Validation of a prediction score model to distinguish acute coronary syndromes from other conditions causing raised cardiac troponin T levels.

Objective: Serum cardiac troponins can be elevated in acute coronary syndromes (ACS) and other non-ACS conditions. We investigated the usefulness of a prediction score model comprising clinical variables to distinguish patients with ACS from other non-ACS conditions.

Methods: Two independent, non-randomized observational cohorts (groups 1 and 2) were examined, comprising consecutive patients who were admitted to a university teaching hospital and found to have a raised serum troponin T level (>or=0.01 microg/l). The international definition was used to confirm acute myocardial infarction. Multivariate logistic regression identified clinical variables in the first cohort, which were used to construct a score model for distinguishing between ACS and non-ACS, and this score was re-evaluated in the second cohort.

Results: Of the 313 patients in group 1, a score model was formulated using logarithm troponin T, ischaemic chest pain, ST depression and atrial fibrillation or flutter. Using a score of more than or equal to 1.5, sensitivity and specificity for predicting non-ACS were 0.81 and 0.84. The area under the curve was 0.900 (95% confidence interval 0.867-0.934). Sensitivity and specificity for predicting non-ACS among the 341 patients in group 2 using the same model and a score of more than or equal to 1.5 were 0.76 and 0.89, respectively, and the area under the curve was 0.918 (confidence interval 0.887-0.945).

Conclusion: A prediction score model using simple clinical variables has been validated, and this can help clinicians in distinguishing patients with ACS from other non-ACS conditions.