Utilization of Machine Learning Approaches to Predict Mortality in Pediatric Warzone Casualties.

BACKGROUND: Identification of pediatric trauma patients at the highest risk for death may promote optimization of care. This becomes increasingly important in austere settings with constrained medical capabilities. This study aimed to develop and validate predictive models using supervised machine learning (ML) techniques to identify pediatric warzone trauma patients at the highest risk for mortality. METHODS: Supervised learning approaches using logistic regression (LR), support vector machine (SVM), neural network (NN), and random forest (RF) models were generated from the Department of Defense Trauma Registry, 2008-2016. Models were tested and compared to determine the optimal algorithm for mortality. RESULTS: A total of 2,007 patients (79% male, median age range 7-12 years old, 62.5% sustaining penetrating injury) met the inclusion criteria. Severe injury (Injury Severity Score > 15) was noted in 32.4% of patients, while overall mortality was 7.13%. The RF and SVM models displayed recall values of .9507 and .9150, while LR and NN displayed values of .8912 and .8895, respectively. Random forest (RF) outperformed LR, SVM, and NN on receiver operating curve (ROC) analysis demonstrating an area under the ROC of .9752 versus .9252, .9383, and .8748, respectively. CONCLUSION: Machine learning (ML) techniques may prove useful in identifying those at the highest risk for mortality within pediatric trauma patients from combat zones. Incorporation of advanced computational algorithms should be further explored to optimize and supplement the diagnostic and therapeutic decision-making process.

Combatting ischemia reperfusion injury from resuscitative endovascular balloon occlusion of the aorta using adenosine, lidocaine and magnesium: A pilot study.

BACKGROUND: Resuscitative endovascular balloon occlusion of the aorta (REBOA), a minimally invasive alternative to resuscitative thoracotomy, has been associated with significant ischemia reperfusion injury (IRI). Resuscitation strategies using adenosine, lidocaine, and magnesium (ALM) have been shown to mitigate similar inflammatory responses in hemorrhagic and septic shock models. This study examined the effects of ALM on REBOA-associated IRI using a porcine model. METHODS: Animals underwent a 20% controlled hemorrhage followed by 30 minutes of supraceliac balloon occlusion. They were assigned to one of four groups: control (n = 5), 4-hour ALM infusion starting at occlusion, 2-hour (n = 5) and 4-hour (n = 5) interventional ALM infusions starting at reperfusion. Adenosine, lidocaine, and magnesium cohorts received a posthemorrhage ALM bolus followed by their respective ALM infusion. Primary outcomes for the study assessed physiologic and hemodynamic parameters. RESULTS: Adenosine, lidocaine, and magnesium infusion after reperfusion cohorts demonstrated a significant improvement in lactate, base deficit, and pH in the first hour following systemic reperfusion. At study endpoint, continuous ALM infusion initiated after reperfusion over 4 hours resulted in an overall improved lactate clearance when compared with the 2-hour and control cohorts. No differences in hemodynamic parameters were noted between ALM cohorts and controls. CONCLUSION: Adenosine, lidocaine, and magnesium may prove beneficial in mitigating the inflammatory response seen from REBOA-associated IRI as evidenced by physiologic improvements early during resuscitation. Despite this, further refinement should be sought to optimize treatment strategies.

Pediatric adjusted reverse shock index multiplied by Glasgow Coma Scale as a prospective predictor for mortality in pediatric trauma.

INTRODUCTION: Shock index and its pediatric adjusted derivative (pediatric age-adjusted shock index [SIPA]) have demonstrated utility as prospective predictors of mortality in adult and pediatric trauma populations. Although basic vital signs provide promise as triage tools, factors such as neurologic status on arrival have profound implications for trauma-related outcomes. Recently, the reverse shock index multiplied by Glasgow Coma Scale (GCS) score (rSIG) has been validated in adult trauma as a tool combining early markers of physiology and neurologic function to predict mortality. This study sought to compare the performance characteristics of rSIG against SIPA as a prospective predictor of mortality in pediatric war zone injuries. METHODS: Retrospective review of the Department of Defense Trauma Registry, 2008 to 2016, was performed for all patients younger than 18 years with documented vital signs and GCS on initial arrival to the trauma bay. Optimal age-specific cutoff values were derived for rSIG via the Youden index using receiver operating characteristic analyses. Multivariate logistic regression was performed to validate accuracy in predicting early mortality. RESULTS: A total of 2,007 pediatric patients with a median age range of 7 to 12 years, 79% male, average Injury Severity Score of 11.9, and 62.5% sustaining a penetrating injury were included in the analysis. The overall mortality was 7.1%. A total of 874 (43.5%) and 685 patients (34.1%) had elevated SIPA and pediatric rSIG scores, respectively. After adjusting for demographics, mechanism of injury, initial vital signs, and presenting laboratory values, rSIG (odds ratio, 4.054; p = 0.01) was found to be superior to SIPA (odds ratio, 2.742; p < 0.01) as an independent predictor of early mortality. CONCLUSION: Reverse shock index multiplied by GCS score more accurately identifies pediatric patients at highest risk of death when compared with SIPA alone, following war zone injuries. These findings may help further refine early risk assessments for patient management and resource allocation in constrained settings. Further validation is necessary to determine applicability to the civilian population. LEVEL OF EVIDENCE: Prognostic study, level IV.