Doctors, nurses and other clinicians at Stanford Health Care are now able to interact directly with electronic medical documentation through a modern program tool known as Chatehr.
Why does it matter
The tool, built by data scientists from Stanford Medicine, is piloted as a way to facilitate clinicians more smoothly in engaging in its EHR, as reported last week in Stanford Subject of the messageAnd facilitate relieve the administrative burden, allowing them to ask questions about the medical history of their patients.
A tool based on a immense language model (LLM), similar to GPT-4 OPENAI, can also facilitate to automate the summary of patient charts, among others. Chatehr uses information from a person’s medical records to ensure his answer.
The technology has been going on since the last two years, when scientists have first realized the potential of LLM to make EHR more user-friendly-and increasing work efficiency for clinicians.
Chatehr is currently available only to a compact group of clinicians in Stanford hospital: about 30 doctors, nurses, PA and, for example, who are designed to monitor its accuracy and facilitate improve its capabilities.
But the hope, as IT leaders say, consists in the fact that the software, although it will not be able to provide medical advice or decision support, will be able to collect data from EHR on the patient’s longitudinal register and answer questions about his health history, which will save time on clinicians at a care point.
“Chatehr opens a new way for clinicians to interact with electronic medical documentation in a more improved and effective way, regardless of whether he asks for a summary of the entire chart or recovering specific data points relevant to the patient’s care,” said Dr. Michael Pfeffer, chief information and digital officer for Stanford Health Care and School of Medicine. Subject of the message.
[Watch our CIO Spotlight Q&A with Pfeffer from 2023, where he discusses the health system’s AI priorities.]
“This is a unique example of integration of LLM’s capabilities directly with the practice and flow of clinicists’ work,” he added.
Greater trend
Interactions with the EHR system with EHR systems are of course not modern, because the processing of natural language (NLP) and other tools have helped facilitate documentation and charts for doctors and nurses.
But the noise around ChatgPT in 2022 helped to cause universal modern interest in NLP and generative artificial intelligence, and in the last three years it has become obvious that LLM can do great things for healthcare when they are arranged.
Stanford, a longtime leader in AI innovation, is well prepared for innovation. And the chatehr tool looks like a significant value for worked clinicians.
At the same time, as the ponderous implementation of the pilot project shows, as the two latest profiles of AI leaders from Stanford-System of Health Care show, he undertakes a cautious and prudent approach to implementing these tools in clinical conditions.
On the plate
“Making electronic medical documentation more user -friendly means that doctors can spend less time searching for each corner and its needed information,” said Dr. Sneha Jain, assistant to clinical medicine and early Adopter Chatehr, in Stanford Subject of the message.
“AI can extend the practice of doctors and other healthcare professionals, but this is not helpful unless it is embedded in their work, and the information that uses the algorithm is in a medical context,” added Nigam Shah, the main officer of the Stanford science. “Chatehr is secure; he pulls out directly from the relevant medical data; and is built into the electronic system of medical records, which makes it easy and accurate for clinical use.”