Résumé
Background: Multiple studies have evaluated the diagnostic and prognostic performance of conventional troponin (cTn) and high-sensitivity troponin (hs-cTn). We performed a collaborative meta-analysis comparing cTn and hs-cTn for diagnosis of acute myocardial infarction (AMI) and assessment of prognosis in patients with chest pain.Methods: MEDLINE/PubMed, Cochrane CENTRAL, and EMBASE were searched for studies assessing both cTn and hs-cTn in patients with chest pain. Study authors were contacted and many provided previously unpublished data.Results: From 17 included studies, there were 8,644 patients. Compared with baseline cTn, baseline hs-cTn had significantly greater sensitivity (0.884 vs 0.749, P < .001) and negative predictive value (NPV; 0.964 vs 0.935, P < .001), whereas specificity (0.816 vs 0.938, P < .001) and positive predictive value (0.558 vs 0.759, P < .001) were significantly reduced. Based on summary receiver operating characteristic curves, test performance for the diagnosis of AMI was not significantly different between baseline cTn and hs-cTn (0.90 [95% CI 0.85-0.95] vs 0.92 [95% CI 0.90-0.94]). In a subanalysis of 6 studies that alternatively defined AMI based on hs-cTn, cTn had lower sensitivity (0.666, P < .001) and NPV (0.906, P < .001). Elevation of baseline hs-cTn, but negative baseline cTn, was associated with increased risk of death or nonfatal myocardial infarction during follow-up (P < .001) compared with both negative.Conclusion: High-sensitivity troponin has significantly greater early sensitivity and NPV for the diagnosis of AMI at the cost of specificity and positive predictive value, which may enable early rule in/out of AMI in patients with chest pain. Baseline hs-cTn elevation in the setting of negative cTn is also associated with increased nonfatal myocardial infarction or death during follow-up.