Daily Cardiology Research Analysis
A pragmatic cluster randomized trial showed that AI-enabled ECG alerts increased anticoagulant prescribing and atrial fibrillation diagnosis by noncardiologists. A U.S. registry analysis of >2.4 million PCIs using instrumental variable methods confirmed radial access reduces in-hospital mortality and bleeding but slightly increases ischemic stroke. A CARDIA cohort analysis linked cumulative apoB/LDL-P/TRL-P exposure in young adults with later ASCVD and suggested a usual apoB target <75 mg/dL in
Summary
A pragmatic cluster randomized trial showed that AI-enabled ECG alerts increased anticoagulant prescribing and atrial fibrillation diagnosis by noncardiologists. A U.S. registry analysis of >2.4 million PCIs using instrumental variable methods confirmed radial access reduces in-hospital mortality and bleeding but slightly increases ischemic stroke. A CARDIA cohort analysis linked cumulative apoB/LDL-P/TRL-P exposure in young adults with later ASCVD and suggested a usual apoB target <75 mg/dL in early adulthood.
Research Themes
- AI-enabled cardiovascular diagnostics and implementation
- Access-site strategy in PCI and patient safety
- Life-course lipid exposure and ASCVD prevention
Selected Articles
1. Artificial Intelligence-Enabled ECGs for Atrial Fibrillation Identification and Enhanced Oral Anticoagulant Adoption: A Pragmatic Randomized Clinical Trial.
In a cluster randomized trial across two hospitals, AI-ECG alerts to noncardiologists increased new AF diagnoses and non–vitamin K antagonist oral anticoagulant (NOAC) prescriptions within 90 days. There were no differences in echocardiography orders, cardiology referrals, or short-term ischemic/cardiovascular outcomes.
Impact: This pragmatic RCT demonstrates that AI can close care gaps by changing clinician behavior, improving anticoagulation for at-risk AF patients identified outside cardiology services.
Clinical Implications: Health systems can integrate AI-ECG alerts to support AF detection and timely NOAC initiation by noncardiologists, with governance to monitor downstream testing and safety.
Key Findings
- AI-ECG alerts increased NOAC prescriptions (23.3% vs 12.0%; HR 1.85, 95% CI 1.11–3.07).
- AF diagnosis rates increased with alerts (HR 1.40, 95% CI 1.03–1.90).
- No significant differences in echocardiography orders, cardiologist visits, ischemic stroke, cardiovascular death, or all-cause death.
Methodological Strengths
- Pragmatic, cluster randomized design with physician-level randomization across two hospitals.
- Pre-registered trial (NCT05127460) with objective, EHR-captured process outcomes.
Limitations
- Open-label design and limited to two hospitals in Taiwan; generalizability may be constrained.
- Primary benefits were process measures; trial was underpowered for hard clinical outcomes.
Future Directions: Evaluate long-term clinical outcomes, cost-effectiveness, patient-centered outcomes, and scalable governance for AI alert fatigue and appropriate anticoagulation in broader health systems.
2. Cumulative exposure to atherogenic lipoprotein particles in young adults and subsequent incident atherosclerotic cardiovascular disease.
In CARDIA, higher cumulative and usual early-adult exposure to apoB, LDL-P, and TRL-P predicted incident ASCVD after age 40, with adjusted HRs ≈1.30 per SD. Risk rose beyond a usual apoB exposure of ~75 mg/dL/year, supporting earlier life-course lipid targets.
Impact: Quantifies life-course risk from particle-based exposure and proposes a practical apoB threshold for young adults, shifting prevention earlier.
Clinical Implications: Primary prevention should track apoB (and particle metrics) beginning in early adulthood, with consideration of targets <75 mg/dL to minimize cumulative exposure and later ASCVD risk.
Key Findings
- Among 4366 participants, 241 ASCVD events occurred after age 40 over ~19.3 years.
- Each 1 SD higher cumulative apoB/LDL-P/TRL-P exposure associated with adjusted HR ≈1.30 for ASCVD.
- ASCVD hazard increased beyond a usual apoB exposure ≈75 mg/dL/year in ages 18–<40.
Methodological Strengths
- Prospective, population-based cohort with repeated particle measurements and long follow-up.
- Robust modeling of cumulative and usual exposures with adjusted Cox regression.
Limitations
- Observational design limits causal inference; residual confounding possible.
- Generalizability bounded by CARDIA demographics and measurement intervals.
Future Directions: Test apoB-centered early-life interventions and quantify benefits of maintaining apoB <75 mg/dL across diverse populations, integrating lifetime risk modeling and cost-effectiveness.
3. Radial vs femoral access for percutaneous coronary intervention: temporal trends and outcomes in the USA.
In a national registry (6.66 million PCIs), radial access rose from 20.3% to 57.5% (2013–2022). Instrumental variable analyses in 2.42 million eligible cases linked radial access to lower in-hospital mortality, major bleeding, and vascular complications, but a small absolute increase in ischemic stroke risk.
Impact: Confirms contemporary, nationwide benefits of radial PCI with rigorous causal methods while alerting to a small stroke trade-off, informing guideline implementation and operator training.
Clinical Implications: Radial access should remain the default PCI strategy given mortality and bleeding benefits; teams should mitigate stroke risk via technique optimization (e.g., catheter manipulation, embolic vigilance) and patient selection.
Key Findings
- Radial PCI adoption increased from 20.3% (2013) to 57.5% (2022) across all indications and regions.
- Radial access reduced in-hospital mortality (ARD −0.15%), major access-site bleeding (ARD −0.64%), and other major vascular complications (ARD −0.21%).
- Radial access was associated with a small absolute increase in ischemic stroke (ARD +0.05%) with no signal on a falsification endpoint.
Methodological Strengths
- Massive nationwide registry with contemporary practice spanning a decade.
- Instrumental variable analysis leveraging operator preference and falsification endpoint for causal inference robustness.
Limitations
- Retrospective design; instrumental variable assumptions may be imperfect.
- Outcomes limited to in-hospital events; mechanisms behind stroke risk not elucidated.
Future Directions: Investigate mechanisms and mitigation strategies for periprocedural stroke during radial PCI and evaluate long-term outcomes with patient-level linkage.