Machine Learning-Based Models for Prediction of Critical Illness at Community, Paramedic, and Hospital Stages

Comparison of the new-Poisoning Mortality Score and the Modified Early Warning Score for predicting in-hospital mortality in patients with acute poisoning

INTRODUCTION

The evaluation of acute poisoning is challenging due to varied toxic substances and clinical presentations. The new-Poisoning Mortality Score was recently developed to assess patients with acute poisoning and showed good performance in predicting in-hospital mortality. The objective of this study is to externally validate the performance of the new-Poisoning Mortality Score and to compare it with the Modified Early Warning Score.

METHODS

This retrospective analysis used data from the 2019-2020 Injury Surveillance Cohort, established by the Korea Center for Disease Control and Prevention, to perform external validation of the new-Poisoning Mortality Score. The statistical performances of the new-Poisoning Mortality and Modified Early Warning Scores were assessed and compared in terms of discrimination and calibration. Discrimination analysis involved metrics such as sensitivity, specificity, accuracy, and the area under the receiver operating characteristic curve. For calibration analysis, the Hosmer-Lemeshow goodness-of-fit test was utilized and calibration curves for each score were generated to elucidate the relationship between observed and predicted mortalities.

RESULTS

This study analysed 16,570 patients with acute poisoning. Significant differences were observed between survivors and those who died in-hospital, including age, sex, and vital signs. The new-Poisoning Mortality Score showed better performance over the Modified Early Warning Score in predicting in-hospital mortality, in terms of the area under the receiver operating characteristic curve (0.947 versus 0.800), sensitivity (0.863 versus 0.667), specificity (0.912 versus 0.817), and accuracy (0.911 versus 0.814). When evaluated through calibration curves, the new-Poisoning Mortality Score showed better concordance between predicted and observed mortalities. In subgroup analyses, the score system consistently showed strong performance, excelling particularly in substances with high mortality indices and remaining superior in all substances as a group.

CONCLUSIONS

Our study has helped to validate the new-Poisoning Mortality Score as an effective tool for predicting in-hospital mortality in patients with acute poisoning in the emergency department. The score system demonstrated superior performance over the Modified Early Warning Score in various metrics. Our findings suggest that the new-Poisoning Mortality Score can contribute to the enhancement of clinical decision-making and patient management.

Reimagining Core Entrustable Professional Activities for Undergraduate Medical Education in the Era of Artificial Intelligence

The proliferation of generative artificial intelligence (AI) and its extensive potential for integration into many aspects of health care signal a transformational shift within the health care environment. In this context, medical education must evolve to ensure that medical trainees are adequately prepared to navigate the rapidly changing health care landscape. Medical education has moved toward a competency-based education paradigm, leading the Association of American Medical Colleges (AAMC) to define a set of Entrustable Professional Activities (EPAs) as its practical operational framework in undergraduate medical education. The AAMC's 13 core EPAs for entering residencies have been implemented with varying levels of success across medical schools. In this paper, we critically assess the existing core EPAs in the context of rapid AI integration in medicine. We identify EPAs that require refinement, redefinition, or comprehensive change to align with the emerging trends in health care. Moreover, this perspective proposes a set of "emerging" EPAs, informed by the changing landscape and capabilities presented by generative AI technologies. We provide a practical evaluation of the EPAs, alongside actionable recommendations on how medical education, viewed through the lens of the AAMC EPAs, can adapt and remain relevant amid rapid technological advancements. By leveraging the transformative potential of AI, we can reshape medical education to align with an AI-integrated future of medicine. This approach will help equip future health care professionals with technological competence and adaptive skills to meet the dynamic and evolving demands in health care.

Predicting short-term outcomes in atrial-fibrillation-related stroke using machine learning

Background

Prognostic prediction and the identification of prognostic factors are critical during the early period of atrial-fibrillation (AF)-related strokes as AF is associated with poor outcomes in stroke patients.

Methods

Two independent datasets, namely, the Korean Atrial Fibrillation Evaluation Registry in Ischemic Stroke Patients (K-ATTENTION) and the Korea University Stroke Registry (KUSR), were used for internal and external validation, respectively. These datasets include common variables such as demographic, laboratory, and imaging findings during early hospitalization. Outcomes were unfavorable functional status with modified Rankin scores of 3 or higher and mortality at 3 months. We developed two machine learning models, namely, a tree-based model and a multi-layer perceptron (MLP), along with a baseline logistic regression model. The area under the receiver operating characteristic curve (AUROC) was used as the outcome metric. The Shapley additive explanation (SHAP) method was used to evaluate the contributions of variables.

Results

Machine learning models outperformed logistic regression in predicting both outcomes. For 3-month unfavorable outcomes, MLP exhibited significantly higher AUROC values of 0.890 and 0.859 in internal and external validation sets, respectively, than those of logistic regression. For 3-month mortality, both machine learning models exhibited significantly higher AUROC values than the logistic regression for internal validation but not for external validation. The most significant predictor for both outcomes was the initial National Institute of Health and Stroke Scale.

Conclusion

The explainable machine learning model can reliably predict short-term outcomes and identify high-risk patients with AF-related strokes.