Abstract

Objective: To investigate the combination of clinical data, exercise testing and biomarkers for the evaluation of patients with chest pain without ST-segment deviation or troponin elevation.

Design: Prospective cohort design.

Settting: Two teaching hospitals in Spain.

Patients: 422 patients presenting to the emergency department were studied. Leukocyte count, C-reactive protein (CRP), pregnancy-associated plasma protein A (PAPP-A) and N-terminal pro-brain natriuretic peptide (NT-proBNP) were determined. A validated clinical risk score (number of points according to pain characteristics and risk factors) was used for clinical evaluation and early exercise testing was performed.

Main outcome measures: Adverse events (death, myocardial infarction or revascularisation) during a median 60 weeks follow-up.

Results: By receiver operating characteristic curve analysis, the association between death or myocardial infarction and adverse events was not significant with leukocyte count (p = 0.3, p = 0.3) or CRP (p = 0.5, p = 0.8), was borderline significant with PAPP-A (p = 0.07, p = 0.04) and strongly significant with NT-pro-BNP (p = 0.0001, p = 0.0001). By Cox regression including clinical risk score, exercise testing result and biomarkers, exercise testing was the independent predictor of revascularisation (p = 0.0001), whereas risk score (p = 0.03) and NT-proBNP (p = 0.0004) predicted death or myocardial infarction. The inclusion of NT-proBNP improved the accuracy of the model for death or myocardial infarction (C-statistic 0.84 versus 0.76, p = 0.01). The combination of clinical score and NT-proBNP afforded the stratification in high (17.2%, p = 0.0001), intermediate (5.3%) and low (1.1%) risk categories of death or myocardial infarction.

Conclusions: NT-proBNP provides incremental prognostic information above that given by clinical history and exercise testing in patients with chest pain without ST-segment deviation and negative troponin.