Article Impact Level: HIGH Data Quality: STRONG Summary of Clinical Cardiology, 46(5), 477–483. https://doi.org/10.1002/clc.23995 Yumin Kim et al.
Points
- The study compared clinical coronary computed tomography angiography (CCTA) interpretation with AI-QCT (artificial intelligence-quantitative computed tomography) for non-emergent invasive coronary angiography (ICA) referral in stable patients based on ACC/AHA guidelines.
- AI-QCT identified that 9% of patients had no coronary artery disease (CAD) compared to 34% with clinical CCTA interpretation.
- AI-QCT significantly reduced the need for ICA by 87% and 95% when identifying obstructive coronary stenosis at ≥50% and ≥70% thresholds, respectively.
- Patients with non-obstructive stenosis identified through AI-QCT had excellent clinical outcomes, with no cardiovascular death or acute myocardial infarction in 78% of cases with maximum stenosis < 50%.
- Implementing an AI-QCT-guided referral management strategy reduced overall costs by 26% and 34% for patients with <50% or <70% stenosis, respectively, without affecting 1-year MACE outcomes, demonstrating the potential of AI-QCT in optimizing patient care and healthcare cost-efficiency.
Summary
In this study, the diagnostic performance, economic implications, and association with major adverse cardiovascular events (MACE) were compared between clinical coronary computed tomography angiography (CCTA) interpretation and a semi-automated approach employing artificial intelligence and machine learning, referred to as atherosclerosis imaging-quantitative computed tomography (AI-QCT). The research focused on patients referred for non-emergent invasive coronary angiography (ICA) based on American College of Cardiology (ACC)/American Heart Association (AHA) guidelines.
The study analyzed CCTA data from 747 stable patients aged 60 ± 12.2 years, including 49% women. Traditional site interpretations of CCTAs were juxtaposed with those analyzed using cloud-based software (Cleerly, Inc.) that employed AI-QCT for stenosis determination, coronary vascular measurements, and quantification and characterization of atherosclerotic plaque.
The results indicated that the AI-QCT approach revealed a significant disparity compared to traditional clinical interpretation. Specifically, AI-QCT identified that 9% of patients had no coronary artery disease (CAD), whereas clinical CCTA interpretation yielded this result for only 34%. Moreover, the application of AI-QCT for identifying obstructive coronary stenosis at the ≥50% and ≥70% thresholds could have significantly reduced the need for ICA by 87% and 95%, respectively. Clinical outcomes for patients identified with non-obstructive stenosis through AI-QCT were excellent, with no cardiovascular death or acute myocardial infarction in 78% of cases with maximum stenosis < 50%.
Furthermore, implementing an AI-QCT-guided referral management strategy reduced overall costs by 26% and 34% for patients with <50% or <70% stenosis, respectively, without any change in 1-year MACE outcomes. This research highlights the potential of artificial intelligence and machine learning in reducing unnecessary ICA rates and costs for stable patients referred for guideline-indicated non-emergent ICA. It underscores the clinical and economic benefits of AI-QCT in improving the management of patients with suspected coronary artery disease while maintaining patient safety and healthcare cost-efficiency.
Link to the article: https://onlinelibrary.wiley.com/doi/10.1002/clc.23995
References Kim, Y., Choi, A. D., Telluri, A., Lipkin, I., Bradley, A. J., Sidahmed, A., Jonas, R., Andreini, D., Bathina, R., Baggiano, A., Cerci, R., Choi, E., Choi, J., Choi, S., Chung, N., Cole, J., Doh, J., Ha, S., Her, A., … Chang, H. (2023). Atherosclerosis Imaging Quantitative Computed Tomography (Ai‐qct) to guide referral to invasive coronary angiography in the randomized controlled CONSERVE trial. Clinical Cardiology, 46(5), 477–483. https://doi.org/10.1002/clc.23995
