RESUMEN
BACKGROUND & AIMS: Guidelines recommend use of risk stratification scores for patients presenting with gastrointestinal bleeding (GIB) to identify very-low-risk patients eligible for discharge from emergency departments. Machine learning models may outperform existing scores and can be integrated within the electronic health record (EHR) to provide real-time risk assessment without manual data entry. We present the first EHR-based machine learning model for GIB. METHODS: The training cohort comprised 2546 patients and internal validation of 850 patients presenting with overt GIB (ie, hematemesis, melena, and hematochezia) to emergency departments of 2 hospitals from 2014 to 2019. External validation was performed on 926 patients presenting to a different hospital with the same EHR from 2014 to 2019. The primary outcome was a composite of red blood cell transfusion, hemostatic intervention (ie, endoscopic, interventional radiologic, or surgical), and 30-day all-cause mortality. We used structured data fields in the EHR, available within 4 hours of presentation, and compared the performance of machine learning models with current guideline-recommended risk scores, Glasgow-Blatchford Score, and Oakland Score. Primary analysis was area under the receiver operating characteristic curve. Secondary analysis was specificity at 99% sensitivity to assess the proportion of patients correctly identified as very low risk. RESULTS: The machine learning model outperformed the Glasgow-Blatchford Score (area under the receiver operating characteristic curve, 0.92 vs 0.89; P < .001) and Oakland Score (area under the receiver operating characteristic curve, 0.92 vs 0.89; P < .001). At the very-low-risk threshold of 99% sensitivity, the machine learning model identified more very-low-risk patients: 37.9% vs 18.5% for Glasgow-Blatchford Score and 11.7% for Oakland Score (P < .001 for both comparisons). CONCLUSIONS: An EHR-based machine learning model performs better than currently recommended clinical risk scores and identifies more very-low-risk patients eligible for discharge from the emergency department.
RESUMEN
BACKGROUND: Despite advances in detection and treatments, biliary tract cancers continue to have poor survival outcomes. Currently, there is limited data investigating the significance of socioeconomic status, race/ethnicity, and environmental factors in biliary tract cancer survival. AIM: To investigate how socioeconomic status and race/ethnicity are associated with survival. METHODS: Data from the Surveillance, Epidemiology, and End Results database for biliary and gallbladder adenocarcinomas were extracted from 1975 to 2016. Socioeconomic data included smoking, poverty level, education, adjusted household income, and percentage of foreign-born persons and urban population. Survival was calculated with Cox proportional hazards models for death in the 5-year period following diagnosis. RESULTS: Our study included 15883 gallbladder, 11466 intrahepatic biliary, 12869 extrahepatic biliary and 7268 ampulla of Vater adenocarcinoma cases. When analyzing county-specific demographics, patients from counties with higher incomes were associated with higher survival rates [hazard ratio (HR) = 0.97, P <0.05]. Similarly, counties with a higher percentage of patients with a college level education and counties with a higher urban population had higher 5-year survival rates (HR = 0.96, P = 0.002 and HR = 0.97, P = 0.004, respectively). CONCLUSION: Worse survival outcomes were observed in lower income counties while higher income and education level were associated with higher 5-year overall survival among gallbladder and biliary malignancies.
