Use of conventional cardiac troponin assay for diagnosis of non-ST-elevation myocardial infarction: ‘The Ottawa Troponin Pathway’

Abstract

Background Serial conventional cardiac troponin (cTn) measurements 6-9 hours apart are recommended for non-ST-elevation MI (NSTEMI) diagnosis. We sought to develop a pathway with 3-hour changes for major adverse cardiac event (MACE) identification and assess the added value of the HEART [History, Electrocardiogram (ECG), Age, Risk factors, Troponin] score to the pathway. Methods We prospectively enrolled adults with NSTEMI symptoms at two-large emergency departments (EDs) over 32-months. Patients with STEMI, unstable angina and one cTn were excluded. We collected baseline characteristics, Siemens Vista conventional cTnI at 0, 3 or 6-hours after ED presentation; HEART score predictors; disposition and ED length of stay (LOS). Adjudicated primary outcome was 15-day MACE (acute MI, revascularization, or death due to cardiac ischemia/unknown cause). We analyzed multiples of 99th percentile cutoff cTnI values (45, 100 and 250ng/L). Results 1,683 patients (mean age 64.7 years; 55.3% female; median LOS 7-hours; 88 patients with 15-day MACE) were included. 1,346 (80.0%) patients with both cTnI�45 ng/L; and 155 (9.2%) of the 213 patients with one value�100ng/L but both<250ng/L or �20% change did not suffer MACE. Among 124 patients (7.4%) with one of the two values>45ng/L but<100ng/L based on 3 or 6-hour cTnI, one patient with absolute change<10ng/L and 6 of the 19 patients with�20ng/L were diagnosed with NSTEMI (patients with Δ10-19ng/L between first and second cTnI had third one at 6-hours). Based on the results, we developed the Ottawa Troponin Pathway (OTP) with a 98.9% sensitivity (95% CI 93.8-100%) and 94.6%

A Methodological Appraisal of the HEART Score and Its Variants

“…This has not yet been the subject of meta-analysis; however, we describe the reports from our search that detail sensitivity in Table 5. 10,12,13,18,32,50,[57][58][59][60][61] Most are underpowered, and the 2 with the strongest lower confidence interval bounds report these values as 96%, ie, compatible with missing 4% of major adverse cardiac events and outside of the 1% to 2% desirable threshold discussed previously. We note limitations to these 2 studies as follows.…”

“…One HEART study used a somewhat more ED-centric endpoint: major adverse cardiac events at 15 days. 18 Neither HEART nor major adverse cardiac events account for disparities in follow-up. In some settings, it is regularly possible to arrange cardiology or primary care clinic visits within 1 to 2 days-as well as prompt outpatient objective cardiac testing-whereas for other settings or patients a prompt, effective reevaluation is at best uncertain.…”

Evaluating the performance of the HEART score in a Tanzanian emergency department

“…We excluded patients lost to follow-up since we were unable to determine MACE outcomes for these patients, which could possibly bias results; however, only 1% of participants were lost to follow-up. This with little change in predictive ability, 14,45 and others have suggested shorter follow-up periods may be more relevant to ED clinicians. 41 Moreover, a longer follow-up period would have identified even more MACE events, which likely would not have improved the performance of the HEART score in our setting.…”

“…Our follow‐up data were obtained at 30 days, while the original HEART score study defined MACE outcomes at 6 weeks 10 . However, this is unlikely to have affected outcomes data as other validation studies used time points of 30 days or less with little change in predictive ability, 14,45 and others have suggested shorter follow‐up periods may be more relevant to ED clinicians 41 . Moreover, a longer follow‐up period would have identified even more MACE events, which likely would not have improved the performance of the HEART score in our setting.…”