The Journal of emergency medicine · 2025
Integration of Nitrogen Oxides Into a Triage-Based Index for Predicting Adverse Outcomes in ST-Segment Elevation Myocardial Infarction Patients.
СтатьяНазвание и аннотация: Перевод в очереди
Обработка статьи · этапы, ошибки и управление
Загрузка состояния…
Ошибки и ограничения
История · было → стало
Состояние обновляется каждые 5 секунд, пока блок открыт. Процент показывается только там, где известен объём работы. Пауза останавливает продолжение этапа; текущий запрос к источнику может завершиться.
Ting Man-Ju, Hsieh Chien-Chieh, Yang Hsiao-Yu, Jaw Fu-Shan, Chen Pau-Chung
Аннотация
BACKGROUND: Early risk stratification in ST-segment elevation myocardial infarction (STEMI) remains challenging. The Hypoxia-Age-Shock Index (HASI), incorporating SpO₂, age, heart rate, and systolic blood pressure, offers improved prediction over traditional indices but may benefit from machine learning (ML) and environmental data.
OBJECTIVES: To compare HASI with established shock indices for mortality prediction in STEMI triage and assess whether adding machine learning and ambient NOx data improves early risk assessment.
METHODS: We retrospectively analyzed 711 STEMI patients. HASI was compared with the Shock Index (SI) and Age-Adjusted Shock Index (ASI) for predicting in-hospital mortality. ML models (logistic regression, random forest, support vector machine, XGBoost) were developed using HASI variables. Ambient nitrogen oxides (NOx) data were matched to emergency department arrival times. Model performance was evaluated using area under the receiver operating characteristic curve (AUC), sensitivity, specificity, positive predictive value (PPV), negative predictive value (NPV) and area under the precision-recall curve (AUPRC). SHapley Additive exPlanations (SHAP) analysis assessed variable importance.
RESULTS: Among 711 STEMI patients, 41 (5.8%) died during hospitalization and 77 (10.8%) underwent endotracheal intubation. HASI outperformed both SI and ASI (AUC: 0.747 vs. 0.628 and 0.700; p < 0.05). The application of machine learning further improved predictive performance, with the random forest model achieving an AUC of 0.961 and sensitivity of 0.750. Incorporating ambient NOx further enhanced prediction, increasing the AUPRC to 0.907 and the XGBoost sensitivity to 0.833. NOx ranked third in feature importance.
CONCLUSIONS: HASI, combined with ML and ambient NOx exposure, provides a rapid and interpretable tool based on SHAP analysis for transparent feature contribution and early mortality risk assessment in STEMI triage.
OBJECTIVES: To compare HASI with established shock indices for mortality prediction in STEMI triage and assess whether adding machine learning and ambient NOx data improves early risk assessment.
METHODS: We retrospectively analyzed 711 STEMI patients. HASI was compared with the Shock Index (SI) and Age-Adjusted Shock Index (ASI) for predicting in-hospital mortality. ML models (logistic regression, random forest, support vector machine, XGBoost) were developed using HASI variables. Ambient nitrogen oxides (NOx) data were matched to emergency department arrival times. Model performance was evaluated using area under the receiver operating characteristic curve (AUC), sensitivity, specificity, positive predictive value (PPV), negative predictive value (NPV) and area under the precision-recall curve (AUPRC). SHapley Additive exPlanations (SHAP) analysis assessed variable importance.
RESULTS: Among 711 STEMI patients, 41 (5.8%) died during hospitalization and 77 (10.8%) underwent endotracheal intubation. HASI outperformed both SI and ASI (AUC: 0.747 vs. 0.628 and 0.700; p < 0.05). The application of machine learning further improved predictive performance, with the random forest model achieving an AUC of 0.961 and sensitivity of 0.750. Incorporating ambient NOx further enhanced prediction, increasing the AUPRC to 0.907 and the XGBoost sensitivity to 0.833. NOx ranked third in feature importance.
CONCLUSIONS: HASI, combined with ML and ambient NOx exposure, provides a rapid and interpretable tool based on SHAP analysis for transparent feature contribution and early mortality risk assessment in STEMI triage.