Vanderbilt University Medical Center (VUMC) delves into pioneering AI initiatives, piloting Nuance’s DAX Copilot and developing in-house AI solutions. The introduction of Nuance’s tool aims to revolutionize clinical documentation by automating note creation and reducing administrative burdens. Ten physicians will assess the system, reflecting VUMC’s commitment to streamlining workflows and improving medical record-keeping. Moreover, VUMC researchers developed aiChat, an internal version of ChatGPT, providing clinicians with a secure alternative. These efforts align with VUMC’s extensive history in AI tools, marking a significant milestone in their exploration of AI’s potential in healthcare.
Vanderbilt University Medical Center (VUMC) has initiated a groundbreaking step by piloting Nuance’s DAX Copilot, an artificial intelligence tool designed to assist physicians in clinical documentation. This move signifies VUMC’s commitment to exploring AI’s potential in revolutionizing clinician workflows and enhancing medical record-keeping efficiency, as stated by the health system’s Chief Medical Information Officer (CMIO).
The implementation involves engaging ten physicians to utilize and evaluate the Nuance DAX Copilot system. This tool aims to automate the creation of clinical notes, alleviating the administrative burden associated with charting in the electronic health record (EHR). Leveraging natural language processing technology, the AI employs generative techniques to produce real-time clinic visit notes with relevant headings and context by analyzing physician-patient interactions. Subsequently, doctors can review and edit these notes before seamlessly integrating them into the Epic EHR system.
Initially, physicians from VUMC’s Division of General Internal Medicine and Public Health, along with those from the Department of Orthopaedic Surgery, will test Nuance’s technology. However, Dr. Dara Mize mentioned that the health system remains open to evaluating additional AI-powered documentation assistants across various departments.
Apart from Nuance’s DAX Copilot, Vanderbilt researchers have developed aiChat, an in-house iteration of OpenAI’s large language model. Deployed within VUMC’s Azure cloud tenant, aiChat offers clinicians an alternative to the public version of ChatGPT, disabled on VUMC’s networks due to privacy considerations. This internal platform ensures stricter data control, with VUMC maintaining oversight of all information submitted to the HIPAA-certified tool, covered under the institution’s Business Associates Agreement with Microsoft.
Travis Osterman, Associate Vice President for Research Informatics and Associate CMIO for VUMC, emphasized the collaborative efforts between Health IT and the Department of Biomedical Informatics in creating a secure platform for the VUMC community to explore large language models safely.
This initiative aligns with VUMC’s extensive experience in leveraging clinical AI tools, including the development of proprietary solutions. A decade ago, biomedical informaticists at VUMC created an automated application predicting elective surgical case volume, enhancing staffing level management proactively. Additionally, the Vanderbilt-Ingram Cancer Center adopted AI to optimize treatment scheduling, reducing patient wait times, aiding nurse scheduling, and enhancing throughput at infusion centers.
Recent AI initiatives at VUMC involve collaborations with the Gordon and Betty Moore Foundation, focusing on establishing a model framework for AI technology oversight in health systems. The institution has explored AI and Smart on FHIR integration in voice assistants, crafting a natural language processing voice assistant for its Epic EHR.
Dr. Mize emphasized the significance of the DAX Copilot initiative, highlighting its role in enhancing healthcare documentation quality and efficiency. He emphasized how generative AI holds promise in streamlining clinician workflows and improving medical record-keeping while reducing documentation time, marking a pivotal moment in VUMC’s pursuit of AI’s potential in healthcare enhancement.
Vanderbilt University Medical Center (VUMC) stands at the forefront of AI integration in healthcare, as evidenced by its proactive adoption of Nuance’s DAX Copilot and the development of aiChat. The piloting of Nuance’s tool by ten physicians signals VUMC’s commitment to enhancing efficiency in clinical documentation and medical record-keeping. Moreover, the creation of aiChat emphasizes VUMC’s dedication to data security and exploration of tailored AI solutions. With a decade-long history of AI innovation, VUMC continues to pave the way for transformative advancements in healthcare practices. These initiatives mark an essential stride in VUMC’s relentless pursuit of leveraging AI’s potential for the betterment of patient care and clinician workflows.
