| Diagnostic performance across STEMI activation pathways against the outcome of confirmed occlusive myocardial infarction | ||||||
| Triage pathway | Total N |
Performance metrics
|
||||
|---|---|---|---|---|---|---|
| Sensitivity (%) | Specificity (%) | PPV (%) | NPV (%) | Accuracy (%) | ||
| Combined Cohort | ||||||
| Patient taken to lab | 295 | 97.27 | 85.41 | 79.85 | 98.14 | 89.83 |
| AI ECG | 295 | 94.55 | 89.19 | 83.87 | 96.49 | 91.19 |
| (Centre A) Autonomous paramedic activation | ||||||
| Paramedic Referrals | 177 | N/A | N/A | 42.37 | N/A | 42.37 |
| Patient taken to lab | 177 | 100.00 | 80.39 | 78.95 | 100.00 | 88.70 |
| AI ECG | 177 | 94.67 | 89.22 | 86.59 | 95.79 | 91.53 |
| (Centre B) Activation by remote cardiology nurse-led ECG and clinical history review | ||||||
| Paramedic Referrals | 118 | N/A | N/A | 29.66 | N/A | 29.66 |
| Patient taken to lab | 118 | 91.43 | 91.57 | 82.05 | 96.20 | 91.53 |
| AI ECG | 118 | 94.29 | 89.16 | 78.57 | 97.37 | 90.68 |