Construction and Validation of a Risk Prediction Model for Acute Asthma Exacerbations Based on Machine Learning

Objective: The purpose of this project is to build a risk prediction model based on the NHANES 2017–2018 database and to identify high-risk indicators for acute asthma exacerbations. METHODS: Methods: A variety of variables, including past medical history, physical activity, lifestyle factors, demographics, and laboratory markers, were taken from the NHANES database for this study. First, high-risk indicators were identified using logistic regression, and then the significant variables were used to build a risk prediction model. Results: A number of clinical variables, including age, gender, glycated hemoglobin, eosinophil count, coronary heart disease, annual income, and high-sensitivity C-reactive protein (HSCRP), were found to be significant high-risk factors. Based on these important factors, the logistic regression model performed well in terms of prediction, with an AUC of 0.67 for the validation set and 0.64 for the training set. Conclusion: This work established a risk prediction model for acute asthma exacerbations that can be used in clinical risk assessment to help identify high-risk individuals early and enable tailored therapies.