Dataset Information

Predictors of Mortality in Traumatic Intracranial Hemorrhage: A National Trauma Data Bank Study.

ABSTRACT: Background/Objective: Traumatic intracranial hemorrhage (tICH) accounts for significant trauma morbidity and mortality. Several studies have developed prognostic models for tICH outcomes, but previous models face limitations, including poor generalizability and limited accuracy. The objective was to develop a prognostic model and determine predictors of mortality using the largest trauma database in the U.S., applying rigorous analytical methodology with true hold-out-set model validation. Methods: We identified 248,536 patients in the National Trauma Data Bank (NTDB) from 2012 to 2016 with a diagnosis code associated with tICH. For each admission, we collected demographic information, systolic blood pressure, blood alcohol level (BAL), Glasgow Coma Score (GCS), Injury Severity Score (ISS), presence of epidural/subdural/subarachnoid/intraparenchymal hemorrhage, comorbidities, complications, trauma center level, and trauma center region. Our final study population was 212,666 patients following exclusion of records with missing data. The dependent variable was patient death. Linear support vector machine (SVM) classification was carried out with recursive feature selection. Model performance was assessed using holdout 10-fold cross-validation. Results: Cross-validation demonstrated a mean accuracy of 0.792 (95% CI 0.783-0.799). Accuracy, precision, recall, and AUC were 0.827, 0.309, 0.750, and 0.791, respectively. In the final model, high ISS, advanced age, subdural hemorrhage, and subarachnoid hemorrhage were associated with increased mortality, while high GCS verbal and motor subscores, current smoker, BAL beyond the legal limit, and level 1 trauma center were associated with decreased mortality. Conclusions: A linear SVM model was developed for tICH, with nine features selected as predictors of mortality. These findings are applicable to multiple hemorrhage subtypes and may benefit the triage of high risk patients upon admission. While many studies have attempted to create models to predict mortality in TBI, we sought to confirm those predictors using modern modeling approaches, machine learning, and true hold-out test sets, using the largest available TBI database in the U.S. We find that while the predictors we identify are consistent with prior reports, overall prediction accuracy is somewhat lower than prior reports when assessed more rigorously.

SUBMITTER: Wu E

PROVIDER: S-EPMC7705094 | biostudies-literature | 2020

REPOSITORIES: biostudies-literature

ACCESS DATA

Publications

Predictors of Mortality in Traumatic Intracranial Hemorrhage: A National Trauma Data Bank Study.

Wu Esther E Marthi Siddharth S Asaad Wael F WF

Frontiers in neurology 20201117

<b>Background/Objective:</b> Traumatic intracranial hemorrhage (tICH) accounts for significant trauma morbidity and mortality. Several studies have developed prognostic models for tICH outcomes, but previous models face limitations, including poor generalizability and limited accuracy. The objective was to develop a prognostic model and determine predictors of mortality using the largest trauma database in the U.S., applying rigorous analytical methodology with true hold-out-set model validation ...[more]

PMID: 33281725

Similar Datasets

Project description:There have been large changes over the past several decades to patient demographics in those presenting with mild traumatic brain injury (mTBI) with intracranial hemorrhage (ICH; complicated mTBI) with the potential to affect the use of neurosurgical interventions. The objective of this study was to characterize long-term trends of neurosurgical interventions in patients with complicated mTBI using 13 years of the National Trauma Data Bank (NTDB). This was a retrospective cohort study of adult (≥18 years) trauma patients included in the NTDB from 2007 to 2019 who had an emergency department Glasgow Coma Scale score 13-15, an intracranial hemorrhage (ICH), and no skull fracture. Neurosurgical intervention time trends were quantified for each ICH type using mixed-effects logistic regression with random slopes and intercepts for hospitals, as well as covariates for time and 14 demographic, injury, and hospital characteristics. In total, 666,842 ICH patients across 1060 hospitals were included. The four most common hemorrhages were isolated subdural hemorrhage (36%), isolated subarachnoid hemorrhage (24%), multiple hemorrhage types (24%), and isolated unspecified hemorrhages (9%). Overall, 49,220 (7%) patients received a neurosurgical intervention. After adjustment, the odds of neurosurgical intervention significantly decreased every 10 years by the following odds ratios (odds ratio [95% confidence interval]): 0.85 [0.78, 0.93] for isolated subdural, 0.63 [0.51, 0.77] for isolated subarachnoid, 0.50 [0.41, 0.62] for isolated unspecified, and 0.79 [0.73, 0.86] for multiple hemorrhages. There were no significant temporal trends in neurosurgical intervention odds for isolated epidural hemorrhages (0.87 [0.68, 1.12]) or isolated contusions/lacerations (1.03 [0.75, 1.41]). In the setting of complicated mTBI, the four most common ICH types were associated with significant declines in the odds of neurosurgical intervention over the past decade. It remains unclear whether changing hemorrhage characteristics or practice patterns drove these trends.

Project description:BackgroundIntracranial hemorrhage (ICH) is a recognized complication of adults treated with extracorporeal membrane oxygenation (ECMO) and is associated with increased morbidity and mortality. However, the predictors of ICH in this patient category are poorly understood. The purpose of this study was to identify predictors of ICH in ECMO-treated adult patients.MethodsWe conducted a retrospective review of adult patients (?18 years) treated with ECMO at the Karolinska University Hospital (Stockholm, Sweden) between September 2005 and June 2016, excluding patients with ICH upon admission or those who were treated with ECMO for less than 12 h. In a comparative analysis, the primary end-points were the difference in baseline characteristics and predictors of hemorrhage occurrence (ICH vs. non-ICH cohorts). The secondary end-point was difference in mortality between groups. Paired testing and uni- and multivariate regression models were applied.ResultsTwo hundred and fifty-three patients were included, of which 54 (21%) experienced an ICH during ECMO treatment. The mortality for patients with ICH was 81% at 1 month and 85% at 6 months, respectively, compared to 28 and 33% in patients who did not develop ICH. When comparing ICH vs. non-ICH cohorts, pre-admission antithrombotic therapy (p?=?0.018), high pre-cannulation Sepsis-related Organ Failure Assessment (SOFA) coagulation score (p?=?0.015), low platelet count (p?<?0.001), and spontaneous extracranial hemorrhage (p?=?0.045) were predictors of ICH. In a multivariate regression model predicting ICH, pre-admission antithrombotic therapy and low platelet count demonstrated independent risk association. When comparing the temporal trajectories for coagulation variables in the days leading up to the detection of an ICH, plasma antithrombin significantly increased per patient over time (p?=?0.014). No other temporal trajectories were found.ConclusionsICH in adult ECMO patients is associated with a high mortality rate and independently associated with pre-admission antithrombotic therapy and low platelet count, thus highlighting important areas of potential treatment strategies to prevent ICH development.

Dataset Information

Predictors of Mortality in Traumatic Intracranial Hemorrhage: A National Trauma Data Bank Study.

Publications

Predictors of Mortality in Traumatic Intracranial Hemorrhage: A National Trauma Data Bank Study.

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure

Tweets