BACKGROUND:Existing mortality prediction models have attempted to quantify injury burden following trauma-related admissions with the most notable being the Injury Severity Score (ISS). Although easy to calculate, it requires additional administrative coding. International Classification of Diseases (ICD)-based models such as the Trauma Mortality Prediction Model (TMPM-ICD10) circumvent these limitations, but they use linear modeling, which may not adequately capture the intricate relationships of injuries on mortality.Using ICD-10 coding and machine learning (ML) algorithms, the present study used the National Trauma Data Bank to develop mortality prediction models whose performance was compared with logistic regression, ISS, and TMPM-ICD10. METHODS:The 2015 to 2017 National Trauma Data Bank was used to identify adults following trauma-related admissions. Of 8,021 ICD-10 codes, injuries were categorized into 1,495 unique variables. The primary outcome was in-hospital mortality. eXtreme Gradient Boosting (XGBoost), a ML technique that uses iterations of decision trees, was used to develop mortality models. Model discrimination was compared with logistic regression, ISS, and TMPM-ICD10 using receiver operating characteristic curve and probabilistic accuracy with calibration curves. RESULTS:Of 1,611,063 patients, 54,870 (3.41%) experienced in-hospital mortality. Compared with those who survived, those who died more frequently suffered from penetrating trauma and had a greater number of injuries. The XGBoost model exhibited superior receiver operating characteristic curve (0.863 [95% confidence interval (CI), 0.862-0.864]) compared with logistic regression (0.845 [95% CI, 0.844-0.846]), ISS (0.828 [95% CI, 0.827-0.829]), and TMPM-ICD10 (0.861 [95% CI, 0.860-0.862]) (all p < 0.001). Importantly, the ML model also had significantly improved calibration compared with other methodologies (XGBoost, coefficient of determination (R 2 ) = 0.993; logistic regression, R 2 = 0.981; ISS, R 2 = 0.649; TMPM-ICD10, R 2 = 0.830). CONCLUSION:Machine learning models using XGBoost demonstrated superior performance and calibration compared with logistic regression, ISS, and TMPM-ICD10. Such approaches in quantifying injury severity may improve its utility in mortality prognostication, quality improvement, and trauma research.

No abstract

Background: Although patients with opioid use disorder have been shown to be more susceptible to traumatic injury, the impact of opioid use disorder after trauma-related admission remains poorly characterized. The present nationally representative study evaluated the association of opioid use disorder on clinical outcomes after traumatic injury warranting operative intervention. Methods: The 2010 to 2018 Nationwide Readmissions Database was used to identify adult trauma victims who underwent major operative procedures. Injury severity was quantified using International Classification of Diseases Trauma Mortality Prediction Model. Entropy balancing was used to adjust for intergroup differences. Multivariable regression models were developed to assess the association of opioid use disorder on in-hospital mortality, perioperative complications, resource utilization, and readmissions. Results: Of an estimated 5,089,003 hospitalizations, 54,097 (1.06%) had a diagnosis of opioid use disorder with increasing prevalence during the study period. Compared with others, opioid use disorder had a lower proportion of extremity injuries and falls but greater predicted mortality measured by Trauma Mortality Prediction Model. After adjustment, opioid use disorder was associated with decreased odds of in-hospital mortality (adjusted odds ratio: 0.61; 95% confidence interval, 0.53e0.70) but had greater likelihood of pneumonia, infectious complications, and acute kidney injury. Additionally, opioid use disorder was associated with longer hospitalization duration as well as greater index costs and risk of readmission within 30 days (adjusted odds ratio: 1.36; 95% confidence interval, 1.25e1.49). Conclusion: Opioid use disorder in operative trauma has significantly increased in prevalence and is associated with decreased in-hospital index mortality but greater resource utilization and readmission.

No abstract

No abstract

No abstract

No abstract

Purpose: Safety net hospitals (SNH) have been associated with inferior surgical outcomes and increased resource use. Utilization and outcomes for extracorporeal membrane oxygenation (ECMO), a rescue modality for patients with respiratory or cardiac failure, may vary by safety net status. We hypothesized SNH to be associated with inferior outcomes and costs of ECMO in a national cohort. Materials and Methods: The 2008-2017 National Inpatient Sample was queried for ECMO hospitalizations and safety net hospitals were identified. Multivariable regression was used to perform risk-adjusted comparisons of mortality, complications and resource utilization at safety net and non-safety net hospitals. Results: Of 36,491 ECMO hospitalizations, 28.2% were at SNH. On adjusted comparison SNH was associated with increased odds of mortality (AOR: 1.23), tracheostomy use (AOR: 1.51), intracranial hemorrhage (AOR: 1.39), as well as infectious complications (AOR: 1.21, all P < .05), with NSNH as reference. SNH was also associated with increased hospitalization duration (β=+4.5 days) and hospitalization costs (β=+$32,880, all P < .01). Conclusions: We have found SNH to be associated with inferior survival, increased complications, and higher costs compared to NSNH. These disparate outcomes warrant further studies examining systemic and hospital-level factors that may impact outcomes and resource use of ECMO at SNH.