The Need to Improve Derivation and Description of Algorithms to Rule-Out Patients With Possible Myocardial Infarction

[摘要] Decision- making algorithms for use within the emergency department to rule out acute myocardial infarction ( MI), acute coronary syndrome, or major adverse cardiovascular events require a very low miss rate ( for simplicity, only MI is used from hereon). Sensitivities of = 99% or negative predictive values ( NPVs) of = 99.5% are typically targeted by algorithm developers. These algorithms involve a troponin concentration threshold, or a concentration threshold plus a change in concentration threshold ( delta), or risk score threshold and ECG interpretation. Several have been shown to be clinically useful. Nevertheless, there are a myriad of algorithms to choose from, and it is difficult to know which are optimal ( ie, which would safely rule out the most patients). This article contends that we can improve how we derive and describe these algorithms.