A new study led by researchers from Google and Beth Israel Deaconess Medical Center (BIDMC), published in The Lancet, explores whether AI can safely assess patient needs before they arrive at a doctor's office. This research marks a significant milestone as Google’s first-ever publication in the main journal of The Lancet.
Clinical Implementation and Results
In a real-world study conducted at BIDMC’s ambulatory primary care clinic, 98 patients consulted AMIE—Google's research diagnostic AI chatbot—prior to their urgent care visits. Supervising physicians monitored these interactions in real time; notably, not a single conversation required interruption based on predefined safety criteria.
The study's findings highlight the potential utility of AI in clinical settings:
- AMIE’s differential diagnoses matched the final clinical diagnoses in 90% of cases.
- 75% of clinicians reported that AI-generated summaries assisted their preparation for patient visits.
- In more than half of the cases, the AI influenced the physician's approach to care.
Future Outlook
While the results suggest that AI could enhance the patient-physician relationship and alleviate strain on healthcare professionals, the researchers emphasize the need for larger clinical trials to assess these tools at scale. The study represents progress in defining the boundaries of AI applications within science and global health.
