AI and ML Implementation for Patient Message Identification

This case study details the collaboration between Baystate Health and InterSystems to implement an AI/ML-based methodology for identifying urgent patient messages within the MyBaystate patient portal. The initiative aims to enhance patient care, reduce risks, and improve staff productivity by enabling healthcare professionals to quickly identify time-sensitive communications. The project involved developing a custom large language model (LLM) that accurately reflects the local language characteristics of Baystate patients. A multidisciplinary team from both organizations worked together to refine and train the model, which was validated through a pilot program. The results indicate significant improvements in care quality and operational efficiency, as urgent messages are now easily identifiable. The study outlines future plans to further leverage AI to streamline communication processes and enhance overall healthcare delivery.