| >3×cut-off# | >cut-off# | ≤cut-off# | p-value | |
| Patients n | 15 | 9 | 218 | |
| Any chest pain | 9 (60) | 4 (44) | 111 (51) | 0.77 |
| High-risk chest pain | 2 (13) | 0 | 48 (22) | 0.29 |
| First-admission ECG | ||||
| Diagnostic ECG | 4 (27) | 3 (33) | 28 (13) | 0.06 |
| Equivocal ECG | 5 (33) | 3 (33) | 80 (36) | >0.99 |
| Sinus tachycardia | 8 (53) | 7 (78) | 97 (44) | 0.14 |
| Left bundle block | 0 | 0 | 12 (6) | >0.99 |
| Right bundle block | 3 (20) | 0 | 12 (6) | 0.08 |
| P pulmonale | 2 (13) | 1 (11) | 17 (8) | 0.36 |
| Right axis deviation | 2 (13) | 1 (11) | 17 (8) | 0.36 |
| Serial ECG changes | 7 (50) | 4 (50) | 72 (40) | 0.39 |
Data are presented as n (%), unless otherwise stated. #: cut-offs differed between sites and were based on the lowest concentration where the coefficient of variation was <10% as recommended by the 2007 Universal Definition [11] as follows. Royal Infirmary of Edinburgh, Edinburgh, UK (ARCHITECT troponin I assay; Abbott Diagnostics, Abbott Park, IL, USA): 0.15 ng·mL−1 prior to introduction of a new assay (November 17, 2007) and 0.05 ng·mL−1 thereafter; Glasgow Royal Infirmary, Glasgow, UK (ARCHITECT): 0.04 ng·mL−1; Monklands Hospital, Airdrie and Crosshouse Hospital, Kilmarnock, UK (troponin T assay; Roche, Basel, Switzerland): 0.03 ng·mL−1.