New LLM Developed to Manage Adverse Drug Reactions
A new large language model-based clinical decision support system named HELIOT has been developed to help clinicians manage adverse drug reactions. The system integrates literature-guided insights to inform real-time decision-making in clinical settings. This demonstrates a growing trend of using specialized AI tools to augment clinical judgment for critical patient safety issues.
- The HELIOT system aims to reduce alert fatigue by learning from patient-specific medication tolerances documented in unstructured clinical notes, which could potentially decrease interruptive alerts by over 50% compared to traditional clinical decision support systems. - Integrating new AI tools into Epic-centric environments leverages interoperability standards like HL7 FHIR. Epic supports external AI system connections primarily through FHIR APIs, allowing for secure, real-time data exchange without direct database access. Technologies like CDS Hooks enable AI-driven alerts and suggestions to be embedded directly into the Epic workflow at points like ordering and signing. - A significant pain point for ICU nurses is the design of EHRs, with issues like redundant data entry, poor workflow navigation, and excessive clicking contributing to documentation burden and burnout. A survey of over 9,000 nurses revealed that more than two-thirds feel digital documentation burden and poor EHR usability contribute to job dissatisfaction. - Transitioning from an ICU role to nursing informatics can be leveraged by highlighting experience as a super-user, participating in EHR implementation teams, and developing skills in data analysis tools like SQL and Tableau. This clinical experience is crucial for translating clinician needs to technology partners. - To become a board-certified informatics nurse (NI-BC) through the ANCC, a nurse generally needs an active RN license, a BSN, two years of full-time practice, 30 hours of continuing education in informatics, and have completed at least 2,000 hours of practice in informatics nursing within the last three years. - Federal regulations from the ONC and CMS, under the 21st Century Cures Act, mandate the use of FHIR-based APIs to improve patient data access and prevent information blocking. This regulatory push shapes health IT priorities, requiring systems to be more open and interoperable. - Epic is actively integrating AI to enhance clinical efficiency, with features that use generative AI to summarize chart notes, draft clinical documentation from patient conversations (ambient scribes), and draft end-of-shift notes for nurses. Some health systems using these tools report up to a 50% reduction in documentation time. - The HL7 organization is developing standards specifically for AI in healthcare, building on frameworks like FHIR to ensure that data used for training AI models is high-quality and that the outputs are transparent and traceable.
