Coronary disease & ACSCohort study
Implementation of a national artificial intelligence (AI) technology program on cardiovascular outcomes and the health system
- Observational cohort study
- national implementation program
- 27 hospitals
Authors' conclusion
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.
Study context
- Population
- Patients with chest pain investigated with coronary computed tomographic angiography (CCTA) during the NHS national AI implementation programall patients investigated with CCTA
- Participants
- 90,553 patients
- Exposure
- FFR-CT implementation vs none
- FFR-CTn = 7,863
- Outcome
- CV outcomes
Main finding
As reported in the abstract
All-cause and CV mortality
FFR-CTAdjusted hazard ratio (CI)P value
all-cause mortality
1.00(0.93–1.08)
P = 0.97CV mortality
0.96(0.85–1.08)
P = 0.48- All-cause mortality, n (%), as printed
- 1,134 (3.2%) versus 1,612 (2.9%)
- CV mortality, n (%), as printed
- 465 (1.3%) versus 617 (1.1%)
Observational data: associations, not effects.
Further findings
Invasive coronary angiograms
| Comparison | aHR (CI) | P value | |
|---|---|---|---|
| FFR-CT | 0.93 (0.90–0.97) | P < 0.001 | |
| 0.512Adjusted hazard ratio (log scale) |
- Invasive coronary angiograms — Invasive coronary angiograms, n (%), as printed
- 5,720 (16%) versus 8,183 (14.9%)
Noninvasive cardiac tests
- Noninvasive cardiac tests per 1,000 patients, as printed
- 189/1,000 patients versus 167/1,000
- Reduced noninvasive cardiac tests
- P < 0.001
Original publication
Implementation of a national AI technology program on cardiovascular outcomes and the health system. Nat Med. Published 4 April 2025. doi:10.1038/s41591-025-03620-y
Summarised from the published abstract; every number is checked against it