Artificial intelligence-generated encounter summaries: early insights from ambulatory clinicians at a large academic health system

Objectives Assess time impact and provider perception of AI-generated encounter summaries. Materials and Methods An artificial intelligence (AI) clinical note summarization tool was deployed in ambulatory practices for 22 providers. A 6-month pre-post analysis evaluated changes in EHR time, and survey feedback assessed tool utility. Results Among 17 providers with complete Epic Signal time data, average time spent in clinical review per visit decreased from 3:22 minutes pre-intervention (8345 visits) to 3:04 minutes post-intervention (7924 visits). Higher baseline review time was significantly associated with greater post-implementation reductions, with each additional minute at baseline linked to an 8.6-second decrease in review time (P = .046). The majority of providers reported high tool accuracy and time saving. Discussion AI-generated summaries appear most beneficial for clinicians who initially spend the longest time in chart review, suggesting a targeted efficiency gain. Providers expressed optimism about the tool’s potential to streamline clinical workflows. Conclusion Early experience with AI-generated summaries shows promise in improving provider workflow.