Implementation of a national AI technology program on cardiovascular outcomes and the health system

Coronary artery disease (CAD) is a major cause of ill health and death worldwide. Coronary computed tomographic angiography (CCTA) is the first-line investigation to detect CAD in symptomatic patients. This diagnostic approach risks greater second-line heart tests and treatments at a cost to the patient and health system. The National Health Service funded use of an artificial intelligence (AI) diagnostic tool, computed tomography (CT)-derived fractional flow reserve (FFR-CT), in patients with chest pain to improve physician decision-making and reduce downstream tests. This observational cohort study assessed the impact of FFR-CT on cardiovascular outcomes by including all patients investigated with CCTA during the national AI implementation program at 27 hospitals (CCTA n = 90,553 and FFR-CT n = 7,863). FFR-CT was safe, with no difference in all-cause (n = 1,134 (3.2%) versus 1,612 (2.9%), adjusted-hazard ratio (aHR) 1.00 (0.93-1.08), P = 0.97) or cardiovascular mortality (n = 465 (1.3%) versus 617 (1.1%), aHR 0.96 (0.85-1.08), P = 0.48), while reducing invasive coronary angiograms (n = 5,720 (16%) versus 8,183 (14.9%), aHR 0.93 (0.90-0.97), P < 0.001) and noninvasive cardiac tests (189/1,000 patients versus 167/1,000), P < 0.001). Implementation of an AI-diagnostic tool as part of a health intervention program was safe and beneficial to the patient pathway and health system with fewer cardiac tests at 2 years.

Authors

Institutions

Publication Details

Journal
Nature Medicine
Published
2025-04-04
DOI
https://doi.org/10.1038/s41591-025-03620-y
Citations
25
Primary Topic
Cardiac Imaging and Diagnostics
Type
article
Field-Weighted Citation Impact
34.00

Funders

Controls
|||
ALL TIME
JAN
FEB
MAR
APR
MAY
JUN
JUL
AUG
SEP
OCT
article

Implementation of a national AI technology program on cardiovascular outcomes and the health system

John Pierre Greenwood, Justin Carter, Imran Sunderji, Sarah Mullen et al.
25 citations
Nature Medicine
Cardiac Imaging and Diagnostics
34.00
article

Implementation of a national AI technology program on cardiovascular outcomes and the health system

John Pierre Greenwood, Justin Carter, Imran Sunderji, Sarah Mullen, Edward Nicol, Sumeet Sharma, Anna Beattie, Gerry P McCann, Oliver John Watson, Andrew M. Beale, Jennifer Crooks, Campbell Rogers, Sri Iyenger, Nicholas G. Bellenger, Russell Bull, Benjamin Hudson, Jonathan Richard Weir-McCall, Hatef Mansoubi, Peter Haworth, Peter O’Kane, Matthias Schmitt, Victoria Parish, Vinod Venugopal, Manish Motwani, Matthew Shaw, Liam Mullen, Jonathan Rodrigues, Laurence P. Tidbury, Aparna Deshpande, Sawan Waidyanatha, Girvan Burnside, Chee Gee Liew, Timothy A. Fairbairn, Anoop Chauhan, Simon Duckett, Gregory Y. H. Lip, Joban Sehmi, Anna Reid, Ian Kemp
article en
25 citations

Abstract

Coronary artery disease (CAD) is a major cause of ill health and death worldwide. Coronary computed tomographic angiography (CCTA) is the first-line investigation to detect CAD in symptomatic patients. This diagnostic approach risks greater second-line heart tests and treatments at a cost to the patient and health system. The National Health Service funded use of an artificial intelligence (AI) diagnostic tool, computed tomography (CT)-derived fractional flow reserve (FFR-CT), in patients with chest pain to improve physician decision-making and reduce downstream tests. This observational cohort study assessed the impact of FFR-CT on cardiovascular outcomes by including all patients investigated with CCTA during the national AI implementation program at 27 hospitals (CCTA n = 90,553 and FFR-CT n = 7,863). FFR-CT was safe, with no difference in all-cause (n = 1,134 (3.2%) versus 1,612 (2.9%), adjusted-hazard ratio (aHR) 1.00 (0.93-1.08), P = 0.97) or cardiovascular mortality (n = 465 (1.3%) versus 617 (1.1%), aHR 0.96 (0.85-1.08), P = 0.48), while reducing invasive coronary angiograms (n = 5,720 (16%) versus 8,183 (14.9%), aHR 0.93 (0.90-0.97), P < 0.001) and noninvasive cardiac tests (189/1,000 patients versus 167/1,000), P < 0.001). Implementation of an AI-diagnostic tool as part of a health intervention program was safe and beneficial to the patient pathway and health system with fewer cardiac tests at 2 years.

Nature MedicineVol. 31(6)
North Tees and Hartlepool NHS Foundation Trust (GB), University of Liverpool (GB), University of Leicester (GB), King's College - North Carolina (US), King's College London (GB), University of Hull (GB), Leeds Teaching Hospitals NHS Trust (GB), University Hospitals of North Midlands NHS Trust (GB), West Hertfordshire Hospitals NHS Trust (GB), Royal Devon and Exeter Hospital (GB), Newcastle upon Tyne Hospitals NHS Foundation Trust (GB), Dudley Group NHS Foundation Trust (GB), University Hospitals of Leicester NHS Trust (GB), Harefield Hospital (GB), Frimley Health NHS Foundation Trust (GB), United Lincolnshire Hospitals NHS Trust (GB), Liverpool Heart and Chest Hospital (GB), Portsmouth Hospitals NHS Trust (GB), Manchester University NHS Foundation Trust (GB), Sheffield Teaching Hospitals NHS Foundation Trust (GB), Papworth Hospital NHS Foundation Trust (GB), Blackpool Teaching Hospitals NHS Foundation Trust (GB), HeartFlow (United States) (US), Great Western Hospitals NHS Foundation Trust (GB), Ashford and St Peter's Hospitals NHS Foundation Trust (GB), University Hospitals Sussex NHS Foundation Trust (GB), nView Medical (United States) (US), Royal United Hospital Bath NHS Trust (GB), Bournemouth University (GB)
Medical Research Council
Good health and well-being
Openalex Percentile: Top 0%
Cardiac Imaging and Diagnostics
34.00
AI Navigator

Ask Laika to Summarize, Analyze, and Connect papers live on the map.

Summarize Papers & Methodologies

Extract key findings, datasets, and comparative methods across publications.

Benchmark Rankings & Visual Analytics

Rank top research institutions, authors, funders, topics, and journals by Field-Weighted Citation Impact (FWCI) and paper volume with instant charts.

Connect Distant Disciplines

Bridge topological clusters on the map to find hidden collaborative intersections.