Objective To validate a triage flowchart to rule out acute coronary syndrome (ACS) in chest pain patients attending the emergency department (ED).

Methods An observational cohort study of consecutive patients. In all cases, a previously derived five-step triage flowchart (age ≤40 years, absence of diabetes, not previously known coronary artery disease, non-oppressive and non-retrosternal pain) was applied. Patients meeting all five discriminators were grouped as ‘five-step triage non-ACS’, the rest as ‘five-step triage ACS’. The same strategy was used with a four-step model (without age ≤40 years). After ED study and 1-month follow-up, patients were definitively classified as ‘true ACS’ or ‘true non-ACS’. Validity indexes and receiver operating characteristics curves were calculated.

Results 4231 patients were included: 918 (21.7%) were ‘true ACS’, 3303 (78.1%) ‘true non-ACS’; 10 (0.2%) were lost to follow-up. The five-step triage flowchart classified 4000 (94.8%) as ‘triage ACS’ and 221 (5.2%) as ‘triage non-ACS’; none of the latter was ‘true ACS’. The four-step model classified 3194 (75.6%) as ‘triage ACS’ and 1027 (24.4%) as ‘triage non-ACS’. A ‘true ACS’ was seen in 26 patients from the latter group. Accordingly, five-step triage flowchart specificity and positive predictive value (PPV) to rule out ACS were 100% (95% CI 100% to 100%). For the four-step model specificity and PPV were 97% (95% CI 96% to 98%).

Conclusion The five-step triage flowchart identifies chest pain patients without an ACS. However, only 5% of these patients meet these five criteria. A simpler model allows greater patient inclusion but a higher risk of misclassification of true ACS.