Multicenter Evaluation of a 0-Hour/1-Hour Algorithm in the Diagnosis of Myocardial Infarction With High-Sensitivity Cardiac Troponin T
Christian Mueller, Evangelos Giannitsis, Michael Christ, Jorge Ordóñez-Llanos, Christopher deFilippi, James McCord, Richard Body, Mauro Panteghini, Tomas Jernberg, Mario Plebani, Franck Verschuren, John French, Robert Christenson, Silvia Weiser, Garnet Bendig, Peter Dilba, Bertil Lindahl, TRAPID-AMI Investigators, Raphael Twerenbold, Hugo A Katus, Steffen Popp, Miquel Santalo-Bel, Richard M Nowak, Daniel Horner, Alberto Dolci, Martina Zaninotto, Alessandro Manara, Sylvie Menassanch-Volker, Jochen Jarausch, Christian Zaugg, Christian Mueller, Evangelos Giannitsis, Michael Christ, Jorge Ordóñez-Llanos, Christopher deFilippi, James McCord, Richard Body, Mauro Panteghini, Tomas Jernberg, Mario Plebani, Franck Verschuren, John French, Robert Christenson, Silvia Weiser, Garnet Bendig, Peter Dilba, Bertil Lindahl, TRAPID-AMI Investigators, Raphael Twerenbold, Hugo A Katus, Steffen Popp, Miquel Santalo-Bel, Richard M Nowak, Daniel Horner, Alberto Dolci, Martina Zaninotto, Alessandro Manara, Sylvie Menassanch-Volker, Jochen Jarausch, Christian Zaugg
Abstract
Study objective: We aim to prospectively validate the diagnostic accuracy of the recently developed 0-h/1-h algorithm, using high-sensitivity cardiac troponin T (hs-cTnT) for the early rule-out and rule-in of acute myocardial infarction.
Methods: We enrolled patients presenting with suspected acute myocardial infarction and recent (<6 hours) onset of symptoms to the emergency department in a global multicenter diagnostic study. Hs-cTnT (Roche Diagnostics) and sensitive cardiac troponin I (Siemens Healthcare) were measured at presentation and after 1 hour, 2 hours, and 4 to 14 hours in a central laboratory. Patient triage according to the predefined hs-cTnT 0-hour/1-hour algorithm (hs-cTnT below 12 ng/L and Δ1 hour below 3 ng/L to rule out; hs-cTnT at least 52 ng/L or Δ1 hour at least 5 ng/L to rule in; remaining patients to the "observational zone") was compared against a centrally adjudicated final diagnosis by 2 independent cardiologists (reference standard). The final diagnosis was based on all available information, including coronary angiography and echocardiography results, follow-up data, and serial measurements of sensitive cardiac troponin I, whereas adjudicators remained blinded to hs-cTnT.
Results: Among 1,282 patients enrolled, acute myocardial infarction was the final diagnosis for 213 (16.6%) patients. Applying the hs-cTnT 0-hour/1-hour algorithm, 813 (63.4%) patients were classified as rule out, 184 (14.4%) were classified as rule in, and 285 (22.2%) were triaged to the observational zone. This resulted in a negative predictive value and sensitivity for acute myocardial infarction of 99.1% (95% confidence interval [CI] 98.2% to 99.7%) and 96.7% (95% CI 93.4% to 98.7%) in the rule-out zone (7 patients with false-negative results), a positive predictive value and specificity for acute myocardial infarction of 77.2% (95% CI 70.4% to 83.0%) and 96.1% (95% CI 94.7% to 97.2%) in the rule-in zone, and a prevalence of acute myocardial infarction of 22.5% in the observational zone.
Conclusion: The hs-cTnT 0-hour/1-hour algorithm performs well for early rule-out and rule-in of acute myocardial infarction.
Copyright © 2016 American College of Emergency Physicians. Published by Elsevier Inc. All rights reserved.
Source: PubMed