"Check Symptoms & Get Care": Mount Sinai's AI Triage Solution

Mount Sinai Health System (MSHS), one of New York City's largest academic medical centers, faced a common patient-access challenge: individuals presenting with new symptoms often did not know where to turn, contributing to delayed care, unnecessary clinic and emergency department visits, and inefficient use of provider and facility resources. To improve care navigation, MSHS implemented a scalable, evidence-based, artificial intelligence (AI)-driven, digital self-triage solution. After a competitive market evaluation and request for proposal process beginning in 2021, MSHS selected Clearstep, a platform built in partnership with Dr. Barton Schmitt, the coauthor of the Schmitt-Thompson telephone triage protocols. Clearstep combined a probabilistic natural-language processing layer with a rules-based clinical expert system. The solution, branded "Check Symptoms & Get Care," was deployed across the MSHS public website and integrated within the MyMountSinai mobile app (powered by Epic) in early 2023. From over 60,000 visits, approximately 22,000 patients completed digital triage sessions (37% initiation and approximately 80% completion), with high satisfaction (a system usability scale score of 85.5 and approximately 75% of users rating greater than or equal to 8 out of 10), broad after-hours utilization (71% of use outside business hours), and zero reported safety incidents. Blinded clinician comparison and continuous post-deployment review demonstrated 88%-96% concordance with physician triage. The authors describe the team, hurdles, metrics, and a tiered road map so that organizations of varying technical readiness can adapt this model.