AI Chatbot Assists Boston Urgent-Care Patients in First-of-Its-Kind Trial

Last year, Beth Israel Deaconess Medical Center in Boston enrolled nearly 100 patients in an eight-month trial that paired a Google-powered chatbot with the hospital’s urgent-care visits. After scheduling an appointment, participants described their symptoms to the bot, which explored possible causes, suggested treatments, and generated a summary for the attending physician before the in-person consultation.

New York Times

The trial’s results, published in *The Lancet*, reported that primary-care physicians found the chatbot helpful in roughly 75 % of cases and that it altered the physicians’ approach in more than half of those encounters. Across hundreds of interactions, the system “hallucinated” – produced fabricated information – only once. Patients expressed trust in the technology, but primarily because it operated under physician supervision.

New York Times

Resident physician Peter Brodeur likened the bot’s performance to that of a high-performing medical student or resident. Internist and medical historian Adam Rodman noted that patient confidence grew only after they realized the chatbot was a hand-off to their doctor rather than an autonomous source of advice.

New York Times

Beth Israel’s primary-care team emphasized that the chatbot served as a decision-support tool, not a replacement for clinical judgment. The hospital highlighted the low incidence of misinformation and the overall positive physician feedback as evidence that the model could be integrated into routine urgent-care workflows, pending further validation.

New York Times

• “They compared it to a high-performing medical student or resident,” — Peter Brodeur, a resident physician at Beth Israel

New York Times

• “None of our patients trusted it until they realized that it was handing-off their doctor,” — Adam Rodman, a Beth Israel internist and medical historian

New York Times