Logo image
Back
Patient-Centered Design Recommendations for AI-Generated Discharge Summaries
Conference proceeding

Patient-Centered Design Recommendations for AI-Generated Discharge Summaries

Pantea Habibi, Haleh Vatani, Paul Landes, Barbara Di Eugenio, Richard Cameron, Andrew D. Boyd, Pamela Martyn-Nemeth, Carolyn Dickens, Karen Dunn Lopez and Debaleena Chattopadhyay
Proceedings (IEEE International Conference on Healthcare Informatics. Online), pp.680-689
06/01/2026
DOI: 10.1109/ICHI69079.2026.00088

View Online

Abstract

Hospitals are rapidly adopting generative AI to draft discharge documentation, increasing the importance of how discharge information is organized, consumed, and shared-especially for people living with chronic illnesses whose capacity to engage in care and information needs can change over time. We conducted an inpatient interview study with 20 adults hospitalized with chronic heart failure that included task-based design probe activities with a digital discharge summary. Current engagement with discharge summaries was limited: 13 participants recalled receiving paper summaries, but only two had viewed the digital version after discharge, and only four reported reviewing summaries at all. During the design probe tasks, participants preferred listening over reading summaries and wanted easier ways to share discharge information with caregivers. We translate these findings into patient-centered design recommendations for AI-generated discharge summaries in portal settings, emphasizing reduced navigational burden, multimodal access, and ease of sharing.
 Occupational Health Artificial intelligence Design methodology Discharge summary Discharges (electric) Hospitals Medical services Modeling Patient-centered design Portals Printing Probes Qualitative study User study

Details

Metrics

1 Record Views
Logo image