Performance of the 0/2‐hour high‐sensitivity cardiac troponin T diagnostic protocol in a multisite United States cohort

The diagnostic performance of the high‐sensitivity troponin T (hs‐cTnT) 0/2‐h algorithm is unclear among U.S. emergency department (ED) patients with acute chest pain.


INTRODUC TI ON
Approximately 6.5 million emergency department (ED) visits in the United States each year are for patients experiencing chest pain. 1,2celerated diagnostic protocols (ADPs) have been developed to evaluate these patients' risk for acute coronary syndrome (ACS) and guide their ED disposition.9][10] Originally developed and validated in Europe and Australia, the hs-cTnT 0/2-h algorithm is an ADP that relies solely on hs-cTnT measures.2][13][14] A Canadian validation study found an NPV of 99.5% and 97.0% for 30-day CDMI and 30-day major adverse cardiovascular events (MACE).However, the hs-cTnT 0/2-hour algorithm developed by Reichlin et al. has not been validated in a U.S. cohort.
The primary objective of this study was to evaluate the diagnostic performance of the hs-cTnT 0/2-h algorithm for 30-day CDMI within the hs-cTnT to Optimize Chest Pain Risk Stratification (STOP-CP) cohort.A secondary objective was to evaluate the diagnostic performance for index-visit CDMI, CDMI at 90 days, and MACE (defined as cardiac death, MI, or coronary revascularization) at index and 30 and 90 days.Finally, we evaluated whether the combination of the hs-cTnT 0/2-h with the history, ECG, age, and risk factor (HEAR) score improves diagnostic performance. 15

Study design and setting
This is a preplanned secondary analysis of the STOP-CP (Gen 5 STAT assay; Clini calTr ials.gov NCT02984436) prospective, multicenter cohort study.STOP-CP enrolled patients with symptoms concerning for ACS at eight U.S. EDs from January 25, 2017, to September 6,   2018.Study sites are detailed in the primary STOP-CP article. 16stitutional review board approval was obtained at all sites.Written informed consent was obtained for enrollment.7][18] The Standards for Reporting of Diagnostic Accuracy Studies (STARD) guidelines helped direct the research and manuscript development processes. 19

Study population
We prospectively enrolled ED patients ≥21 years of age with serial troponins ordered for the evaluation of possible ACS.Exclusion criteria included ST-elevation myocardial infarction, systolic blood pressure <90 mm Hg, life expectancy <90 days, a noncardiac illness requiring admission, inability to provide consent or be contacted for follow-up, non-English speaking, pregnancy, being a prisoner, or previous enrollment in the study.

Conclusions:
The hs-cTnT 0/2-h algorithm ruled out most patients.With NPV of <99% for 30-day CDMI, the hs-cTnT 0/2-h algorithm, many emergency physicians may not consider it safe to use for U.S. ED patients.When combined with a low-risk HEAR score, NPV was >99% for 30-day CDMI at the cost of reduced efficacy.
sex, race and ethnicity, and risk factors (current or prior tobacco use, hypertension, hyperlipidemia, diabetes, family history of coronary artery disease [CAD], obesity, prior cerebrovascular accident, peripheral vascular disease, and end-stage renal disease).ECG findings of acute ischemia were indicated by the treating physician.

0/2-Hour algorithm
The 0/2-h algorithm used was originally described by Reichlin et al. 21 each patient, hs-cTnT measures were used to stratify patients into rule-out, observation, and rule-in zones using established assayspecific cut-points shown in Figure 1. 9,16,22However, the hs-cTnT 0/2-h algorithm's 0-h rule-out cut-point of 5 ng/L (the limit of detection) was modified to 6 ng/L (the limit of quantification), because the U.S. Food and Drug Administration does not allow reporting below the limit of quantification.[25]

Outcomes
The primary outcome was 30-day CDMI, inclusive of index visit events.Secondary outcomes included: (1) index and 90-day CDMI; (2) index, 30-day, and 90-day MACE (cardiac death, MI, and coronary revascularization); (3) the individual MACE components at index, 30 days, and 90 days; and (4) efficacy, defined as the proportion of patients classified into the rule-out zone during the index visit. 5,26dical record review and telephone follow-up through 90 days were  27 If the cause of death could not be determined, it was considered cardiac.MI was determined by the Fourth Universal Definition of MI: rise and fall of troponin (with at least one value >99th percentile URL) with symptoms of ischemia, ECG evidence of ischemia, imaging evidence of new nonviable myocardium, a new regional wall motion abnormality, or evidence of thrombus on angiography. 28Event rates at 30 days are inclusive of index events, and rates at 90 days are inclusive of both 30 day and index events.

Statistical analysis
Counts, percentages, means and standard deviations, or medians and interquartile ranges (IQRs) were used to describe the study population.To evaluate the performance of the hs-cTnT 0/2-h algorithm, sensitivity, specificity, NPV and positive predictive value (PPV), and negative and positive likelihood ratios (−LR and +LR) for index, 30-day, and 90-day outcomes were calculated.For efficacy, sensitivity, specificity, NPV, PPV, −LR, and + LR, exact 95% confidence intervals (95% CIs) were computed.Consistent with prior studies, sensitivity, NPV, and -LR were calculated for the rule-out zone (i.e., rule-out vs. rule-in or observation) and specificity, PPV, and +LR were calculated for the rule-in zone (i.e., rule-in vs. observation or rule-out). 8,15,16,24,29In addition, we evaluated the diagnostic performance of the combination of the hs-cTnT 0/2-h algorithm with a HEAR score.For this combination, patients were classified to the rule-out zone only if they met both the hs-cTnT 0/2-h algorithm rule-out cut-points and had a low-risk HEAR score of 0-3.Patients with a HEAR score of ≥7 were classified to the rule-in zone regardless of hs-cTnT measures.Patients meeting the hs-cTnT 0/2-hour algorithm's rule-out criteria, who had a HEAR score of 4-6, were reclassified to the observation zone.

RE SULTS
This preplanned secondary analysis included 1307 patients.The patient flow diagram is shown in Figure 2. The cohort was 53.6% (700/1307) male and 58.6% (762/1307) white, with a mean ± SD age of 57.5 ± 12.7 years.Patient demographics are presented in Table 1.

Combination of hs-cTnT 0/2 ADP with the HEAR score
Incorporation of the HEAR score into the algorithm, as detailed in Figure 4 and

DISCUSS ION
In this secondary analysis of a multicenter prospective U.S.-based study of ED patients with suspected ACS, the hs-cTnT 0/2-h algorithm did not achieve an NPV of ≥99% to rule out index or 30-or 90-day CDMI or MACE.A previous investigation of physicians' comfort with chest pain risk stratification demonstrated most physicians desire an NPV of ≥99% for 30-day MACE. 30This acceptable miss rate is echoed in 2021 guidelines on reasonable and appropriate care of ED patients with chest pain and is commonly used in other algorithms aimed at evaluating ED patients with chest pain. 29,31Thus, our findings suggest that many physicians may not be comfortable with the diagnostic performance of the hs-cTnT 0/2-h algorithm when used alone.However, the combination of the hs-cTnT 0/2-h algorithm with a HEAR score was able to improve safety and yield an NPV of ≥99% for index and 30-and 90-day CDMI and for index and 30-day MACE at the cost of substantively reduced efficacy.
Our finding of a low NPV for 30-day outcomes differs from prior studies of the hs-cTnT 0/2-h algorithm.Among European, Australian, and Canadian populations, the hs-cTnT 0/2-h algorithm achieved nearly 100% NPV for 30-day MI and composite of CDMI. 11,21I G U R E 4 hs-cTnT 0/2-h ADP + HEAR score outcomes at index and 30 and 90 days.ACS, acute coronary syndrome; HEAR, history, electrocardiogram, age, and risk factor; hs-cTnT, high sensitivity troponin T; MACE, major adverse cardiovascular event; MI, myocardial infarction; NPV, negative predictive value; PPV, positive predictive value.
NPVs are higher than those found in our study.
However, the validation in Canada, by McRae et al. 11 had an NPV for 30-day MACE of 97.0%, which is similar to our results.The difference in our findings to prior studies may be due to differences in the baseline characteristics of our population compared to previous international studies.For example, the rate of hypertension and diabetes in our study was much higher than in Reichlin's validation cohort (68% vs. 52% and 30% vs. 15%, respectively). 21Similarly, the rates of known CAD, prior MI, and peripheral artery disease were higher in our cohort compared to Reichlin's (33% vs. 26%, 22% vs. 20%, and 7% vs. 2%, receptively).Efficacy of the hs-cTnT 0/2-h algorithm in this cohort was high and similar to results found in prior studies. 11,21This was also similar to the efficacies reported for the hs-cTnI 0/2-h algorithm. 12,13In our study, the efficacy of the hs-cTnT 0/2-h protocol was dramatically reduced when the HEAR score was incorporated.This is consistent with prior studies that have shown that the addition of clinical and historical features, such as risk scores, improve safety, but at the cost of efficacy. 6,12Wildi et al. 12 found that efficacy was reduced from 63.6% to 36.2% and 67.6% to 40.7% in two separate cohorts when the hs-cTnT 0/2-h algorithm was combined with additional clinical features.
Overall, our findings suggest the hs-cTnT 0/2-h algorithm used by itself among U.S. ED patients with chest pain does not achieve a sufficiently high NPV to be considered safe by most clinicians.
However, the combination of this pathway with a HEAR score increases NPVs above 99%, though substantively reducing efficacy.
Thus, clinicians and hospitals must weigh enhanced safety versus a large reduction in efficacy when considering whether to add a risk score to the 0/2-h algorithm.Further study is necessary to determine methods to improve efficacy of chest pain risk stratification pathways while maintaining safety.In this study, the hs-cTnT 0/2-h algorithm combined with the HEAR score reclassified about half of the hs-cTnT 0/2-h algorithm rule-out zone patients to the observation zone.Identifying which observation zone patients can be safety discharged and receive outpatient care may be one future strategy to improve efficacy while maintaining safety.

LI M ITATI O N S
Although this study was conducted at eight U.S. EDs, these were mostly academic sites, which limits generalizability to other care settings.Informed consent was required to participate in STOP-CP, resulting in possible selection bias.We describe a safety threshold of NPV ≥99% for 30-day MACE, but acknowledge that some clinicians are willing to accept lower NPV thresholds. 35The 30-day CDMI and MACE rates in STOP-CP are higher than in previous U.S. cohorts, and this increased prevalence may impact NPV. 29,36Event rates among hs-cTnT 0/2-hour algorithm zones were low.This study used only the Roche hs-cTnT assay.Therefore, these conclusions cannot be applied to 0/2-hour hs-cTnI algorithm derivations.Time from chest pain onset is included in chest pain pathways, however, in the primary STOP-CP analysis we found no difference in performance as a function of time of chest pain onset and was not included in this analysis. 9,16,34This study was observational and as such, the hs-cTnT 0/2-h algorithm was not used to guide patient care; thus the clinical impact of the algorithm is unknown.We include index events in test characteristics to provide the most accurate estimation of the effects of the hs-cTnT 0/2-h algorithm application to STOP-CP cohort patients but recognize the study design may cause these index event rates in the rule-out group to be overestimated compared to

CON CLUS IONS
In this multisite, prospective U.S. cohort study, the high-sensitivity cardiac troponin T 0/2-h algorithm did not achieve a negative predictive value of ≥99% for 30-day cardiac death or myocardial infarction or major adverse cardiovascular events, suggesting many emergency physicians may not find it sufficient to rule out acute coronary syndrome among U.S. ED patients.Adding a HEAR score to the high-sensitivity cardiac troponin T 0/2-h algorithm improved safety at 30 and 90 days with an negative predictive value of >99% at the expense of efficacy.Further study of algorithms incorporating clinical features with the high-sensitivity cardiac troponin T 0/2-h algortihm is warranted to achieve a better balance between safety and efficacy.
completed to determine outcomes.Expert reviewers adjudicated any patient who experienced death or a clinical diagnosis of MI or had an elevated contemporary cTn.Expert adjudicators (Mate Huis in't Veld, MD; Michael Massoomi, MD; Jason P. Stopyra, MD, MS; and James McCord, MD) classified deaths as cardiac or noncardiac based on the Action to Control Cardiovascular Risk in Diabetes (ACCORD) trial definition, with the exception of death due to stroke, which was classified as a noncardiac death.
Note: Data are reported as mean (±SD), n (%), or median (IQR).TA B L E 2Abbreviations: CAD, coronary artery disease; CDMI, cardiac death or myocardial infarction; ECG, electrocardiogram; hs-cTnT, high sensitivity troponin T; MACE, major adverse cardiovascular events; MI, myocardial infarction.TA B L E 3 Test characteristics at index and 30 and 90 days and efficacy for hs-cTnT 0/2-h ADP with and without the HEAR score.Abbreviations: ADP, accelerated diagnostic protocol; HEAR, history, electrocardiogram, age, and risk factor score; hs-cTnT, high sensitivity troponin T; MACE, major adverse cardiovascular events; MI, myocardial infarction; NPV, negative predictive value; PPV, positive predictive value.
TA B L E 4