Machine Learning-Based Hospital Readmission Prediction and Risk Analysis in the United States Healthcare System

Abstract

Hospital readmission is a significant challenge in the United States healthcare system, leading to increased healthcare costs, hospital overcrowding, and reduced quality of patient care. Early prediction of hospital readmission risk can help healthcare providers identify high-risk patients and take preventive measures before discharge. This study presents an explainable machine learning-based hospital readmission prediction framework using patient demographic and clinical data from the United States healthcare system. The dataset includes important patient information such as age, length of hospital stays, previous admission history, comorbidities, insurance status, disease type, diabetes, heart disease, and follow-up visit status. Several machine learning algorithms, including Logistic Regression, Random Forest, Extreme Gradient Boosting (XGBoost), and Artificial Neural Network, were used to predict hospital readmission risk. The performance of the models was evaluated using accuracy, precision, recall, F1-score, and Receiver Operating Characteristic (ROC) curve. Among all models, XGBoost achieved the highest prediction performance compared to other machine learning models. Feature importance analysis showed that length of hospital stay, comorbidities, insurance status, previous admission history, and follow-up visit were the most significant factors influencing hospital readmission in the United States. In addition to the machine learning prediction model, this study proposes an IoT-based post-discharge patient monitoring system to reduce hospital readmission risk through continuous patient health monitoring and early medical intervention. The proposed system can help healthcare providers monitor patients remotely, identify high-risk patients in real time, and take preventive actions to reduce hospital readmission rates. This research contributes to the development of intelligent healthcare prediction systems and demonstrates how machine learning and IoT technologies can improve patient care, reduce readmission rates, and support data-driven decision-making in the United States healthcare system.