AI-guided refinement of coronary revascularization need in patients suspected of acute coronary syndrome

Aims Overdiagnosis in patients suspected of acute coronary syndrome (ACS) leads to unnecessary coronary angiographies, particularly in cases with non-specifically elevated troponin levels. We established machine learning (ML) models integrating sequentially available prehospital and in-hospital variables to improve early prediction of the need for coronary re while minimizing overdiagnosis. Methods and results Retrospective cohort study analyzing patients with suspected ACS from 2016 to 2020. ML models were trained using data available at different diagnostic time points, including prehospital assessment, arterial blood gas (ABG) analysis, full laboratory results, and sequential troponin measurements. A total of 2,756 patients were included, identified through emergency physician protocols for ACS-related complaints. Patients with incomplete data or prehospital mortality were excluded. Model performance improved with additional diagnostic data. Model 1 (prehospital data only) achieved an AUROC of 0.76 (95% CI 0.72-0.79), while Model 4 (including sequential troponin testing) reached 0.87 (95% CI 0.83-0.91). Adding early hospital diagnostics (Model 2) significantly improved accuracy compared to Model 1 (0.65 vs. 0.78). Sequential troponin testing in Model 4 did not substantially enhance performance compared to single troponin testing in Model 3 (AUROC 0.87, 95% CI 0.83-0.91 vs. 0.86, 95% CI 0.82-0.91). Misclassification analysis revealed that underdiagnosed patients were typically older females with dyspnea and known CAD but no ST-elevations. Overdiagnosed patients had higher BMI, ST-elevations, regional wall motion abnormalities, and impaired LVEF but lacked significant sequential troponin elevation. Conclusion Prehospital assessments combined with early in-hospital diagnostics provide reliable stratification of PCI need, potentially optimizing clinical decision-making and resource utilization.