Diabetes Patients’ Readmission Prediction

Mahalakshmi Nathan; Dasantila Sherifi; Veerabahu Muthusamy

Diabetes Patients’ Readmission Prediction

Author(s)	Ms. Mahalakshmi Nathan, Prof. Dr. Dasantila Sherifi, Dr. Veerabahu Muthusamy
Country	United States
Abstract	Hospital readmission within 30 days of discharge is one of the most important quality indicators used to assess healthcare performance and financial efficiency. Such readmissions contribute significantly to avoidable healthcare costs and patient morbidity. Among chronic conditions, diabetes mellitus presents a particularly high risk, as patients with diabetes experience markedly greater 30-day readmission rates compared to the general inpatient population. This increased vulnerability highlights the urgent need for predictive modeling to identify high-risk individuals before discharge and implement targeted post-hospital interventions. In this study, we apply and compare multiple supervised machine-learning algorithms—including Logistic Regression, k-Nearest Neighbors, Decision Trees, Random Forests, Gradient Boosting, and Stochastic Gradient Descent classifiers—on a ten-year electronic health record (EHR) dataset comprising over 100,000 diabetes-related hospital encounters. We perform comprehensive data preprocessing that includes removal of irrelevant identifiers, imputation of missing values, feature normalization, and encoding of categorical variables. Feature engineering focuses on clinically relevant attributes such as prior inpatient visits, length of stay, number of diagnoses, number of medications prescribed, and patient demographics. To address severe class imbalance between readmitted and non-readmitted patients, techniques such as Synthetic Minority Oversampling (SMOTE) and class weighting are employed. Models are evaluated using key performance metrics including area under the receiver operating characteristic curve (AUC), precision, recall, and F1 score. Results demonstrate that ensemble methods, particularly Gradient Boosting, achieve superior discrimination and robustness following hyperparameter optimization. The findings suggest that systematic use of predictive analytics can help healthcare institutions identify high-risk diabetic patients, prioritize discharge planning, and optimize follow-up care. Finally, we present actionable business recommendations for clinical integration and discuss model limitations, ethical considerations, and future research directions to enhance predictive performance and real-world applicability
Keywords	Diabetes; hospital readmission; machine learning; electronic health records; feature engineering; gradient boosting.
Field	Computer > Artificial Intelligence / Simulation / Virtual Reality
Published In	Volume 16, Issue 4, October-December 2025
Published On	2025-10-24

View / Download PDF File

doi

CrossRef DOI is assigned to each research paper published in our journal.

IJSAT DOI prefix is
10.71097/IJSAT

Downloads

Research Paper Format Copyright Permission Form and Undertaking Form Cover Page Vol 16 Isu 3 Cover Page Vol 16 Isu 2 Cover Page Vol 16 Isu 1

All research papers published on this website are licensed under Creative Commons Attribution-ShareAlike 4.0 International License, and all rights belong to their respective authors/researchers.

CC-BY-SA

About IJSAT Fees & Payment Current Issue Publication Archive	Submit Research Paper Track Submission Status Publication Guidelines Publication Ethics Peer Review & Plagiarism	Join as a Reviewer Editors & Reviewers Reviewer Referral Program Get Reviewer Membership Certi.	Website/Journal Policies Usage Policy Content Policies Privacy Policy

Contact Us	Message on WhatsApp	+91-9687-182-185	editor@ijsat.org

International Journal on Science and Technology

A Widely Indexed Open Access Peer Reviewed Multidisciplinary Bi-monthly Scholarly International Journal

Diabetes Patients’ Readmission Prediction

Share this