Machine Learning for Predicting Medical Error Risks in Greek Surgery Departments

Patient safety is a critical global health priority, with surgical errors, including in-hospital infections and procedural mishaps, causing over 7 million adverse events and 1 million deaths annually. This study evaluates machine learning (ML) to predict medical error risks in the general surgery department of a Greek Tertiary/University Hospital. Leveraging a 10-year dataset of 19,965 patient records, we applied different ML algorithms, achieving 94.3 % accuracy in detecting errors such as healthcare-associated infections, medication errors, and equipment-related failures. Key predictors included hospitalization duration and initial diagnosis, enabling targeted risk identification. These findings suggest ML can pinpoint risks stemming from staff performance, equipment malfunctions, or clinical management errors, facilitating the development of department-specific safety guidelines. Integration with tools like the WHO Surgical Safety Checklist could enhance proactive error prevention. Such AI-driven models can be seamlessly integrated into future internet-enabled healthcare systems for real-time, proactive patient safety management. However, limitations, including potential data biases from retrospective records and challenges in embedding ML into clinical workflows, may hinder applicability. Ethical concerns, such as patient data privacy, algorithmic fairness, and clinician trust in predictive models, require careful consideration. By combining ML-driven predictive analytics with clinician expertise, healthcare systems can transition from reactive to proactive error mitigation, improving patient outcomes and reducing costs. Future multi-center studies are needed to validate these findings across diverse settings, ensuring generalizability and equitable implementation in resource-constrained environments like Greece, and will benefit from scalable, internet-based platforms for data aggregation and model deployment.