Algorithms based on high-sensitivity troponin quickly and accurately rule out acute myocardial infarction

Référence

Lee CC, Huang SS, Yeo YH, et al. High-sensitivity-cardiac troponin for accelerated diagnosis of acute myocardial infarction: A systematic review and meta-analysis. Am J Emerg Med 2020;38(7):1402-1407.

Sommaire

These authors searched 2 databases and manually searched the reference lists of included papers to identify studies that examined the diagnostic value of high-sensitivity troponin assays for the diagnosis of AMI. Two authors independently evaluated studies for inclusion and assessed the methodologic quality of the included studies. They ultimately included 56 studies with 67,945 patients. The algorithms used troponin values sequentially at baseline and after 2 hours (in the case of troponin T, after 1 hour). Unfortunately, while the authors identified heterogeneity in the diagnostic criteria and are properly concerned about verification bias, they failed to address the lack of an independent gold standard that does not include one of the tests under consideration. This latter problem, called incorporation bias, will inflate estimates of the diagnostic power of tests. Regardless, for both high-sensitivity troponin I and T, the overall accuracy based on the area under the receiver operating characteristic curve ranged from 89% to 97% (with narrow confidence intervals) with low to moderate heterogeneity. The sensitivity at each time interval ranged from 92% to 95% and the specificity ranged from 68% to 82%. This translates to a positive likelihood ratio range of 2.9 to 6.7 (modestly accurate for ruling in AMI) and a negative likelihood ratio range of 0.05 to 0.11 (very good at ruling out AMI). Even with the inflated estimates, it is likely these tests are very good at ruling out AMI, but may be less accurate at ruling it in.