Background: Troponin point-of-care assays do not achieve a comparable level of analytical sensitivity, compared to cardiac troponin assays measured in a central laboratory, when either the 10% CV or the 99th percentile of a healthy population is used as the cut-off.
Methods: We compared 2 point-of-care assays for troponin I with a central laboratory assay in 86 subjects presenting with acte coronary syndrome (ACS) in a cardiovascular center.
Result: Using the published cut offs the sensitivity, specificity, PPV and NPV of i-STAT were 75%, 100%, 100% and 72%. The sensitivity, specificity, PPV and NPV of PATHFAST were 84.62%, 100%, 100% and 80.95%. Using decreased cut-offs, sensitivity and PPV increased for both assays to 80.77% and 77.27% for i-STAT and 86.54% and 82.93% for PATHFAST.Furthermore, sensitivity and PPV of both POC analyzers improved when tested using female-specific cut-offs of central laboratory assay. Clinical review showed i-STAT using published cut-off missed 1 patient with confirmed MI and that a lower cut-off allowed its detection. There were 41 clinically diagnosed MI cases and additional 12 more patients detected as positive by hsTnI, mostly females (n=10, 83%).
Conclusion: By decreasing the cut offs of POC analyzers, we improved the clinical sensitivity that allows us to identify undisclosed troponin elevations in additional patients that warrant further investigation. The hsTnI assay identified more female with minimal troponin elevations compared to POC assays which raises our concerns that these female subjects may have been overdiagnosed by hsTnI or clinically underdiagnosed.