. Resultados: Los diferentes AML alcanzaron valores ROC superiores al 0,85, teniendo medianas cercanas a 0,98. No encontramos diferencias significativas entre los AML. Cada AML ofrece un conjunto de variables diferentes, pero con predominancia de las variables hemodinámicas, de reanimación y de deterioro neurológico. Conclusiones: Los AML podrían superar a las escalas tradicionales de predicción en la predicción de hemorragia masiva. (AU)

Objective: Comparison of the predictive ability of various machine learning algorithms (MLA) versus traditional prediction scales for massive hemorrhage in patients with severe traumatic injury (ETG). Design: On a database of a retrospective cohort with prehospital clinical variables and massive hemorrhage outcome, a treatment of the database was performed to be able to apply the different MLA, obtaining a total set of 473 patients (80% training and 20% validation). For modeling, proportional imputation and cross validation were performed. The predictive power was evaluated with the ROC metric and the importance of the variables using the Shapley values. Setting: Out-of-hospital care of patients with ETG. Participants: Patients with ETG treated out-of-hospital by a prehospital medical service from January 2010 to December 2015 and transferred to a trauma center in Madrid. Interventions: None. Main variables of interest: Obtaining and comparing the ROC curve metric of 4 MLAs: random forest, support vector machine, gradient boosting machine and neural network with the results obtained with traditional prediction scales. Results: The different MLA reached ROC values higher than 0.85, having medians close to 0.98. We found no significant differences between MLAs. Each MLA offers a different set of more important variables with a predominance of hemodynamic, resuscitation variables and neurological impairment. Conclusions: MLA may be helpful in patients with massive hemorrhage by outperforming traditional prediction scales. (AU)

Could machine learning algorithms help us predict massive bleeding at prehospital level?

OBJECTIVE: Comparison of the predictive ability of various machine learning algorithms (MLA) versus traditional prediction scales (TPS) for massive hemorrhage (MH) in patients with severe traumatic injury (STI). DESIGN: On a database of a retrospective cohort with prehospital clinical variables and MH outcome, a treatment of the database was performed to be able to apply the different AML, obtaining a total set of 473 patients (80% training, 20% validation). For modeling, proportional imputation and cross validation were performed. The predictive power was evaluated with the ROC metric and the importance of the variables using the Shapley values. SETTING: Out-of-hospital care of patients with STI. PARTICIPANTS: Patients with STI treated out-of-hospital by a out-of-hospital medical service from January 2010 to December 2015 and transferred to a trauma center in Madrid. INTERVENTIONS: None. MAIN VARIABLES OF INTEREST: Obtaining and comparing the "Receiver Operating Characteristic curve" (ROC curve) metric of four MLAs: "random forest" (RF), "vector support machine" (SVM), "gradient boosting machine" (GBM) and "neural network" (NN) with the results obtained with TPS. RESULTS: The different AML reached ROC values higher than 0.85, having medians close to 0.98. We found no significant differences between AMLs. Each AML offers a different set of more important variables with a predominance of hemodynamic, resuscitation variables and neurological impairment. CONCLUSIONS: MLA may be helpful in patients with HM by outperforming TPS.

Associated Risk Factors and Impact in Clinical Outcomes of Multiorgan Failure in Patients with TBI.

BACKGROUND: Individual extracerebral organ dysfunction is common after severe traumatic brain injury (TBI) and impacts outcomes. However, multiorgan failure (MOF) has received less attention in patients with isolated TBI. Our objective was to analyze the risk factors associated with the development of MOF and its impact in clinical outcomes in patients with TBI. METHODS: This was an observational, prospective, multicenter study using data from a nationwide registry that currently includes 52 intensive care units (ICUs) in Spain (RETRAUCI). Isolated significant TBI was defined as Abbreviated Injury Scale (AIS) ≥ 3 in the head area with no AIS ≥ 3 in any other anatomical area. Multiorgan failure was defined using the Sequential-related Organ Failure Assessment as the alteration of two or more organs with a score of ≥ 3. We analyzed the contribution of MOF to crude and adjusted mortality (age and AIS head) by using logistic regression analysis. A multiple logistic regression analysis was performed to analyze the risk factors associated with the development of MOF in patients with isolated TBI. RESULTS: A total of 9790 patients with trauma were admitted to the participating ICUs. Of them, 2964 (30.2%) had AIS head ≥ 3 and no AIS ≥ 3 in any other anatomical area, and these patients constituted the study cohort. Mean age was 54.7 (19.5) years, 76% of patients were men, and ground-level falls were the main mechanism of injury (49.1%). In-hospital mortality was 22.2%. Up to 185 patients with TBI (6.2%) developed MOF during their ICU stay. Crude and adjusted (age and AIS head) mortality was higher in patients who developed MOF (odds ratio 6.28 [95% confidence interval 4.58-8.60] and odds ratio 5.20 [95% confidence interval 3.53-7.45]), respectively. The logistic regression analysis showed that age, hemodynamic instability, the need of packed red blood cells concentrates in the initial 24 h, the severity of brain injury, and the need for invasive neuromonitoring were significantly associated with MOF development. CONCLUSIONS: MOF occurred in 6.2% of patients with TBI admitted to the ICU and was associated with increased mortality. MOF was associated with age, hemodynamic instability, the need of packed red blood cells concentrates in the initial 24 h, the severity of brain injury, and the need for invasive neuromonitoring.

Mentoring Programme: Guiding your first steps towards scientific excellence.

SEMICYUC's first Mentoring Programme aims to support the research careers of the Society's youngest members. Added benefits include acquiring new research and/or clinical skills, increasing the ability of critical thought, and fostering the development of the next generation of research leaders. This project would not be possible without the exceptional team of mentors or research experts willing to embark on the journey with the young trainees. This article sets out the foundations of such a programme and proposes future changes for continuous improvement.

External validation of the Glasgow Coma Scale-Pupils in patients with severe head injury. / Validación externa de la Escala de Coma de Glasgow con valoración pupilar en pacientes con traumatismo craneoencefálico grave.

OBJECTIVES: To compare the ability of the Glasgow Coma Scale (GCS) score, the GCS Pupils (GCS-P) score, and the Pupil Reactivity Score (PRS) to predict mortality in patients with severe head injury. MATERIAL AND METHODS: Retrospective analysis of all patients with severe head injury and initial GCS scores of 8 or lower on initial evaluation for whom records included pupil dilation information and clinical course after admission to intensive care units of participating hospitals. We assessed the ability of each of the 3 scores (GCS, GCS-P, and PRS) to predict mortality using discrimination analysis. Discrimination was estimated by calculating the areas under the receiver operating characteristic curves (AUC) and 95% CIs. RESULTS: A total of 1551 patients with severe head injury and pupil dilation records were studied. The mean age was 50 years, 1190 (76.7%) were males, and 592 (38.2%) died. No pupil dilation was observed in 905 patients (58.3%), 362 (23.3%) had unilateral mydriasis, and 284 (18.3%) had bilateral mydriasis. The GCS-P score was significantly better at predicting mortality, with an AUC of 0.77 (95% CI, 0.74-0.79), versus 0.69 (95% CI, 0.67-0.72) for the GCS, and 0.75 (95% CI, 0.72-0.77) for the PRS. As the GCS-P score decreased, mortality increased. CONCLUSION: The GCS-P was more useful than the GCS for predicting death after severe head injury.

OBJETIVO: Analizar la capacidad para predecir la mortalidad hospitalaria de la Escala de Coma de Glasgow con valoración pupilar (GCS-P) comparado con la Escala de Coma de Glasgow (GCS) y con la escala de reactividad pupilar (PRS) en pacientes con traumatismo craneoencefálico (TCE) grave. METODO: Análisis retrospectivo de cohortes de todos los pacientes con TCE, puntuación en la GCS # 8 en la atención inicial, datos de exploración pupilar inicial y del desenlace hospitalario ingresados en las unidades de cuidados intensivos participantes. Se determinó la capacidad predictiva de mortalidad de la GCS, PRS y la GCS-P mediante un análisis de discriminación. La discriminación se analizó empleando curvas operativas del receptor (COR), el área bajo la curva (ABC) y su intervalo de confianza del 95% (IC 95%). RESULTADOS: Se analizaron 1.551 pacientes con TCE grave y datos sobre exploración pupilar. La edad media fue de 50 años, 1.190 (76,7%) eran hombres, y hubo 592 (38,2%) defunciones. Hubo 905 (58,3%) pacientes sin alteraciones pupilares, 362 (23,3%) con midriasis unilateral y 284 (18,3%) pacientes con midriasis bilateral. El análisis del ABCCOR para predecir la mortalidad hospitalaria mostró de forma significativa una mejor capacidad predictiva del GCS-P con ABC = 0,77 (IC 95% 0,74-0,79) respecto al GCS con ABC = 0,69 (IC 95% 0,67-0,72). La reactividad pupilar mostró un ABC = 0,75 (IC 95% 0,72-0,77). Se observó un incremento de mortalidad con la disminución del GCS-P. CONCLUSIONES: La escala GCS-P presentó mejor rendimiento que la GCS para predecir mortalidad en el TCE grave.

Validación externa de la Escala de Coma de Glasgow con valoración pupilar en pacientes con traumatismo craneoencefálico grave / External validation of the Glasgow Coma Scale-Pupils in patients with severe head injury

Objetivos. Analizar la capacidad para predecir la mortalidad hospitalaria de la Escala de Coma de Glasgow con valoración pupilar (GCS-P) comparado con la Escala de Coma de Glasgow (GCS) y con la escala de reactividad pupilar (PRS) en pacientes con traumatismo craneoencefálico (TCE) grave. Métodos. Análisis retrospectivo de cohortes de todos los pacientes con TCE, puntuación en la GCS # 8 en la atención inicial, datos de exploración pupilar inicial y del desenlace hospitalario ingresados en las unidades de cuidados intensivos participantes. Se determinó la capacidad predictiva de mortalidad de la GCS, PRS y la GCS-P mediante un análisis de discriminación. La discriminación se analizó empleando curvas operativas del receptor (COR), el área bajo la curva (ABC) y su intervalo de confianza del 95% (IC 95%). Resultados. Se analizaron 1.551 pacientes con TCE grave y datos sobre exploración pupilar. La edad media fue de 50 años, 1.190 (76,7%) eran hombres, y hubo 592 (38,2%) defunciones. Hubo 905 (58,3%) pacientes sin alteraciones pupilares, 362 (23,3%) con midriasis unilateral y 284 (18,3%) pacientes con midriasis bilateral. El análisis del ABCCOR para predecir la mortalidad hospitalaria mostró de forma significativa una mejor capacidad predictiva del GCS-P con ABC = 0,77 (IC 95% 0,74-0,79) respecto al GCS con ABC = 0,69 (IC 95% 0,67-0,72). La reactividad pupilar mostró un ABC = 0,75 (IC 95% 0,72-0,77). Se observó un incremento de mortalidad con la disminución del GCS-P. Conclusiones. La escala GCS-P presentó mejor rendimiento que la GCS para predecir mortalidad en el TCE grave. (AU)

Objectives. To compare the ability of the Glasgow Coma Scale (GCS) score, the GCS Pupils (GCS-P) score, and the Pupil Reactivity Score (PRS) to predict mortality in patients with severe head injury. Methods. Retrospective analysis of all patients with severe head injury and initial GCS scores of 8 or lower on initial evaluation for whom records included pupil dilation information and clinical course after admission to intensive care units of participating hospitals. We assessed the ability of each of the 3 scores (GCS, GCS-P, and PRS) to predict mortality using discrimination analysis. Discrimination was estimated by calculating the areas under the receiver operating characteristic curves (AUC) and 95% CIs. Results. A total of 1551 patients with severe head injury and pupil dilation records were studied. The mean age was 50 years, 1190 (76.7%) were males, and 592 (38.2%) died. No pupil dilation was observed in 905 patients (58.3%), 362 (23.3%) had unilateral mydriasis, and 284 (18.3%) had bilateral mydriasis. The GCS-P score was significantly better at predicting mortality, with an AUC of 0.77 (95% CI, 0.74-0.79), versus 0.69 (95% CI, 0.67-0.72) for the GCS, and 0.75 (95% CI, 0.72-0.77) for the PRS. As the GCS-P score decreased, mortality increased. Conclusion. The GCS-P was more useful than the GCS for predicting death after severe head injury. (AU)

Traumatic Brain Injury and Acute Kidney Injury-Outcomes and Associated Risk Factors.

Our objective was to analyze the contribution of acute kidney injury (AKI) to the mortality of isolated TBI patients and its associated risk factors. Observational, prospective and multicenter registry (RETRAUCI) methods were used, from March 2015 to December 2019. Isolated TBI was defined as abbreviated injury scale (AIS) ≥ 3 head with no additional score ≥ 3. A comparison of groups was conducted using the Wilcoxon test, chi-square test or Fisher's exact test, as appropriate. A multiple logistic regression analysis was conducted to analyze associated risk factors in the development of AKI. For the result, overall, 2964 (30.2%) had AIS head ≥ 3 with no other area with AIS ≥ 3. The mean age was 54.7 (SD 19.5) years, 76% were men, and the ground-level falls was 49.1%. The mean ISS was 18.4 (SD 8). The in-hospital mortality was 22.2%. Up to 310 patients (10.6%) developed AKI, which was associated with increased mortality (39% vs. 17%, adjusted OR 2.2). Associated risk factors (odds ratio (OR) (95% confidence interval)) were age (OR 1.02 (1.01-1.02)), hemodynamic instability (OR 2.87 to OR 5.83 (1.79-13.1)), rhabdomyolysis (OR 2.94 (1.69-5.11)), trauma-associated coagulopathy (OR 1.67 (1.05-2.66)) and transfusion of packed red-blood-cell concentrates (OR 1.76 (1.12-2.76)). In conclusion, AKI occurred in 10.6% of isolated TBI patients and was associated with increased mortality.

Chronic critical illness in polytrauma. Results of the Spanish trauma in ICU registry.

PURPOSE: Chronic critical illness after trauma injury has not been fully evaluated, and there is little evidence in this regard. We aim to describe the prevalence and risk factors of chronic critical illness (CCI) in trauma patients admitted to the intensive care unit. MATERIAL AND METHODS: Retrospective observational multicenter study (Spanish Registry of Trauma in ICU (RETRAUCI)). Period March 2015 to December 2019. Trauma patients admitted to the ICU, who survived the first 48 h, were included. Chronic critical illness (CCI) was considered as the need for mechanical ventilation for a period greater than 14 days and/or placement of a tracheostomy. The main outcomes measures were prevalence and risk factors of CCI after trauma. RESULTS: 1290/9213 (14%) patients developed CCI. These patients were older (51.2 ± 19.4 vs 49 ± 18.9); p < .01) and predominantly male (79.9%). They presented a higher proportion of infectious complications (81.3% vs 12.7%; p < .01) and multiple organ dysfunction syndrome (MODS) (27.02% vs 5.19%; p < .01). CCI patients required longer stays in the ICU and had higher ICU and overall in-hospital mortality. Age, injury severity score, head injury, infectious complications, and development of MODS were independent predictors of CCI. CONCLUSION: CCI in trauma is a prevalent entity in our series. Early identification could facilitate specific interventions to change the trajectory of this process.

Risk Factors Associated with Mortality in Severe Chest Trauma Patients Admitted to the ICU.

Our objective was to determine outcomes of severe chest trauma admitted to the ICU and the risk factors associated with mortality. An observational, prospective, and multicenter registry of trauma patients admitted to the participating ICUs (March 2015-December 2019) was utilized to collect the patient data that were analyzed. Severe chest trauma was defined as an Abbreviated Injury Scale (AIS) value of ≥3 in the thoracic area. Logistic regression analysis was used to evaluate the contribution of severe chest trauma to crude and adjusted ORs for mortality and to analyze the risk factors associated with mortality. Overall, 3821 patients (39%) presented severe chest trauma. The sample's characteristics were as follows: a mean age of 49.88 (19.21) years, male (77.6%), blunt trauma (93.9%), a mean ISS of 19.9 (11.6). Crude and adjusted (for age and ISS) ORs for mortality in severe chest trauma were 0.78 (0.68-0.89) and 0.43 (0.37-0.50) (p < 0.001), respectively. In-hospital mortality in the severe chest trauma patients without significant traumatic brain injury (TBI) was 5.63% and was 25.71% with associated significant TBI (p < 0.001). Age, the severity of injury (NISS and AIS-head), hemodynamic instability, prehospital intubation, acute kidney injury, and multiorgan failure were risk factors associated with mortality. The contribution of severe chest injury to the mortality of trauma patients admitted to the ICU was very low. Risk factors associated with mortality were identified.

Elevated monocyte distribution width in trauma: An early cellular biomarker of organ dysfunction.

INTRODUCTION: Traumatic injury elicits an inflammatory response such as the one occurring during systemic infection. Monocyte distribution width (MDW) has been found to distinguish sepsis in a pool of patients with suspected infection. We hypothesized that an elevated MDW in trauma patients would be associated with the development of multiple organ dysfunction syndrome (MODS) and an increased mortality. MATERIALS AND METHODS: Observational study in a dedicated trauma Intensive Care Unit (ICU) in Madrid during 2019-2020. Patients were classified according to their first MDW value on admission, as greater or lesser than 21 U. Clinical data was obtained and univariate and multivariate analysis were realized, as well as a test performance analysis. RESULTS: 354 patients were studied, with a median age of 46 years, 78% male. Half presented with severe trauma ISS > 15, mostly with a blunt mechanism of injury. A MDW ≥ 21 U on admission was found in 17% of cases. These patients were more likely to present with hemodynamic instability and MODS. They had a higher length of stay (3.8 vs 2 days) and higher mortality (21 vs 5%) compared to the low MDW group. These findings remained statistically significant in the multivariate analysis, with an OR 4.6 (IC 95% 1.7-12) for MODS and 3.1 (IC 95% 1.2-8.3) for mortality. CONCLUSIONS: In trauma patients, a MDW ≥ 21 U on admission was independently associated with a greater risk of MODS, a higher mortality and a higher length of stay. This biomarker could be useful in predicting severity in the initial evaluation of trauma patients.

Development of a new score for early mortality prediction in trauma ICU patients: RETRASCORE.

BACKGROUND: Severity scores are commonly used for outcome adjustment and benchmarking of trauma care provided. No specific models performed only with critically ill patients are available. Our objective was to develop a new score for early mortality prediction in trauma ICU patients. METHODS: This is a retrospective study using the Spanish Trauma ICU registry (RETRAUCI) 2015-2019. Patients were divided and analysed into the derivation (2015-2017) and validation sets (2018-2019). We used as candidate variables to be associated with mortality those available in RETRAUCI that could be collected in the first 24 h after ICU admission. Using logistic regression methodology, a simple score (RETRASCORE) was created with points assigned to each selected variable. The performance of the model was carried out according to global measures, discrimination and calibration. RESULTS: The analysis included 9465 patients: derivation set 5976 and validation set 3489. Thirty-day mortality was 12.2%. The predicted probability of 30-day mortality was determined by the following equation: 1/(1 + exp (- y)), where y = 0.598 (Age 50-65) + 1.239 (Age 66-75) + 2.198 (Age > 75) + 0.349 (PRECOAG) + 0.336 (Pre-hospital intubation) + 0.662 (High-risk mechanism) + 0.950 (unilateral mydriasis) + 3.217 (bilateral mydriasis) + 0.841 (Glasgow ≤ 8) + 0.495 (MAIS-Head) - 0.271 (MAIS-Thorax) + 1.148 (Haemodynamic failure) + 0.708 (Respiratory failure) + 0.567 (Coagulopathy) + 0.580 (Mechanical ventilation) + 0.452 (Massive haemorrhage) - 5.432. The AUROC was 0.913 (0.903-0.923) in the derivation set and 0.929 (0.918-0.940) in the validation set. CONCLUSIONS: The newly developed RETRASCORE is an early, easy-to-calculate and specific score to predict in-hospital mortality in trauma ICU patients. Although it has achieved adequate internal validation, it must be externally validated.

Factores asociados con la mortalidad en pacientes traumáticos con puntuación de 3 en la escala de coma de Glasgow y midriasis bilateral arreactiva / Factors associated with death due to trauma in patients with a Glasgow Coma Scale score of 3 and bilateral fixed dilated pupils

Objetivo. Comparar los pacientes traumáticos con una puntuación de 3 en la escala de coma de Glasgow (Glasgow Coma Scale, GCS) en función de la reactividad pupilar e investigar los factores asociados a la mortalidad hospitalaria en los pacientes con GCS 3 y midriasis bilateral arreactiva.Método. Estudio observacional, prospectivo y multicéntrico. Se incluyeron todos los pacientes traumáticos recogidos con GCS 3 ingresados en las unidades de cuidados intensivos (UCI) participantes desde marzo 2015 hasta diciembre 2019. Se realizó un análisis de regresión logística para el estudio de los factores asociados a la mortalidad hospitalaria en pacientes con GCS 3 puntos y midriasis bilateral arreactiva.Resultados. De los 933 pacientes con GCS 3 puntos, 454 (48,7%) presentaron pupilas simétricas y reactivas, 201 (21,5%) anisocoria arreactiva y 278 (29,8%) midriasis bilateral arreactiva. La mortalidad hospitalaria fue elevada en los 3 grupos: 32,5% con pupilas normales, 54,6% con anisocoria arreactiva y 91,0% con midriasis bilateral arreactiva. La edad, la puntación de 3 o más en el Abbreviated Injury Scale (cabeza) y el shock o shock refractario se aso-ciaron de forma significativa con la mortalidad hospitalaria, siendo la lesión difusa tipo I y II y la lesión masa evacuada factores protectores en los pacientes con 3 puntos en la GCS y midriasis bilateral arreactiva. De los 26 pacientes que sobrevivieron con GCS 3 y midriasis bilateral arreactiva, 12 (46,1%) tuvieron un GCS de 14-15 al alta hospitalaria.Conclusiones. La mortalidad hospitalaria de los pacientes traumáticos con 3 puntos en la GCS y midriasis bilateral arreactiva fue del 91%. La edad, la puntación de 3 o más en el Abbreviated Injury Scale (cabeza) y el shock o shock refractario se asociaron de forma significativa con la mortalidad hospitalaria, siendo la lesión difusa tipo I y II, y la lesión masa evacuada factores protectores en este grupo de pacientes. (AU)

Objectives. To compare patients with a Glasgow Coma Scale (GCS) score of 3 stratified according to pupillary reaction and to explore factors associated with in-hospital death in those with bilateral fixed dilated pupils.Methods. Prospective, observational, multicenter study. We included all patients with trauma and GCS scores of 3 admitted to the intensive care unit from March 2015 to December 2019. Factors associated with in-hospital mortality in the patients with bilateral dilated pupils were explored using multiple regression analysis.Results. Of the 933 patients included, 454 (48.7%) had responsive pupils, 201 (21.5%) had a single fixed dilated pupil, and 278 (29.8%) had bilateral dilation. Hospital mortality was high in all 3 groups: 32.5% in those with normal responsive pupils, 54.6% in those with a single unreactive pupil, and 91.0% in those with bilateral dilation. Factors significantly associated with in-hospital death were age, a score of 3 or more on the Abbreviated Injury Scale for the head, and shock or refractory shock. Types I or II diffuse lesions and evacuated mass lesions were protective in patients with GCS scores of 3 and bilateral dilated pupils. Twelve of the 26 patients (46.1%) with bilateral dilated pupils and GCS scores of 3 had GCS scores of 14 or 15 on discharge from the hospital.Conclusions. The in-hospital mortality was 91% in this study of trauma patients with GCS scores of 3 and bilateral dilated pupils. Factors significantly associated with in-hospital death were age, a score of 3 or more on the Abbreviated Injury Scale for the head, and shock or refractory shock. Types I or II diffuse lesions and evacuated mass lesions were protective in patients with GCS scores of 3 and bilateral dilated pupils. (AU)

Dexamethasone for the treatment of traumatic brain injured patients with brain contusions and pericontusional edema: Study protocol for a prospective, randomized and double blind trial.

BACKGROUND: Traumatic brain injury (TBI) constitutes a leading cause of death and disability. Patients with TBI and cerebral contusions developing pericontusional edema are occasionally given dexamethasone on the belief that this edema is similar to that of tumors, in which the beneficial effect of dexamethasone has been demonstrated. METHODS: The DEXCON TBI trial is a multicenter, pragmatic, randomized, triple-blind, placebo controlled trial to quantify the effects of dexamethasone on the prognosis of TBI patients with brain contusions and pericontusional edema. Adult patients who fulfill the elegibility criteria will be randomized to dexamethasone/placebo in a short and descending course: 4 mg/6 h (2 days); 4 mg/8 hours (2 days); 2 mg/6 hours (2 days); 2 mg/8 hours (2 days); 1 mg/8 hours (2 days); 1 mg/12 hours (2 days). The primary outcome is the Glasgow Scale Outcome Extended (GOSE) performed 1 month and 6 months after TBI. Secondary outcomes are: number of episodes of neurological deterioration; symptoms associated with TBI; adverse events; volume of pericontusional edema before and after 12 days of treatment; results of the neuropsychological tests one month and 6 months after TBI. The main analysis will be on an "intention-to-treat" basis. Logistic regression will estimate the effect of dexamethasone/placebo on GOSE at one month and at 6 months, dichotomized in unfavorable outcome (GOSE 1-6) and favorable outcome (GOSE 7-8). Efficacy will also be analyzed using the 'sliding dichotomy'. An interim and safety analysis will be performed including patients recruited during the first year to calculate the conditional power. A study with 600 patients would have 80% power (2 sided alpha = 5%) to detect a 12% absolute increase (from 50% to 62%) in good recovery. DISCUSSION: This is a confirmative trial to elucidate the therapeutic efficacy of dexamethasone in a very specific group of TBI patients: patients with brain contusions and pericontusional edema. This trial could become an important milestone for TBI patients as nowadays there is no effective treatment in this type of patients. TRIAL REGISTRATION: eudraCT: 2019-004038-41; Clinical Trials.gov: NCT04303065.

Machine learning techniques for mortality prediction in critical traumatic patients: anatomic and physiologic variables from the RETRAUCI study.

BACKGROUND: Interest in models for calculating the risk of death in traumatic patients admitted to ICUs remains high. These models use variables derived from the deviation of physiological parameters and/or the severity of anatomical lesions with respect to the affected body areas. Our objective is to create different predictive models of the mortality of critically traumatic patients using machine learning techniques. METHODS: We used 9625 records from the RETRAUCI database (National Trauma Registry of 52 Spanish ICUs in the period of 2015-2019). Hospital mortality was 12.6%. Data on demographic variables, affected anatomical areas and physiological repercussions were used. The Weka Platform was used, along with a ten-fold cross-validation for the construction of nine supervised algorithms: logistic regression binary (LR), neural network (NN), sequential minimal optimization (SMO), classification rules (JRip), classification trees (CT), Bayesian networks (BN), adaptive boosting (ADABOOST), bootstrap aggregating (BAGGING) and random forest (RFOREST). The performance of the models was evaluated by accuracy, specificity, precision, recall, F-measure, and AUC. RESULTS: In all algorithms, the most important factors are those associated with traumatic brain injury (TBI) and organic failures. The LR finds thorax and limb injuries as independent protective factors of mortality. The CT generates 24 decision rules and uses those related to TBI as the first variables (range 2.0-81.6%). The JRip detects the eight rules with the highest risk of mortality (65.0-94.1%). The NN model uses a hidden layer of ten nodes, which requires 200 weights for its interpretation. The BN find the relationships between the different factors that identify different patient profiles. Models with the ensemble methodology (ADABOOST, BAGGING and RandomForest) do not have greater performance. All models obtain high values ​​in accuracy, specificity, and AUC, but obtain lower values ​​in recall. The greatest precision is achieved by the SMO model, and the BN obtains the best recall, F-measure, and AUC. CONCLUSION: Machine learning techniques are useful for creating mortality classification models in critically traumatic patients. With clinical interpretation, the algorithms establish different patient profiles according to the relationship between the variables used, determine groups of patients with different evolutions, and alert clinicians to the presence of rules that indicate the greatest severity.

Associated risk factors and outcomes of acute kidney injury in severe trauma: Results from the Spanish trauma ICU registry (RETRAUCI).

INTRODUCTION: Acute kidney injury (AKI) constitutes a common complication after severe trauma. Our objective was to analyse the associated risk factors and outcomes of AKI in a large, multicentre sample of trauma ICU patients. MATERIALS AND METHODS: Observational, prospective and multicentre nationwide registry (RETRAUCI). We included all patients admitted to the participating ICUs from November 2013 to May 2017. We analysed the impact of AKI evaluated by the Risk, Injury, Failure, Loss of kidney function and End-stage kidney disease (RIFLE) definition. Comparison of groups was performed using Wilcoxon test, Chi-Square Test or Fisher's exact test as appropriate. A multiple logistic regression analysis was performed to analyse associated factors to the development of AKI. Logistic regression was used to calculate AKI-related mortality. A P value<0.05 was considered significant. RESULTS: During the study period, 5882 trauma patients were admitted. Complete data were available for 5740 patients. Among them, 871 had AKI (15.17%), distributed by RIFLE R 458 (7.98%), RIFLE I 234 (4.08%) and RIFLE F 179 (3.12%). Associated risk factors were: age (OR 3.05), haemodynamic instability (OR 2.90 to OR 8.34 depending on the severity of hypotension), coagulopathy (OR 1.82), rhabdomyolysis (OR 4.67) and AIS abdomen (OR 1.54). AKI was associated with mortality (crude OR 1.93 (1.59-2.36)), even after adjusting by potential confounders (adjusted OR 1.40 (1.13-1.73)). CONCLUSION: In our large sample of trauma ICU patients we found an incidence of AKI of 15%, which was associated with an increased mortality.