NOTIFY-DISCUSS: Patient and clinician perspectives on opportunistic screening for coronary artery calcification using artificial intelligence
6 affiliations
- 1Division of Cardiovascular Medicine, Department of Medicine, Stanford University, Stanford, CA, USA
- 2Division of Cardiovascular Medicine, University of California San Francisco, San Francisco, CA, USA
- 3Evaluation Sciences Unit, Department of Medicine, Stanford University, Palo Alto, CA, USA
- 4Division of Primary Care and Population Health, Department of Medicine, Stanford University, Stanford, CA, USA
- 5Division of Cardiovascular Medicine, Department of Medicine, Stanford University, Stanford, CA, USA; Stanford Prevention Research Center, Stanford University, Stanford, CA, USA; Stanford Cardiovascular Institute, Stanford University, Stanford, CA, USA
- 6Division of Cardiology, Department of Medicine, University of California Los Angeles, Los Angeles, CA, USA; Greater Los Angeles Veterans Affairs Healthcare System, Los Angeles, CA, USA
Abstract
Background: NOTIFY-1 was a quality improvement study that utilized a validated deep learning algorithm to perform opportunistic coronary artery calcification (CAC) screening on previously acquired non-gated chest CT scans, significantly increasing statin prescribing and coronary artery disease testing. NOTIFY-DISCUSS explored patient and clinician perceptions of this real-world clinical application of an artificial intelligence (AI) tool. Methods: Semi-structured interviews and focus groups were conducted with n = 20 NOTIFY-1 participants ( n = 11 patients, n = 9 clinicians). All patients had received their incidental CAC results prior to interviews. Topics included attitudes toward AI use in medicine, experiences with this AI-assisted tool, preferences for communicating incidental findings, and perceived barriers to guideline-directed statin prescribing. Transcripts were coded by a qualitative researcher in NVivo and reviewed by a qualitative expert; analysis of transcripts utilized thematic analysis and consensus approaches. Results: Most participants appreciated the notification tool for providing compelling personalized data to support shared decision-making and behavior change around statin initiation. Participants also noted its potential to improve identification of incidental findings and improve health equity. However, several themes indicated concerns: consent to the use of AI tools; data rights, usage, and security; the time-burden of discussing and contextualizing incidental CAC findings; potential for unintended harm; and need for timely notifications. Conclusions: NOTIFY-DISCUSS offered a unique opportunity to assess perceptions of an implemented AI-driven opportunistic cardiovascular screening tool. Concerns around consent, security, burden and unintended harms will be essential to address when designing and deploying future AI tools, and patient perspectives should be included in ongoing monitoring efforts.
Related
- Potential to improve early detection of lung cancer through expansion of focused and opportunistic screening approaches: a scoping review focusing on potential expansion of screening risk factors and adoption of artificial intelligence models.
- Artificial intelligence in medicine: a clinician-oriented guide and operational framework to evaluate new AI devices
- The Causal Artificial Intelligence Clinician for early haemodynamic management of septic shock in ICU
- Is there a role for artificial intelligence in end-of-life care for persons with dementia? Perspectives of and recommendations by caregivers and hospice staff.
- Artificial Intelligence in Medical Imaging: A Review of Radiodiagnosis in Indian Perspectives.
BibTeX
@article{keamyminor2026notifydiscuss,
title = {{NOTIFY-DISCUSS: Patient and clinician perspectives on opportunistic screening for coronary artery calcification using artificial intelligence}},
author = {Emily Keamy-Minor and Ramzi Dudum and Cassandra L. Bragdon and Summer Ngo and Shiqin Xu and Steven M. Asch and David J. Maron and Cati Brown-Johnson and Alexander T. Sandhu and Fatima Rodriguez},
journal = {American Heart Journal Plus},
year = {2026},
month = oct,
doi = {10.1016/j.ahjo.2026.100989},
url = {https://doi.org/10.1016/j.ahjo.2026.100989},
note = {CC BY-NC-ND},
}
RIS
TY - JOUR
TI - NOTIFY-DISCUSS: Patient and clinician perspectives on opportunistic screening for coronary artery calcification using artificial intelligence
AU - Emily Keamy-Minor
AU - Ramzi Dudum
AU - Cassandra L. Bragdon
AU - Summer Ngo
AU - Shiqin Xu
AU - Steven M. Asch
AU - David J. Maron
AU - Cati Brown-Johnson
AU - Alexander T. Sandhu
AU - Fatima Rodriguez
PY - 2026
DA - 2026/10/22
JO - American Heart Journal Plus
DO - 10.1016/j.ahjo.2026.100989
UR - https://doi.org/10.1016/j.ahjo.2026.100989
AB - Background: NOTIFY-1 was a quality improvement study that utilized a validated deep learning algorithm to perform opportunistic coronary artery calcification (CAC) screening on previously acquired non-gated chest CT scans, significantly increasing statin prescribing and coronary artery disease testing. NOTIFY-DISCUSS explored patient and clinician perceptions of this real-world clinical application of an artificial intelligence (AI) tool. Methods: Semi-structured interviews and focus groups were conducted with n = 20 NOTIFY-1 participants ( n = 11 patients, n = 9 clinicians). All patients had received their incidental CAC results prior to interviews. Topics included attitudes toward AI use in medicine, experiences with this AI-assisted tool, preferences for communicating incidental findings, and perceived barriers to guideline-directed statin prescribing. Transcripts were coded by a qualitative researcher in NVivo and reviewed by a qualitative expert; analysis of transcripts utilized thematic analysis and consensus approaches. Results: Most participants appreciated the notification tool for providing compelling personalized data to support shared decision-making and behavior change around statin initiation. Participants also noted its potential to improve identification of incidental findings and improve health equity. However, several themes indicated concerns: consent to the use of AI tools; data rights, usage, and security; the time-burden of discussing and contextualizing incidental CAC findings; potential for unintended harm; and need for timely notifications. Conclusions: NOTIFY-DISCUSS offered a unique opportunity to assess perceptions of an implemented AI-driven opportunistic cardiovascular screening tool. Concerns around consent, security, burden and unintended harms will be essential to address when designing and deploying future AI tools, and patient perspectives should be included in ongoing monitoring efforts.
SN - 2666-6022
ER -