Ruling Out AMI
Studies evaluating the diagnostic performance of hs-cTn assays for the early diagnosis of AMI usually define AMI on the basis of a rising and/or falling pattern of current generation cTn values. However, defining AMI on the basis of the less sensitive current generation assay results in an underestimation of the true prevalence of AMI and an overestimation of negative predictive value of the experimental assay. It also significantly shortens the time it takes to rule in all the AMIs and thus to definitively exclude AMI as it ignores the new AMIs more sensitively detected by the hs-cTn assay. Thus, in the study by Hammarsten et al., the time to exclude all AMIs was 8.5 hours when all of the AMIs detected with the high-sensitivity assay were included, whereas others that do not include these additional events report this can be done in 3 to 4 hours. In our view, Hammarsten is correct.
This does not mean that hs-cTn cannot help in excluding AMI. Body et al. reported that patients who present with undetectable values (less than the LOB of the hs-cTnT assay) were unlikely to have adverse events during follow-up. If that group of patients is added to those who present later than 6 hours, then perhaps a significant proportion of patients with possible acute coronary syndrome (ACS) could have that diagnosis excluded with the initial value. Studies need to continue to evaluate cTn values for at least 6 h to define the frequency of additional AMIs detected in that manner. Using follow-up evaluations of patients with small event rates who are likely to have additional care during the follow-up period are likely to be underpowered. It may be that better initial risk stratification may help with this, as recently reported. Low-risk patients who have good follow-up after an ED visit may be a group that can be released as early as 2 h after presentation.