Since the NEJM publication by Pope et al in 2000 found 2% of patients with myocardial infarctions were discharged home from the Emergency Department, we have been in search of a tool capable of identifying a cohort of patients who are truly safe for discharge1. Many have anointed the high-sensitivity troponin assay (hs-cTnI) to this role. In theory, the added sensitivity would identify the subset of ACS patients who would otherwise remain occult during their ED stay. Such a sensitivity could come with a loss of specificity, a heavy price when the incidence of disease is low. Despite the theoretical advantages of hs-cTnI, evidence supporting its use is inconsistent at best. With the recent publication of a cluster-RCT by Shah et al in the Lancet, we now have evidence demonstrating the clinical consequences of high sensitivity assays on patient care2.
The authors performed a stepped-wedge, cluster-randomized controlled trial in 10 Emergency Departments. Patients presenting with signs and symptoms concerning for ACS underwent an evaluation, during which both a standard troponin and a high-sensitivity assay were drawn. The study was divided into the validation phase, in which the hs-cTnI was drawn with results not divulged to the treating clinicians, and the implementation phase, during which the hs-cTnI was incorporated into clinical decision making. All sites underwent at least 6-months of a validation phase, after which they were randomly assigned to either an immediate implementation phase or a delayed implementation phase following an additional 6-months of the validation.
Over a 3-year period the authors enrolled 48,282 patients. Of these 18,978 (39%) were enrolled during the validation phase and 29,304 (61%) during the implementation phase. 10,360 (21%) of the patients had an hs-cTnI value that was above the 99th percentile. The majority of these patients (83%) also had an elevated contemporary troponin assay, with only 17% reclassified using the Hs-Tn. The authors found no difference in their primary outcome, the rate of subsequent myocardial infarction (type 1 or type 4b), or cardiovascular death within 1 year following the initial presentation to hospital (6% vs 5% in the validation and implementation phases, respectively). Patients who were reclassified using the hs-cTnI were more likely than those who had congruent negative troponin assays to suffer a subsequent myocardial infarction or death. Despite this increased risk of adverse events, clinician awareness of the elevations in hs-cTnI did not improve patient oriented outcomes. While hs-cTnI was a predictor of a sicker subset of the population, it did not identify a group of patients who benefited from a more aggressive treatment strategy.
Despite our hopes, it would appear that the introduction of hs-cTnI added very little to the work-up of patients presenting with symptoms concerning for ACS. In fact, the utilization likely led to a significant degree of over-diagnosis, downstream testing and over-treatment. While the use of hs-cTnI increased the number of patients diagnosed with myocardial injury by 17%, only a third of these patients had a diagnosis of type 1 myocardial infarction. Patients reclassified by the high-sensitivity assay were more likely to undergo coronary angiography in the implementation phase compared with the validation phase, (11% in the implementation phase vs 4% in the validation phase), and more frequently received prescriptions for antiplatelet agents, dual-antiplatelet agents, statins and beta-blockers. All this without improving patient important outcomes. Finally, the full extent of the overdiagnosis is likely understated in this dataset as the authors utilized the hs-cTnI as their diagnostic gold standard when ascertaining which patients experienced a MI during the 1-year follow up period.
These results should not come as a surprise. They are simply the consequence of the utilization of a more sensitive assay. How many patients will be falsely diagnosed with ACS and suffer the downstream consequences of such a diagnostic burden? Despite many years of progression in the understanding of ACS, a host of novel biomarkers, and a brand new high-sensitivity troponin assay, not much has changed since the publication of the infamous Pope trial. We are still missing approximately 1-2% of MIs following a negative ED workup. And it would seem the introduction of the hs-cTnI has not changed that number. This study is far from perfect. The 1-year outcome that was selected as the primary endpoint is less than ideal for evaluating the downstream consequences of a diagnostic strategy from the Emergency Department. There could be small differences in outcomes that were missed due to the study design. Despite these shortcomings, the results from Shah et al should give us pause before we anoint hs-cTnI as the diagnostic savior in the evaluation of chest pain patients.