Machine Learning Algorithms for the Early Identification of Sepsis in Geriatric Post-Surgical Patients: A Bibliographic Review Integrating Internal Medicine and General Surgery Perspectives

Abstract

Introduction: Sepsis represents a critical threat to geriatric patients following surgical procedures, with early detection hampered by atypical presentations and complex comorbidities. This bibliographic review examines the current evidence on machine learning (ML) applications for early sepsis identification in geriatric post-surgical populations. Objective: To synthesize and analyze existing literature on ML algorithms that integrate surgical and internal medicine parameters for early sepsis detection in geriatric post-operative patients. Methods: A comprehensive literature search was conducted across PubMed, Scopus, and Web of Science databases from 2015 to 2024. Studies focusing on ML applications for sepsis prediction in geriatric surgical populations were included and critically appraised. Results: The review identified 28 relevant studies demonstrating that ML models, particularly ensemble methods like Random Forest and XGBoost, consistently outperform traditional scoring systems. Integration of surgical parameters (operative duration, blood loss) with internal medicine metrics (comorbidity indices, laboratory trends) significantly enhanced predictive accuracy. Conclusion: ML algorithms show substantial promise for improving early sepsis detection in geriatric surgical patients through interdisciplinary data integration. Future research should focus on clinical implementation, model interpretability, and ethical considerations.

Ähnliche Arbeiten