Shah Lab

Automations are fixed prompts applied to defined sets of patients. Across seven operational workflows, these automations convert high-volume, repetitive review into focused, actionable work embedded in the EHR. They help place patients in the right care setting, identify surgical patients who may need medical co-management, strengthen infection surveillance, accelerate referral triage, surface patients who may benefit from inpatient hospice, standardize complex orthopedic referrals, and validate billing accounts. Collectively, they narrow 150 daily charts to about 30 likely Sequoia candidates and roughly 600 charts to 10–40 potential hospice candidates; reduce surgical-site-infection manual review by 40%; process more than 1,000 referrals per day with 17.8% higher throughput; cut referral backlog from five to two business days; and save substantial staff time in orthopedic triage and billing review—demonstrating how validated, workflow-integrated LLM automation can expand capacity, improve consistency, and accelerate patient care while retaining human oversight.

===== User Interface ===== When users open a ChatEHR tab in the EHR, they first see a welcome menu (left) to select which patient data to load into context for the LLM, by selecting data sources and a time range. After launching ChatEHR, they see a chat window (right) where they can submit queries and receive responses, provide feedback (thumbs up/down), or start a new session by returning to the welcome menu. ==== Testimonials ==== The UI pilot concluded in June 2025, with a total of 150 users, 3000 chat sessions, and dozens of testimonials (below) and led to a full enterprise roll out in Sept 2025. In the first 3 months, 1075 users completed the training video and engaged in over 23,000 sessions with over 19B tokens processed. 180 APPs, 424 physicians, 151 residents, and 60 fellows used the UI at least once in the first 3 months of launch. ==== Current usage ==== The panels below show current usage as of July 2026, with 3400+ active users, 150,000+ sessions, and 140B tokens processed. The top left panel shows usage by role and to top right shows usage by service line. The bottom left panel shows usage by time and day of week, and the bottom right panel shows the top 20 tasks performed by the users. —- ===== Our Journey ===== ChatEHR was made possible by embedding the data science team, which includes software engineers, product managers, and data scientists with direct access to academic researchers, in the IT organization. Because of this set up, what began as a sandbox in April 2023 was conceptualized as ChatEHR in November 2023, underwent multiple prototypes by mid-2024, leading to a pilot in early 2025 that scaled to 150 clinical users by August 2025, and a health system-wide roll out by September 2025. ===== News coverage ===== - StatNews - Stanford’s health AI validation tools you should know about - StatNews - Why some hospitals are making their own ChatGPTs for patient records - Stanford News - Clinicians can ‘chat’ with medical records through new AI software, ChatEHR - Stanford News - AI lets clinicians ‘chat’ with medical records - MobiHealthNews - Stanford Medicine's ChatEHR expedites the chart review process - HealthcareITNews - New 'ChatEHR' tool enables clinical conversation at Stanford - Becker's - Stanford pilots ChatEHR - Stanford Medicine Introduces ChatEHR for Efficient Chart Reviews - Stanford’s ChatEHR lets doctors talk to the chart — and it talks back