Implementing High-Sensitivity Cardiac Troponin T in a US Regional Healthcare System
Yader Sandoval, J. Wells Askew, James S. Newman, Casey M. Clements, Eric D. Grube, Olatunde Ola, Ashok Akula, Marshall Dworak, Scott Wohlrab, Brad S. Karon, Allan S. Jaffe
Abstract
◼ troponin L imited data exist on the transition to high-sensitivity cardiac troponin (hs-cTn) in the United States.This study describes this transition in a multicenter US regional healthcare system.Before implementation of hs-cTnT across the Mayo Clinic enterprise, multidisciplinary discussions involving cardiologists, emergency physicians, hospitalists, and laboratorians were undertaken to facilitate the transition.For the contemporary fourth-generation cardiac troponin T (cTnT) assay, the 99th percentile upper-reference limit was <0.01 ng/mL.For the Elecsys Troponin T Gen 5 STAT (Roche Diagnostics, Switzerland), sex-specific 99th percentile upper-reference limits of 10 ng/L (women) and 15 ng/L (men) were used. 1,2A 0/2-hour protocol was implemented 3 in which serial changes (delta) ≥10 ng/L were identified as diagnostic of acute myocardial injury, a value of ≤3 ng/L as no significant change, and changes of 4 to 9 ng/L as indeterminate.For the indeterminate group, a 6-hour sample was automatically ordered by the laboratory.An empirical change ≥12 ng/L over 6 hours was considered indicative of acute myocardial injury. 3After acute myocardial injury was excluded, further risk stratification was recommended using the history, 12lead ECG, age, and risk factors (HEAR) pathway. 4Extensive educational resources were offered before hs-cTnT implementation, including in-person and online (recorded) lectures, dissemination of forms, pocket cards, slide presentations, online courses, and podcasts.An 856-patient pilot was also done.Furthermore, "champions" were designated in each region for contact if questions arose.The present analysis addresses the implementation of fifth-generation cTnT across the Mayo Clinic Health System hospitals (excluding Mayo Clinic Rochester, which transitioned to a new electronic health record system and fifth-generation cTnT at a separate time).Institutional review board approval was not required because analyses did not involve patient-level or identifiable data.With the use of electronic health records-reporting software (Cogito SlicerDicer, EPIC), hospital-level data/codes were examined to evaluate the transition (12 months before [September 11, 2017--September 11, 2018] and after [September 12, 2018-September 12, 2019]) from the fourth-to fifth-generation cTnT assays across 18 Mayo Clinic Health System hospitals.Local procedural data confirmed the accuracy of SlicerDicer.Acute myocardial infarction (MI) diagnoses, including type 2 MI (International Classification of Diseases, 10th Revision, Clinical Modification code I21.A1), were examined with the use of International Classification of Diseases, 10th Revision codes I21 and I22.Troponin testing was done in 27 174 of 285 519 patients (9.5%) during the preimplementation period and 26 633 of 330 585 (8.1%) during the postimplementation period.Acute MI was coded in 6.2% of cases using the fourth-generation cTnT assay compared with 7.4% using the fifth-generation cTnT assay (absolute difference, 1.2% [95% CI, 0.8-1.6];P<0.