HCA Healthcare and Google Cloud are collaborating to enhance clinical workflows using generative AI, aiming to streamline documentation and improve efficiency for physicians. In a pilot program, ER physicians utilized Google’s AI technology via an Augmedix app for natural language processing, converting conversations into notes for electronic health records. The approach received positive feedback, with plans to expand across more hospitals. HCA is also exploring AI’s potential for nurse support and seeks to extend the partnership’s benefits with Google’s specialized language model for medical queries.
HCA Healthcare and Google Cloud have embarked on a collaborative endeavor aimed at harnessing generative AI to enhance clinical workflows and alleviate the burdensome nature of documentation tasks.
The joint effort commenced with a pilot initiative a few months ago, targeting approximately 75 emergency room physicians across four HCA Healthcare hospitals. Through this program, these physicians were granted access to Google’s AI technology, facilitating the more efficient capture of essential medical details derived from patient interactions.
The pilot initiative also included the participation of Augmedix, a company specializing in ambient medical documentation tools and technologies. Leveraging the Augmedix app on hands-free devices, clinicians were able to transcribe clinician-patient conversations into notes using natural language processing, coupled with Google Cloud’s generative AI technology. Subsequently, multi-party medical speech-to-text processing converted the transcribed data into notes for clinicians to review and finalize before integrating them into the electronic health record.
The collaborative team, consisting of HCA’s Care Transformation and Innovation division, Google, and Augmedix, is in the process of refining this approach. Feedback has indicated “high overall satisfaction” among physicians, prompting HCA Healthcare to expand the program to additional hospitals throughout the current year.
HCA is also actively exploring ways in which generative AI can benefit its nursing staff, particularly concerning patient handoffs. By utilizing a Google large language model (LLM), the Care Transformation and Innovation team is developing an automated handoff tool. This tool has the potential to significantly reduce time expenditure while ensuring patient safety and care quality by incorporating human oversight for consistency.
In terms of the AI-generated handoff tool, Google and HCA have strategically designed prompts to guide the LLM toward prioritizing crucial clinical details. These encompass changes in medication, laboratory results, fluctuations in vital signs, patient concerns, and overall treatment responses. The tool’s outputs are meticulously crafted to be easily comprehensible and actionable. As part of its continued refinement, HCA is conducting beta testing of the tool at UCF Lake Nona Hospital.
Looking ahead, HCA Healthcare envisions exploring novel applications for Google’s medically refined Med-PaLM 2 LLM to further assist physicians and nurses. Dr. Michael J. Schlosser, Senior Vice President for Care Transformation and Innovation at HCA Healthcare, highlights the potential utility of a medical-specific LLM for intricate medical queries grounded in scientific and medical knowledge within complex, unstructured medical texts.
The collaboration between HCA Healthcare and Google Cloud originated from their partnership announcement in 2021, which encompasses the multi-year implementation of Meditech Expanse initiated in 2022.
This week, a study by Mass General Brigham unveiled that the ChatGPT LLM demonstrated approximately 72% accuracy in clinical decision-making, maintaining consistent performance across primary and emergency care as well as diverse medical specialties. However, the model encountered challenges in differential diagnoses, as outlined by MGB researchers.
Dr. Schlosser emphasizes the transformative potential of generative AI and new technologies in reshaping interactions within healthcare teams, refining workflows, and furnishing timely, pertinent information to medical professionals. Google Cloud CEO Thomas Kurian echoes this sentiment, underlining how the incorporation of generative AI into solutions can tangibly enhance the daily experiences of doctors and nurses, thereby allowing them to concentrate on patient care.
