Detection and Early Warning for Patient's Critical Condition Using Bayes’ Classifier

Unexpected in-hospital cardiac arrest (IHCA) is defined as when vital signs are measured when entering the emergency department for injury examination, but unexpected cardiac arrest occurs during the stay in the emergency department, requiring immediate emergency treatment to save life. Since unexpected in-hospital cardiac arrest (IHCA) is an emergency medical event, this study selected possible risk factors to collect data based on literature review and analysis, and used the last action of non-IHCA to transfer emergency patients to the Intensive Care Unit (ICU). The data is the control group, and the data is logged in. The collected data were sorted into a total of five correlation modules based on affinity relationships: personal variables, National Early Warning Scoring System (NEWS) scores, laboratory test indicators, Charlson comorbidity index scale, and possible causes of IHCA. Finally, we build a prediction model using Bayes’ Theorem and its classifier.