SHM Converge Scientific Abstract ePoster Gallery-Harnessing Machine Learning Models to Predict and Prevent Hospital Readmissions: A Systematic Review

Pdf Summary

This systematic review examined the role of machine learning and AI in predicting and preventing 30-day hospital readmissions. Using PubMed, the authors screened 469 articles published from January 2013 to January 2024 with the search terms “machine learning” and “hospital readmission.” After four rounds of review and applying inclusion criteria focused on General Internal Medicine, U.S. hospitals, 30-day readmissions, and machine learning models, 9 articles were selected.<br /><br />The review found that hospital readmission rates are an important indicator of healthcare quality, patient outcomes, and cost burden. The literature suggests that AI and machine learning can help reduce readmissions by improving risk prediction, supporting clinical decision-making, and enabling more personalized care. These tools can identify patients at risk for early readmission, predict length of stay, and help clinicians tailor discharge planning, follow-up care, and post-discharge interventions.<br /><br />The review also highlights broader applications of AI across healthcare, including improved diagnostic accuracy, better chronic disease management, reduced postoperative complications, and enhanced documentation. Examples cited include comparable or faster performance than clinicians, a 20% reduction in postoperative complications through remote monitoring in anesthesiology, and a 15% reduction in documentation errors using GPT-4, which may improve postoperative care and reduce readmissions. AI has also shown promise in predicting readmissions for conditions such as COPD exacerbations, heart failure, and sepsis, as well as improving COVID-19 diagnostic accuracy.<br /><br />Overall, the review concludes that integrating AI into hospital care offers a promising, sustainable strategy for reducing readmissions through predictive analytics, streamlined care processes, and more individualized patient management.

Asset Subtitle

Mukul Sharda

Meta Tag

Author List Janavi Wagle, Mukul Sharda, Nathaniel Verhagen, Pinky Jha, Ritisha Mathur, Rupesh Prasad, Saathvik Gowda, Sanjay Bhandari, Saujas Sharma, Sharath Kommu, Shaunak Raikar

Category Research

Concept Machine Learning

Concept Artificial Intelligence

Concept Hospital Readmission

Concept Predictive Analytics

Concept Systematic Review

Distinguished Non-Finalist

Presenter Organization Medical College of Wisconsin

Presenting Author Mukul Sharda

Track Technology in Hospital Medicine

Keywords

machine learning

artificial intelligence

hospital readmission

30-day readmission

predictive analytics

risk prediction

clinical decision support

discharge planning

post-discharge care

patient management

Machine Learning

Artificial Intelligence

Hospital Readmission

Predictive Analytics

Systematic Review