Predictors of short-term outcome in acute chest pain without ST-segment elevation

BACKGROUND: Management of acute chest pain in the emergency room constitutes a challenge. METHODS: Seven hundred and one consecutive patients were evaluated by clinical history (chest pain score and risk factors), ECG, troponin I and early (<24 h) exercise testing in low risk patients (n=165). A composite end-point (recurrent unstable angina, acute myocardial infarction or cardiac death) was recorded during hospital stay or in ambulatory care settings for patients discharged after early exercise testing. RESULTS: The end-point occurred in 122 patients (17%). Multivariate analysis identified the following predictors: chest pain score > or =11 points (OR=1.8, 2-2.8, 95% CI, P=0.007), age > or =68 (OR 1.6, 1.1-2.4 CI 95%, P=0.03), insulin-dependent diabetes mellitus (OR 1.9, 1.1-3.4 CI 95%, P=0.02), a history of coronary surgery (OR 3.3, 1.5-7.2 CI 95%, P=0.003), ST-segment depression (OR 1.9, 1.2-3.0 CI 95%, P=0.009) and troponin I elevation (OR 1.6, 1.1-2.5, CI 95%, P=0.05). ST-segment depression produced a high end-point increase (31 vs. 13%, P=0.0001). Troponin I elevation increased the risk in the subgroup without ST-segment depression (20 vs. 11%, P=0.006) but did not further modify the risk in the subgroup with ST depression (31 vs. 28%, ns). Nevertheless, the negative ECG and troponin I subgroup showed a non-negligible end-point rate (16% when pain score > or =11 or 7% when pain score <11, P=0.004). Finally, no patient with a negative exercise test presented events compared to 7% of those with a non-negative test (RR=2.5, 2.1-3.1 95% CI, P=0.01). CONCLUSIONS: Emergency room evaluation of chest pain should not focus on a single parameter; on the contrary, the clinical history, ECG, troponin and early exercise testing must be globally analysed.