This diagnostic study evaluates and compares the diagnostic performance of the European Society of Cardiology (ESC) 0/1-hour algorithm for 30-day cardiac death or myocardial infarction among patients with and without known coronary artery disease and determine if the algorithm could achieve the negative predictive value rule-out threshold of 99% or higher.
Key Points
Question
Can the European Society of Cardiology (ESC) 0/1-hour algorithm achieve a negative predictive value of 99% or higher for 30-day cardiac death or myocardial infarction among emergency department patients with chest pain who have known coronary artery disease?
Findings
In this diagnostic study of 1430 adults, the high-sensitivity troponin T (hs-cTnT) ESC 0/1-hour algorithm had a negative predictive value of 96.6% (95% CI, 92.8-98.8) for 30-day cardiac death or myocardial infarction among patients with known coronary artery disease.
Meaning
The ESC 0/1-hour hs-cTnT algorithm was unable to safely exclude 30-day cardiac death or myocardial infarction in patients with known coronary artery disease.
Abstract
Importance
The European Society of Cardiology (ESC) 0/1-hour algorithm is a validated high-sensitivity cardiac troponin (hs-cTn) protocol for emergency department patients with possible acute coronary syndrome. However, limited data exist regarding its performance in patients with known coronary artery disease (CAD; prior myocardial infarction [MI], coronary revascularization, or ≥70% coronary stenosis).
Objective
To evaluate and compare the diagnostic performance of the ESC 0/1-hour algorithm for 30-day cardiac death or MI among patients with and without known CAD and determine if the algorithm could achieve the negative predictive value rule-out threshold of 99% or higher.
Design, Setting, and Participants
This was a preplanned subgroup analysis of the STOP-CP prospective multisite cohort study, which was conducted from January 25, 2017, through September 6, 2018, at 8 emergency departments in the US. Patients 21 years or older with symptoms suggestive of acute coronary syndrome without ST-segment elevation on initial electrocardiogram were included. Analysis took place between February and December 2022.
Interventions/Exposures
Participants with 0- and 1-hour high-sensitivity cardiac troponin T (hs-cTnT) measures were stratified into rule-out, observation, and rule-in zones using the ESC 0/1-hour hs-cTnT algorithm.
Main Outcomes and Measures
Cardiac death or MI at 30 days determined by expert adjudicators.
Results
During the study period, 1430 patients were accrued. In the cohort, 775 individuals (54.2%) were male, 826 (57.8%) were White, and the mean (SD) age was 57.6 (12.8) years. At 30 days, cardiac death or MI occurred in 183 participants (12.8%). Known CAD was present in 449 (31.4%). Among patients with known CAD, the ESC 0/1-hour algorithm classified 178 of 449 (39.6%) into the rule-out zone compared with 648 of 981 (66.1%) without CAD (P < .001). Among rule-out zone patients, 30-day cardiac death or MI occurred in 6 of 178 patients (3.4%) with known CAD and 7 of 648 (1.1%) without CAD (P < .001). The negative predictive value for 30-day cardiac death or MI was 96.6% (95% CI, 92.8-98.8) among patients with known CAD and 98.9% (95% CI, 97.8-99.6) in patients without known CAD (P = .04).
Conclusions and Relevance
Among patients with known CAD, the ESC 0/1-hour hs-cTnT algorithm was unable to safely exclude 30-day cardiac death or MI. This suggests that clinicians should be cautious if using the algorithm in patients with known CAD. The negative predictive value was significantly higher in patients without a history of CAD but remained less than 99%.
Introduction
Each year, more than 6.5 million patients visit an emergency department (ED) in the US due to acute chest pain.1,2,3,4 To risk stratify these patients for possible acute coronary syndrome (ACS), clinicians traditionally rely on patients’ history, cardiovascular risk factors, electrocardiogram findings, and cardiac troponin (cTn) measures.3,5,6 Known coronary artery disease (CAD) (prior myocardial infarction [MI], coronary revascularization, or ≥70% coronary stenosis) is a key risk factor for ACS.7,8,9,10,11 Prior studies suggest that CAD is associated with a nearly 4-fold increase in all-cause mortality, MI, and coronary revascularization.12,13 Therefore, many of the validated accelerated diagnostic protocols (ADPs) for patients with ACS include known CAD as a required risk stratification variable.14,15,16,17,18,19 However, the improved precision of high-sensitivity cTn (hs-cTn) assays has led to the development of ADPs that rely primarily on hs-cTn measures without protocolized consideration of patients’ other clinical variables, such as known CAD.20,21,22,23,24
The European Society of Cardiology (ESC) 0/1-hour algorithm is a troponin-only ADP designed to rule in or rule out non–ST-segment elevation MI, which is recommended by ESC and American Heart Association/American College of Cardiology guidelines.3,20,21,22,23,24 Previous studies, conducted primarily in Europe, indicate that the ESC 0/1-hour algorithm classifies nearly 60% of patients in the rule-out zone and as acceptable for early discharge with a negative predictive value (NPV) of 99% or higher.25,26,27,28,29,30,31 However, it does not differentiate risk based on the presence or absence of known CAD, and its performance in patients with known CAD is not well known, particularly in a US chest pain cohort.
To address this evidence gap, this study aims to evaluate and compare the diagnostic performance of the ESC 0/1-hour high-sensitivity cardiac troponin T (hs-cTnT) algorithm among patients with and without known CAD. The primary objective was to determine if the algorithm could achieve the NPV of 99% or higher required to safely rule out 30-day cardiac death or MI among patients with known CAD.32 A secondary objective was to evaluate and compare diagnostic performance for 30-day major adverse cardiovascular events (MACE; defined as cardiac death, MI, or coronary revascularization) in patients with and without known CAD.
Methods
Study Design and Setting
We conducted a preplanned subgroup analysis of the prospective, multicenter cohort study, the High-Sensitivity Cardiac Troponin T (Gen 5 STAT assay) to Optimize Chest Pain Risk Stratification (STOP-CP; NCT02984436). This study enrolled patients with acute chest pain or other symptoms concerning for ACS at 8 US EDs from January 25, 2017, to September 6, 2018. Study sites included the University of Florida, Wake Forest University, Henry Ford Health System, University of Maryland St Joseph Medical Center, University of Maryland Medical Center, University of Maryland Baltimore Washington Medical Center, University of California-Davis, and University of Utah. Each relevant institutional review board approved the study. Informed consent was obtained for STOP-CP enrollment. STOP-CP methods have been previously described.33 The Standards for Reporting of Diagnostic Accuracy (STARD) reporting guideline helped direct the research and manuscript development processes.34
Study Population
Patients evaluated for possible ACS who had serial troponins ordered and were 21 years or older were prospectively enrolled. Data on race and ethnicity were self-reported. Exclusion criteria included ST-segment elevation MI, systolic blood pressure less than 90 mm Hg, life expectancy of less than 90 days, a noncardiac illness requiring admission, inability to provide consent or be contacted for follow-up, and did not speak English, pregnant, or previously enrolled in the study.
Known CAD
A history of known CAD was determined prospectively at the time of enrollment by the patient’s treating clinician through medical record review and patient self-reporting. Based on prior cardiovascular trials and standardized reporting guidelines, known CAD was defined as prior MI, coronary revascularization, or 70% or more coronary stenosis.14,15,35
ESC 0/1-Hour Algorithm
Serial blood samples were collected for hs-cTnT measurement at baseline (<1 hour from first clinical blood draw) and 1 hour later (±30 minutes) in lithium heparin tubes. Study blood samples were centrifuged at 3000g at 4 °C for 15 minutes. Aliquots (1 mL) of plasma were transferred into cryovials, frozen, shipped on dry ice to the University of Maryland, and stored at −70 °C for analyses by a central laboratory. Laboratory personnel were blinded to patient details and outcomes. hs-cTnT was quantified with the Gen 5 STAT assay on the Cobase 601 analyzer (Roche Diagnostics). The assay has a range of 3 to 10 000 ng/L, limit of quantification at 6 ng/L, and a 99th percentile upper reference limit of 19 ng/L in the US with a coefficient of variation of less than 10%.36 hs-cTnT measures were for investigational use only, not for clinical care. Therefore, treating clinicians were blinded to hs-cTnT results and patient care was guided by contemporary cTn results. eTable 1 in Supplement 1 describes the contemporary troponin assays at each site.
hs-cTnT measures were used to stratify patients into rule out (0-hour hs-cTnT <6 ng/L or 0-hour <12 ng/L and Δ0/1-hour <3 ng/L), observation (any hs-cTnT value or Δ not meeting the rule-out or rule-in criteria), and rule in (0-hour ≥52 ng/L or Δ0/1-hour ≥5 ng/L) zones using the previously established assay-specific cut points of the ESC 0/1-hour algorithm (Figure 1).20,21 However, the 0-hour rule-out hs-cTnT cut point was modified from the original ESC 0/1-hour algorithm cut point of 5 ng/L (the limit of detection) to 6 ng/L (the limit of quantification) because the US Food and Drug Administration (FDA) does not allow reporting below the limit of quantification. Based on prior studies, patients stratified to the rule-out zone were expected to have 99% or higher NPV for cardiac death or MI, while patients in the rule-in zone were expected to have a high rate of events.22,25,26,27
Figure 1. The European Society of Cardiology 0/1-Hour Algorithm Among Patients With and Without Known Coronary Artery Disease (CAD) for Index Cardiac Death or Myocardial Infarction (MI), 30-Day Cardiac Death or MI, and 30-Day Major Adverse Cardiovascular Event (MACE).
Abbreviations: −LR, negative likelihood ratio, +LR, positive likelihood ratio; ACS, acute coronary syndrome; ED, emergency department; hs-cTnT, high-sensitivity cardiac troponin T; NPV, negative predictive value; PPV, positive predictive value.
Outcomes
The primary outcome was 30-day cardiac death or MI, inclusive of index visit events. Secondary outcomes included (1) MACE (cardiac death, MI, or coronary revascularization) events from index through 30 days, (2) the individual MACE subcomponents at index (including index MI) and from index through 30 days, and (3) efficacy, defined as the proportion of patients classified into the rule-out zone. Medical record review and telephone follow-up through 30 days were completed to determine outcomes. Patients who died or experienced an MI or had a contemporary cardiac troponin in the higher than 99th percentile upper reference limit were adjudicated by 4 expert reviewers (M.H.i., M.R.M., J.P.S., and J.K.M.). Adjudicators classified deaths as cardiac or noncardiac based on the Action to Control Cardiovascular Risk in Diabetes trial definition, with the exception of death due to stroke being classified as noncardiac death.37 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 1 value >99th percentile upper reference limit) with symptoms of ischemia, electrocardiogram evidence of ischemia, imaging evidence of new nonviable myocardium or a new regional wall motion abnormality, or identification of coronary thrombus by angiography.6 Coronary revascularization was defined as any coronary artery bypass grafting or angioplasty, with or without stent placement.35 Rates of coronary revascularization without MI are also reported.
Statistical Analysis
Counts, percentages, means (SDs), or medians (IQRs) ranges were used to describe the study population. To evaluate the performance of the ESC 0/1-hour algorithm, sensitivity, specificity, NPVs, and positive predictive values (PPVs) were calculated and reported with exact 95% CIs and were compared among the known CAD and no known CAD groups using Fisher exact tests. Likelihood ratios were calculated and reported with 95% CIs calculated using the method of Simel et al.38 The asymptotic hypothesis test developed by Luts et al39 was used to compare likelihood ratios among the known and no known CAD groups. Consistent with prior studies, sensitivity and NPV were calculated for rule out (ie, rule in or observation vs rule out) and specificity and PPV were calculated for rule in (ie, rule in vs observation or rule out).14,15,25,26,33 Fisher exact tests were used to compare cardiac death or MI, MACE, and efficacy rates between patients with and without known CAD at index and 30 days. All hypothesis testing was 2-sided and performed at the α = .05 significance level.
To assess the association of known CAD with index and 30-day cardiac death or MI and 30-day MACE, multivariable logistic regression was performed. Models were adjusted for age, sex, race (White vs non-White [American Indian/Alaska Native, Asian, Native Hawaiian, Black or African American, and other, which included patients who identified as any race not specified above]), hypertension, diabetes, hyperlipidemia, obesity (body mass index [calculated as weight in kilograms divided by height in meters squared], ≥30), current smoking, prior stroke, peripheral vascular disease, and end-stage kidney disease. These variables were selected due to their relevance and inclusion in previous cardiovascular risk stratification work.15 An interaction between ESC 0/1-hour algorithm classification and known CAD was also included in the models. Separate, unadjusted logistic regression models were fit to evaluate the association between known CAD and the primary and secondary outcomes within each ESC 0/1-hour zone (rule out, observation, and rule in). Unadjusted or adjusted odds ratios (aOR) with corresponding 95% CIs were calculated as appropriate for each logistic model. Sensitivity analyses were conducted to evaluate if rates of periprocedural MI varied by known CAD and if performance varied using the traditional ESC 0/1-hour hs-cTnT rule-out cut point of 5 ng/L instead of the FDA-approved 6 ng/L. An additional analysis examined algorithm performance when including a History, Electrocardiogram, Age, Risk factor, and Troponin (HEART) score. In this analysis, patients were classified as ruled out only if they had both a HEART score of 3 or lower and were classified to the rule-out zone by the ESC 0/1-hour hs-cTnT algorithm.40 Patients classified into the rule-out zone by the ESC 0/1-hour algorithm but with a HEART score of 4 to 6 were classified to the observation zone. Patients classified to the rule-out or observation zone by the ESC 0/1-hour algorithm but with a HEART score of 7 or higher were classified to the rule-in zone. A subgroup analysis accounting for early (≤3 hours) vs late presentation (>3 hours) to the ED after chest pain onset was also performed. All analyses were performed using SAS statistical software version 9.4 (SAS Institute Inc) or R version 3.6.2 (R Foundation) with the epiR package. Analysis took place between February and December 2022.
Results
This preplanned subgroup analysis included 1430 patients, of which 449 (31.4%) had known CAD. The eFigure in Supplement 1 shows the study flow diagram. The overall study cohort included 775 male individuals (54.2%) and 826 White individuals (57.8%), with a mean (SD) age of 57.6 (12.8) years. Demographics of patients with and without known CAD are described in Table 1. At 30 days, cardiac death or MI occurred in 183 individuals (12.8%). Cardiac death or MI at 30 days occurred in 88 patients with known CAD (19.6%) compared with 95 patients without CAD (9.7%) (P < .001). MACE was more frequent at 30 days in patients with known CAD vs those without (104 of 449 [23.2%] vs 99 of 981 [10.1%]; P < .001).
Table 1. Cohort Characteristics.
| Characteristic | No. (%) | P valuea | ||
|---|---|---|---|---|
| Known CAD (n = 449) | No known CAD (n = 981) | Total (N = 1430) | ||
| Age, mean (SD), y | 62.3 (11.9) | 55.4 (12.6) | 57.6 (12.8) | <.001 |
| Sex | ||||
| Male | 297 (66.2) | 478 (48.7) | 775 (54.2) | <.001 |
| Female | 152 (33.9) | 503 (51.3) | 655 (45.8) | |
| Race and ethnicity | ||||
| American Indian/Alaska Native | 7 (1.6) | 16 (1.6) | 23 (1.6) | <.001 |
| Asian | 0 | 12 (1.2) | 12 (0.8) | |
| Black or African American | 132 (29.4) | 392 (40.0) | 524 (36.6) | |
| Hispanic or Latino | 16 (3.6) | 41 (4.2) | 57 (4.0) | |
| Native Hawaiian | 1 (0.2) | 1 (0.1) | 2 (0.1) | |
| White | 293 (65.3) | 533 (54.3) | 826 (57.8) | |
| Not Hispanic or Latino | 430 (95.8) | 929 (94.7) | 1359 (95.0) | |
| Otherb | 16 (3.6) | 27 (2.8) | 43 (3.0) | |
| Unknown | 3 (0.7) | 11 (1.1) | 14 (1.0) | |
| Risk factors | ||||
| Current or history of smoking | 282 (62.8) | 509 (51.9) | 791 (55.3) | <.001 |
| Hypertension | 380 (84.6) | 571 (58.5) | 954 (66.7) | <.001 |
| Hyperlipidemia | 322 (71.7) | 361 (36.8) | 683 (47.8) | <.001 |
| Diabetes | 189 (42.1) | 231 (23.6) | 420 (29.4) | <.001 |
| Family history of coronary disease | 253 (56.4) | 408 (41.6) | 661 (46.2) | <.001 |
| BMI >30 | 237 (52.8) | 511 (52.1) | 748 (52.3) | .81 |
| Prior cerebrovascular accident | 93 (20.7) | 61 (6.2) | 154 (10.8) | <.001 |
| Prior peripheral vascular disease | 49 (10.9) | 40 (4.1) | 89 (6.2) | <.001 |
| Prior end-stage kidney disease | 36 (8.0) | 35 (3.6) | 71 (5.0) | <.001 |
| Chest pain onset, hours from arrival | ||||
| ≤3 | 153 (34.4) | 353 (36.0) | 505 (35.5) | .44 |
| >3 | 292 (65.6) | 625 (64.0) | 917 (65.5) | |
| ECG at arrival | <.001 | |||
| Ischemic | 42 (9.4) | 46 (4.7) | 88 (6.2) | |
| Nonischemic | 407 (90.7) | 935 (95.3) | 1342 (93.9) | |
| Initial study hs-cTnT sample, median (IQR), ng/L | 15 (8-35) | 7 (4-16) | 9.0 (5-21) | <.001 |
Abbreviations: BMI, body mass index (calculated as weight in kilograms divided by height in meters squared); CAD, coronary artery disease; ECG, electrocardiogram; hs-cTnT, high-sensitivity troponin T.
Categorical variables were compared between the known CAD and no known CAD groups using χ2 or Fisher exact test. The Wilcoxon rank sum test was used to compare age and hs-cTnT values.
Other included patients who did not identify with any of the specified races.
The ESC 0/1-hour algorithm classified 178 of 449 patients (39.6%) with known CAD into the rule-out zone compared with 648 of 981 patients (66.1%) with no known CAD (P < .001). At 30 days, cardiac death or MI occurred in 6 of 178 rule-out patients (3.4%) with known CAD vs 7 of 648 patients (1.1%) without known CAD (odds ratio [OR], 3.19; 95% CI, 1.05-9.63; P = .04). MACE at 30 days occurred in 14 of 178 rule-out patients (7.9%) with known CAD vs 9 of 648 (1.4%) without known CAD (OR, 6.06; 95% CI, 2.59-14.25; P < .001). The NPV for 30-day cardiac death or MI was 96.6% (95% CI, 92.8-98.8) in patients with known CAD and 98.9% (97.8-99.6) in patients without known CAD (P = .04). Sensitivity for 30-day cardiac death or MI was similar in patients with known CAD (93.2%; 95% CI, 85.7-97.5) vs those without known CAD (92.6%; 95% CI, 85.4-97.0; P = 1.0). Events stratified by ESC 0/1-hour rule-out, observation, and rule-in zones are summarized in Table 2. The diagnostic performance of the algorithm is summarized in Figure 1 and Table 3.
Table 2. Safety Events Among European Society of Cardiology 0/1-Hour High-Sensitivity Cardiac Troponin T Algorithm Patients With and Without Known CAD at 30 Days.
| Event | Known CAD | No known CAD | Total | Odds ratio (95% CI) |
|---|---|---|---|---|
| Rule out | ||||
| No. | 178 | 648 | 826 | NA |
| Index | ||||
| Cardiac death | 1 (0.6) | 0 | 1 (0.1) | NA |
| MI | 3 (1.7) | 5 (0.8) | 8 (1.0) | 2.20 (0.52-9.31) |
| Revascularization | 8 (4.5) | 2 (0.3) | 10 (1.2) | 15.20 (3.20-72.24) |
| Revascularization without MI | 6 (3.4) | 1 (0.2) | 7 (0.9) | 22.57 (2.70-188.72) |
| Cardiac death or MI | 4 (2.3) | 5 (0.8) | 9 (1.1) | 2.96 (0.79-11.13) |
| MACE (cardiac death + MI + revascularization) | 10 (5.6) | 6 (0.9) | 16 (1.9) | 6.37 (2.28-17.77) |
| 30-d (Index + follow-up) | ||||
| Cardiac death | 1 (0.6) | 0 | 1 (0.1) | NA |
| MI | 5 (2.8) | 7 (1.1) | 12 (1.5) | 2.64 (0.83-8.44) |
| Revascularization | 10 (5.6) | 3 (0.5) | 13 (1.6) | 12.80 (3.48-47.02) |
| Revascularization without MI | 8 (4.5) | 2 (0.3) | 10 (1.2) | 15.20 (3.20-72.24) |
| Cardiac death or MI | 6 (3.4) | 7 (1.1) | 13 (1.6) | 3.19 (1.05-9.63) |
| MACE (cardiac death + MI + revascularization) | 14 (7.9) | 9 (1.4) | 23 (2.8) | 6.06 (2.58-14.25) |
| Observation | ||||
| No. | 183 | 231 | 414 | NA |
| Index | ||||
| Cardiac death | 0 | 0 | 0 | NA |
| MI | 29 (15.9) | 22 (9.5) | 51 (12.3) | 1.79 (0.99-3.23) |
| Revascularization | 14 (7.7) | 6 (2.6) | 20 (4.8) | 3.11 (1.17-8.25) |
| Revascularization without MI | 4 (2.2) | 0 | 4 (1.0) | NA |
| Cardiac death or MI | 29 (15.9) | 22 (9.5) | 51 (12.3) | 1.79 (0.99-3.23) |
| MACE (cardiac death + MI + revascularization) | 33 (18.0) | 22 (9.5) | 55 (13.3) | 2.09 (1.17-3.73) |
| 30-d (Index + follow-up) | ||||
| Cardiac death | 0 | 2 (0.9) | 2 (0.5) | NA |
| MI | 30 (16.4) | 27 (11.7) | 57 (13.8) | 1.48 (0.85-2.60) |
| Revascularization | 19 (10.4) | 10 (4.3) | 29 (7.0) | 2.56 (1.16-5.65) |
| Revascularization without MI | 8 (4.4) | 2 (0.9) | 10 (2.4) | 5.24 (1.10-24.96) |
| Cardiac death or MI | 30 (16.4) | 29 (12.6) | 59 (14.3) | 1.37 (0.79-2.37) |
| MACE (cardiac death + MI + revascularization) | 38 (20.8) | 31 (13.4) | 69 (16.7) | 1.69 (1.00-2.84) |
| Rule in | ||||
| No. | 88 | 102 | 190 | NA |
| Index | ||||
| Cardiac death | 1 (1.1) | 0 | 1 (0.5) | NA |
| MI | 47 (53.4) | 58 (56.9) | 105 (55.3) | 0.87 (0.49-1.54) |
| Revascularization | 18 (20.5) | 20 (19.6) | 38 (20.0) | 1.05 (0.52-2.15) |
| Revascularization without MI | 0 | 0 | 0 | NA |
| Cardiac death or MI | 47 (53.4) | 58 (56.9) | 105 (55.3) | 0.87 (0.49-1.54) |
| MACE (cardiac death + MI + revascularization) | 47 (53.4) | 58 (56.9) | 105 (55.3) | 0.87 (0.49-1.54) |
| 30-d (Index + follow-up) | ||||
| Cardiac death | 4 (4.6) | 2 (2.0) | 6 (3.2) | 2.38 (0.43-13.3) |
| MI | 51 (58.0) | 58 (56.9) | 109 (57.4) | 1.05 (0.59-1.86) |
| Revascularization | 0 | 0 | 0 | NA |
| Revascularization without MI | 19 (21.6) | 24 (23.5) | 43 (22.6) | 0.89 (0.45-1.77) |
| Cardiac death or MI | 52 (59.1) | 59 (57.8) | 111 (58.4) | 1.05 (0.59-1.88) |
| MACE (cardiac death + MI + revascularization) | 52 (59.1) | 59 (57.8) | 111 (58.4) | 1.05 (0.59-1.88) |
Abbreviations: CAD, coronary artery disease; MACE, major adverse cardiovascular event; MI, myocardial infarction; NA, not applicable.
Table 3. Test Characteristics of the European Society of Cardiology 0/1-Hour High-Sensitivity Cardiac Troponin T Algorithm for 30-Day Cardiac Death or MI and MACE in Patients With and Without Known CAD.
| Characteristic | % (95% CI) | |||
|---|---|---|---|---|
| 30-d Cardiac death or MI | 30-d MACE | |||
| Known CAD | No known CAD | Known CAD | No known CAD | |
| Rule out | ||||
| Sensitivity | 93.2 (85.7-97.5) | 92.6 (85.4-97.0) | 86.5 (78.4-92.4) | 90.9 (83.4-95.8) |
| NPV | 96.6 (92.8-98.8) | 98.9 (97.8-99.6) | 92.1 (87.1-95.6) | 98.6 (97.4-99.4) |
| Rule in | ||||
| Specificity | 90.0 (86.5-92.9) | 95.1 (93.5-96.5) | 89.6 (85.7-92.6) | 95.1 (93.5-96.4) |
| PPV | 59.1 (48.1-69.4) | 57.8 (47.7-67.6) | 59.1 (48.1-69.5) | 57.8 (47.7-67.6) |
Abbreviations: CAD, coronary artery disease; MACE, major adverse cardiovascular event; MI, myocardial infarction; NPV, negative predictive value; PPV, positive predictive value.
The ESC 0/1-hour algorithm classified 88 of 449 patients (19.6%) with CAD into the rule-in zone vs 102 of 981 patients (10.4%) without CAD (P < .001). Among rule-in patients, the rate of index cardiac death or MI was similar between patients with CAD and those without (47 of 88 [53.4%] vs 58 of 102 [56.9%]; OR, 0.87; 95% CI, 0.49-1.54; P = .66). At 30 days, 52 of 88 patients (59.1%) with known CAD experienced cardiac death or MI compared with 59 of 102 patients (57.8%) without known CAD (OR, 1.05; 95% CI, 0.59-1.88; P = .88). The PPV for 30-day cardiac death or MI was 59.1% (95% CI, 48.1-69.4) in patients with known CAD and 57.8% (95% CI, 47.7-67.6) in patients without known CAD (P = .88). The positive likelihood ratio was lower for patients with known CAD compared with those without (5.9; 95% CI, 4.2-8.5 vs 12.8; 95% CI, 9.2-17.8; P = .002). The test characteristics for the rule-in zone are presented in Figure 1 and Table 3.
The interaction between the ESC 0/1-hour algorithm and known CAD was not significant for 30-day cardiac death or MI (P = .20) but was significant for 30-day MACE (P = .006). Known CAD was associated with increased 30-day MACE events among patients classified to the rule-out zone by the ESC 0/1-hour algorithm (aOR, 6.08; 95% CI, 2.51-14.72). However, in those classified to the observation and rule-in zones, 30-day MACE rates were similar in those with or without known CAD (aOR, 1.71; 95% CI, 0.99-2.95 and aOR, 1.07; 95% CI, 0.54-2.09, respectively). Figure 2 demonstrates the interaction. eTable 2 in Supplement 1 shows the aORs for index and 30-day cardiac death or MI and MACE.
Figure 2. Interaction Between European Society of Cardiology (ESC) 0/1-Hour Algorithm Classification and Known Coronary Artery Disease (CAD) With Adjusted Odds Ratios (aOR).
aOR was adjusted for age, sex, race, hypertension, diabetes, hyperlipidemia, obesity, current smoking, prior stroke, peripheral vascular disease, and end-stage kidney disease as well as the interaction between ESC 0/1-hour classification and known CAD. hs-cTnT indicates high-sensitivity cardiac troponin T; MACE, major adverse cardiovascular event.
A sensitivity analysis was conducted for periprocedural MIs. However, only 1 patient, who had known CAD, experienced a periprocedural MI. The sensitivity analysis using the traditional ESC 0/1-hour hs-cTnT rule-out cut point of 5 ng/L instead of the FDA-approved 6 ng/L resulted in a single patient, who had no known CAD, being reclassified from the rule-out zone to the observation zone. eTable 3 in Supplement 1 shows the analyses for the ESC 0/1-hour hs-cTnT algorithm combined with the HEART score with the test characteristics for this approach summarized in eTable 4 in Supplement 1. eTables 5 and 6 in Supplement 1 show the subgroup analyses for ESC 0/1-hour hs-cTnT performance based on early (≤3 hours of chest pain) vs late presentation (>3 hours of chest pain).
Discussion
The primary finding of this multisite, prospective study is that the ESC 0/1-hour hs-cTnT algorithm had an NPV below the accepted 99% threshold for 30-day cardiac death or MI among patients with known CAD. The algorithm classified 39.6% of patients with known CAD into the rule-out zone, yielding an NPV of 96.6% with an upper bound of the 95% CI less than 99%. This 3.4% missed 30-day death or MI rate is higher than the 1% miss rate most emergency clinicians are generally willing to accept.32 Thus, our results suggest that the ESC 0/1-hour hs-cTnT algorithm does not safely rule out ACS in patients with known CAD. Furthermore, while the NPV for 30-day cardiac death or MI of the ESC 0/1-hour hs-cTnT algorithm was significantly higher among patients without a history of CAD compared with those with known CAD, the point estimate for NPV in patients without CAD remained less than 99%. Similarly, the NPV of 92.1% for 30-day MACE among patients with known CAD was lower than the accepted safety threshold. Given that the ESC 0/1-hour hs-cTnT algorithm achieved less than 99% NPV for 30-day cardiac death or MI and MACE among patients with and without known CAD, it is likely not safe for routine use among US ED patients with chest pain, regardless of CAD status.
Current 2020 ESC guidelines recommend using the ESC 0/1-hour algorithm to evaluate ED patients with acute chest pain (class I recommendation).20 While these guidelines recommend that an electrocardiogram and clinical impression be used to inform care, there is no protocolized consideration of any clinical variable except hs-cTnT, thus potentially exposing patients to increased risk. Based on results from this study, emergency clinicians should be cautious when using the ESC 0/1-hour hs-cTnT algorithm among patients with known CAD. Failing to consider known CAD in those stratified to the rule-out zone could increase missed cardiac events among those discharged from the ED. This is meaningful because patients discharged from the ED with missed ACS have twice the mortality rate of patients who are admitted.41,42 Furthermore, missed ACS is a top cause of litigation against ED clinicians.43
In our multivariable logistic model, there was not a significant interaction between the ESC 0/1-hour algorithm classification and known CAD for 30-day cardiac death or MI. However, there was a significant interaction for 30-day MACE. In the rule-out zone, patients with known CAD were more likely to experience 30-day MACE compared with those without known CAD. There was no statistically significant difference in 30-day MACE rates between patients with and without known CAD in the observation and rule-in zones. However, there was an absolute difference in 30-day MACE of 7.4% between observation zone patients with and without known CAD. This potentially clinically significant difference was likely not statistically significant because the smaller number of patients in the observation zone limited power for this comparison. These interaction findings suggest that known CAD is an important risk factor for 30-day MACE among rule-out patients, but it is not associated with increased odds of 30-day MACE in patients classified to the rule-in zone by the ESC 0/1-hour hs-cTnT algorithm.
Many risk scores and ADPs consider if a patient has known CAD. The HEART score,40 HEART Pathway ADP,14,15 and Emergency Department Assessment of Chest Pain ADP16 each weigh a history of known CAD. The HEART score and HEART Pathway are widely used and validated around the world, with the HEART Pathway having an excellent safety profile.14,15,44,45,46,47 In the HEART Pathway, known CAD excludes a patient from early discharge. Prior research in the STOP-CP cohort demonstrated that adding a low-risk HEART score (0-3) to the ESC 0/1-hour hs-cTnT algorithm improved NPV from 96.8% to 98.2% overall. However, efficacy decreased from 55.3% to 33.7%.33 Similarly, others report that adding clinical variables to the ESC 0/1-hour algorithm improves safety but at the cost of efficacy.48,49 However, in our cohort the addition of the HEART score to the ESC 0/1-hour hs-cTnT algorithm decreased the proportion of patients ruled out with only minimal improvement in NPV for 30-day cardiac death or MI among patients with known CAD. These findings suggest that even when combined with the HEART score, the ESC 0/1-hour algorithm is unable to safely exclude 30-day cardiac death or MI among patients with known CAD.
Results from our analysis differ from prior international studies, which found similar diagnostic performance of the ESC 0/1-hour algorithm among patients with and without CAD. Subgroup analyses of 2 trials found that the ESC 0/1-hour hs-cTnT algorithm had an NPV of 99% or higher regardless of CAD history.25,26 Several possible reasons exist for these differences. One of these trials used index MI as the primary outcome rather than 30-day cardiac death or MI.22,26 The other study enrolled patients at 2 sites in Switzerland and Argentina, with patient populations that may not be generalizable to a US ED population.25 The STOP-CP study enrolled patients across a mixture of US community and academic care centers. Additionally, the STOP-CP patient population was racially diverse, with more than 40% being a race or ethnicity other than White.
The PPV of the ESC 0/1-hour hs-cTnT algorithm was similar in patients with and without known CAD, being less than 60% both for index and 30-day cardiac death or MI. These results differ from prior studies.22,26,28,50 The original ESC 0/1-hour study reported a PPV of 76%.22 Twerenbold et al26 found that PPV was significantly higher in patients with known CAD. Our lower PPV findings, regardless of known CAD, suggest that the ESC 0/1-hour hs-cTnT algorithm may result in greater overtriage in a US population. Surprisingly, the positive likelihood ratio for patients with known CAD was significantly less for patients without known CAD. This is likely because patients with underlying heart disease are more likely to have comorbidities such as end-stage kidney disease, hypertension, and heart failure,5,51 which can result in chronic or acute myocardial injury. Thus, patients with preexisting heart disease are more likely to have hs-cTn measures above the ESC 0/1-hour hs-cTnT algorithm’s 52 ng/L rule-in zone without having ACS.5
Limitations
Although this study was conducted at 8 US 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. Known CAD was determined by medical record review and patient self-reporting, thereby possibly exposing the study to misclassification bias. The 30-day cardiac death or MI and MACE rates were higher than in previous US cohorts.14,29 As required by the FDA, the lowest reportable value for the hs-cTnT assay was 6 ng/L. Outside of the US, the rule-out baseline hs-cTnT value is 5 ng/L. In our sensitivity analysis, only 1 patient was reclassified based on this difference. However, this cut point difference may affect the generalizability of these results outside of the US. This study used only the Roche hs-cTnT assay. Therefore, these conclusions cannot be applied to ESC 0/1-hour hs-cTnI algorithm derivations. The sample size of patients with known CAD classified to the rule-out zone was modest. Combined with a low event rate, this contributed to wide confidence intervals around event rates. Finally, this study was observational and as such, the ESC 0/1-hour hs-cTnT algorithm was not used to guide patient care.
Conclusions
The ESC 0/1-hour hs-cTnT algorithm was unable to safely rule out 30-day cardiac death or MI in this multisite US cohort in patients with known CAD. Among patients with known CAD, the NPV was 96.6% for 30-day cardiac death or MI, indicating that the miss rate was higher than most US clinicians are willing to accept. These results suggest that emergency clinicians should use caution when using the ESC 0/1-hour hs-cTnT algorithm to risk stratify patients with known CAD as the algorithm may not be able to safely rule out cardiac events in this key patient subgroup.
eTable 1. Contemporary troponin analytical data by clinical site
eTable 2. Adjusted odds ratios (aOR) for safety events among ESC 0/1-hour hs-cTnT algorithm patients with known CAD vs. no known CAD
eTable 3. Safety events when using the ESC 0/1-hour hs-cTnT algorithm combined with the HEART score (rule-out ≤ 3, observation 4-6, rule-in ≥ 7)
eTable 4. Test characteristics of the ESC 0/1-hour hs-cTnT algorithm combined with the HEART score (rule-out ≤ 3, observation 4-6, rule-in ≥7)
eTable 5. Subgroup analysis for early presenters (≤ 3 hours of chest pain until ED evaluation)
eTable 6. Subgroup analysis for late presenters (> 3 hours of chest pain until ED evaluation)
eFigure. Study flow diagram
Data sharing statement
References
- 1.Owens PL, Barrett ML, Gibson TB, Andrews RM, Weinick RM, Mutter RL. Emergency department care in the United States: a profile of national data sources. Ann Emerg Med. 2010;56(2):150-165. doi: 10.1016/j.annemergmed.2009.11.022 [DOI] [PubMed] [Google Scholar]
- 2.Benjamin EJ, Muntner P, Alonso A, et al. ; American Heart Association Council on Epidemiology and Prevention Statistics Committee and Stroke Statistics Subcommittee . Heart disease and stroke statistics-2019 update: a report from the American Heart Association. Circulation. 2019;139(10):e56-e528. doi: 10.1161/CIR.0000000000000659 [DOI] [PubMed] [Google Scholar]
- 3.Gulati M, Levy PD, Mukherjee D, et al. ; Writing Committee Members . 2021 AHA/ACC/ASE/CHEST/SAEM/SCCT/SCMR guideline for the evaluation and diagnosis of chest pain: a report of the American College of Cardiology/American Heart Association Joint Committee on Clinical Practice Guidelines. J Am Coll Cardiol. 2021;78(22):e187-e285. doi: 10.1016/j.jacc.2021.07.053 [DOI] [PubMed] [Google Scholar]
- 4.National hospital ambulatory medical care survey: 2017 emergency department summary tables. National Center for Health Statistics. Accessed January 20, 2023. https://www.cdc.gov/nchs/data/nhamcs/web_tables/2017_ed_web_tables-508.pdf
- 5.Januzzi JL Jr, Mahler SA, Christenson RH, et al. Recommendations for institutions transitioning to high-sensitivity troponin testing: JACC scientific expert panel. J Am Coll Cardiol. 2019;73(9):1059-1077. doi: 10.1016/j.jacc.2018.12.046 [DOI] [PubMed] [Google Scholar]
- 6.Thygesen K, Alpert JS, Jaffe AS, et al. ; Executive Group on behalf of the Joint European Society of Cardiology (ESC)/American College of Cardiology (ACC)/American Heart Association (AHA)/World Heart Federation (WHF) Task Force for the Universal Definition of Myocardial Infarction . Fourth universal definition of myocardial infarction (2018). J Am Coll Cardiol. 2018;72(18):2231-2264. doi: 10.1016/j.jacc.2018.08.1038 [DOI] [PubMed] [Google Scholar]
- 7.Ringqvist I, Fisher LD, Mock M, et al. Prognostic value of angiographic indices of coronary artery disease from the Coronary Artery Surgery Study (CASS). J Clin Invest. 1983;71(6):1854-1866. doi: 10.1172/JCI110941 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Litt HI, Gatsonis C, Snyder B, et al. CT angiography for safe discharge of patients with possible acute coronary syndromes. N Engl J Med. 2012;366(15):1393-1403. doi: 10.1056/NEJMoa1201163 [DOI] [PubMed] [Google Scholar]
- 9.Ostrom MP, Gopal A, Ahmadi N, et al. Mortality incidence and the severity of coronary atherosclerosis assessed by computed tomography angiography. J Am Coll Cardiol. 2008;52(16):1335-1343. doi: 10.1016/j.jacc.2008.07.027 [DOI] [PubMed] [Google Scholar]
- 10.Aldrovandi A, Maffei E, Seitun S, et al. Major adverse cardiac events and the severity of coronary atherosclerosis assessed by computed tomography coronary angiography in an outpatient population with suspected or known coronary artery disease. J Thorac Imaging. 2012;27(1):23-28. doi: 10.1097/RTI.0b013e3181f55d0d [DOI] [PubMed] [Google Scholar]
- 11.Min JK, Shaw LJ, Devereux RB, et al. Prognostic value of multidetector coronary computed tomographic angiography for prediction of all-cause mortality. J Am Coll Cardiol. 2007;50(12):1161-1170. doi: 10.1016/j.jacc.2007.03.067 [DOI] [PubMed] [Google Scholar]
- 12.van Werkhoven JM, Schuijf JD, Gaemperli O, et al. Prognostic value of multislice computed tomography and gated single-photon emission computed tomography in patients with suspected coronary artery disease. J Am Coll Cardiol. 2009;53(7):623-632. doi: 10.1016/j.jacc.2008.10.043 [DOI] [PubMed] [Google Scholar]
- 13.Pundziute G, Schuijf JD, Jukema JW, et al. Prognostic value of multislice computed tomography coronary angiography in patients with known or suspected coronary artery disease. J Am Coll Cardiol. 2007;49(1):62-70. doi: 10.1016/j.jacc.2006.07.070 [DOI] [PubMed] [Google Scholar]
- 14.Mahler SA, Riley RF, Hiestand BC, et al. The HEART Pathway randomized trial: identifying emergency department patients with acute chest pain for early discharge. Circ Cardiovasc Qual Outcomes. 2015;8(2):195-203. doi: 10.1161/CIRCOUTCOMES.114.001384 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Mahler SA, Lenoir KM, Wells BJ, et al. Safely identifying emergency department patients with acute chest pain for early discharge. Circulation. 2018;138(22):2456-2468. doi: 10.1161/CIRCULATIONAHA.118.036528 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Than M, Flaws D, Sanders S, et al. Development and validation of the Emergency Department Assessment of Chest Pain Score and 2 h accelerated diagnostic protocol. Emerg Med Australas. 2014;26(1):34-44. doi: 10.1111/1742-6723.12164 [DOI] [PubMed] [Google Scholar]
- 17.Hess EP, Brison RJ, Perry JJ, et al. Development of a clinical prediction rule for 30-day cardiac events in emergency department patients with chest pain and possible acute coronary syndrome. Ann Emerg Med. 2012;59(2):115-25.e1. doi: 10.1016/j.annemergmed.2011.07.026 [DOI] [PubMed] [Google Scholar]
- 18.Christenson J, Innes G, McKnight D, et al. A clinical prediction rule for early discharge of patients with chest pain. Ann Emerg Med. 2006;47(1):1-10. doi: 10.1016/j.annemergmed.2005.08.007 [DOI] [PubMed] [Google Scholar]
- 19.Scheuermeyer FX, Innes G, Grafstein E, et al. Safety and efficiency of a chest pain diagnostic algorithm with selective outpatient stress testing for emergency department patients with potential ischemic chest pain. Ann Emerg Med. 2012;59(4):256-264.e3. doi: 10.1016/j.annemergmed.2011.10.016 [DOI] [PubMed] [Google Scholar]
- 20.Collet JP, Thiele H. The ‘Ten Commandments’ for the 2020 ESC Guidelines for the management of acute coronary syndromes in patients presenting without persistent ST-segment elevation. Eur Heart J. 2020;41(37):3495-3497. doi: 10.1093/eurheartj/ehaa624 [DOI] [PubMed] [Google Scholar]
- 21.Roffi M, Patrono C, Collet JP, et al. ; ESC Scientific Document Group . 2015 ESC Guidelines for the management of acute coronary syndromes in patients presenting without persistent ST-segment elevation: Task Force for the Management of Acute Coronary Syndromes in Patients Presenting without Persistent ST-Segment Elevation of the European Society of Cardiology (ESC). Eur Heart J. 2016;37(3):267-315. doi: 10.1093/eurheartj/ehv320 [DOI] [PubMed] [Google Scholar]
- 22.Reichlin T, Schindler C, Drexler B, et al. One-hour rule-out and rule-in of acute myocardial infarction using high-sensitivity cardiac troponin T. Arch Intern Med. 2012;172(16):1211-1218. doi: 10.1001/archinternmed.2012.3698 [DOI] [PubMed] [Google Scholar]
- 23.Mueller C, Twerenbold R, Reichlin T. Early diagnosis of myocardial infarction with sensitive cardiac troponin assays. Clin Chem. 2019;65(3):490-491. doi: 10.1373/clinchem.2018.298638 [DOI] [PubMed] [Google Scholar]
- 24.Vasile VC, Jaffe AS. High-sensitivity cardiac troponin for the diagnosis of patients with acute coronary syndromes. Curr Cardiol Rep. 2017;19(10):92. doi: 10.1007/s11886-017-0904-4 [DOI] [PubMed] [Google Scholar]
- 25.Twerenbold R, Costabel JP, Nestelberger T, et al. Outcome of applying the ESC 0/1-hour algorithm in patients with suspected myocardial infarction. J Am Coll Cardiol. 2019;74(4):483-494. doi: 10.1016/j.jacc.2019.05.046 [DOI] [PubMed] [Google Scholar]
- 26.Twerenbold R, Neumann JT, Sörensen NA, et al. Prospective validation of the 0/1-h algorithm for early diagnosis of myocardial infarction. J Am Coll Cardiol. 2018;72(6):620-632. doi: 10.1016/j.jacc.2018.05.040 [DOI] [PubMed] [Google Scholar]
- 27.Burgos LM, Trivi M, Costabel JP. Performance of the European Society of Cardiology 0/1-hour algorithm in the diagnosis of myocardial infarction with high-sensitivity cardiac troponin: systematic review and meta-analysis. Eur Heart J Acute Cardiovasc Care. Published online June 29, 2020. doi: 10.1177/2048872620935399 [DOI] [PubMed] [Google Scholar]
- 28.Pickering JW, Greenslade JH, Cullen L, et al. Assessment of the European Society of Cardiology 0-Hour/1-hour algorithm to rule-out and rule-in acute myocardial infarction. Circulation. 2016;134(20):1532-1541. doi: 10.1161/CIRCULATIONAHA.116.022677 [DOI] [PubMed] [Google Scholar]
- 29.Nowak RM, Christenson RH, Jacobsen G, et al. Performance of novel high-sensitivity cardiac troponin i assays for 0/1-hour and 0/2- to 3-hour evaluations for acute myocardial infarction: results from the HIGH-US Study. Ann Emerg Med. 2020;76(1):1-13. doi: 10.1016/j.annemergmed.2019.12.008 [DOI] [PubMed] [Google Scholar]
- 30.Chapman AR, Anand A, Boeddinghaus J, et al. Comparison of the efficacy and safety of early rule-out pathways for acute myocardial infarction. Circulation. 2017;135(17):1586-1596. doi: 10.1161/CIRCULATIONAHA.116.025021 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31.Chapman AR, Fujisawa T, Lee KK, et al. Novel high-sensitivity cardiac troponin I assay in patients with suspected acute coronary syndrome. Heart. 2019;105(8):616-622. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32.Than M, Herbert M, Flaws D, et al. What is an acceptable risk of major adverse cardiac event in chest pain patients soon after discharge from the emergency department?: a clinical survey. Int J Cardiol. 2013;166(3):752-754. doi: 10.1016/j.ijcard.2012.09.171 [DOI] [PubMed] [Google Scholar]
- 33.Allen BR, Christenson RH, Cohen SA, et al. Diagnostic performance of high-sensitivity cardiac troponin T strategies and clinical variables in a multisite US cohort. Circulation. 2021;143(17):1659-1672. doi: 10.1161/CIRCULATIONAHA.120.049298 [DOI] [PubMed] [Google Scholar]
- 34.STARD 2015: an updated list of essential items for reporting diagnostic accuracy studies. Equator Network. Accessed March 29, 2022. https://www.equator-network.org/reporting-guidelines/stard/
- 35.Hollander JE, Blomkalns AL, Brogan GX, et al. ; Multidisciplinary Standardized Reporting Criteria Task Force; Standardized Reporting Criteria Working Group of Emergency Medicine Cardiac Research and Education Group-International . Standardized reporting guidelines for studies evaluating risk stratification of emergency department patients with potential acute coronary syndromes. Ann Emerg Med. 2004;44(6):589-598. doi: 10.1016/j.annemergmed.2004.08.009 [DOI] [PubMed] [Google Scholar]
- 36.Fitzgerald RL, Hollander JE, Peacock WF, et al. Analytical performance evaluation of the Elecsys® Troponin T Gen 5 STAT assay. Clin Chim Acta. 2019;495:522-528. doi: 10.1016/j.cca.2019.05.026 [DOI] [PubMed] [Google Scholar]
- 37.Gerstein HC, Miller ME, Byington RP, et al. ; Action to Control Cardiovascular Risk in Diabetes Study Group . Effects of intensive glucose lowering in type 2 diabetes. N Engl J Med. 2008;358(24):2545-2559. doi: 10.1056/NEJMoa0802743 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 38.Simel DL, Samsa GP, Matchar DB. Likelihood ratios with confidence: sample size estimation for diagnostic test studies. J Clin Epidemiol. 1991;44(8):763-770. doi: 10.1016/0895-4356(91)90128-V [DOI] [PubMed] [Google Scholar]
- 39.Luts J, Nofuentes JAR, del Castillo J de DL, Van Huffel S. Asymptotic hypothesis test to compare likelihood ratios of multiple diagnostic tests in unpaired designs. J Stat Plan Inference. 2011;141(11):3578-3594. doi: 10.1016/j.jspi.2011.05.010 [DOI] [Google Scholar]
- 40.Six AJ, Backus BE, Kelder JC. Chest pain in the emergency room: value of the HEART score. Neth Heart J. 2008;16(6):191-196. doi: 10.1007/BF03086144 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 41.Lee TH, Rouan GW, Weisberg MC, et al. Clinical characteristics and natural history of patients with acute myocardial infarction sent home from the emergency room. Am J Cardiol. 1987;60(4):219-224. doi: 10.1016/0002-9149(87)90217-7 [DOI] [PubMed] [Google Scholar]
- 42.Pope JH, Aufderheide TP, Ruthazer R, et al. Missed diagnoses of acute cardiac ischemia in the emergency department. N Engl J Med. 2000;342(16):1163-1170. doi: 10.1056/NEJM200004203421603 [DOI] [PubMed] [Google Scholar]
- 43.Brown TW, McCarthy ML, Kelen GD, Levy F. An epidemiologic study of closed emergency department malpractice claims in a national database of physician malpractice insurers. Acad Emerg Med. 2010;17(5):553-560. doi: 10.1111/j.1553-2712.2010.00729.x [DOI] [PubMed] [Google Scholar]
- 44.Greenslade JH, Carlton EW, Van Hise C, et al. Diagnostic accuracy of a new high-sensitivity troponin I assay and five accelerated diagnostic pathways for ruling out acute myocardial infarction and acute coronary syndrome. Ann Emerg Med. 2018;71(4):439-451.e3. doi: 10.1016/j.annemergmed.2017.10.030 [DOI] [PubMed] [Google Scholar]
- 45.Fernando SM, Tran A, Cheng W, et al. Prognostic accuracy of the HEART score for prediction of major adverse cardiac events in patients presenting with chest pain: a systematic review and meta-analysis. Acad Emerg Med. 2019;26(2):140-151. doi: 10.1111/acem.13649 [DOI] [PubMed] [Google Scholar]
- 46.Van Den Berg P, Body R. The HEART score for early rule out of acute coronary syndromes in the emergency department: a systematic review and meta-analysis. Eur Heart J Acute Cardiovasc Care. 2018;7(2):111-119. doi: 10.1177/2048872617710788 [DOI] [PubMed] [Google Scholar]
- 47.Six AJ, Cullen L, Backus BE, et al. The HEART score for the assessment of patients with chest pain in the emergency department: a multinational validation study. Crit Pathw Cardiol. 2013;12(3):121-126. doi: 10.1097/HPC.0b013e31828b327e [DOI] [PubMed] [Google Scholar]
- 48.Chapman AR, Hesse K, Andrews J, et al. High-sensitivity cardiac troponin I and clinical risk scores in patients with suspected acute coronary syndrome. Circulation. 2018;138(16):1654-1665. doi: 10.1161/CIRCULATIONAHA.118.036426 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 49.Mokhtari A, Borna C, Gilje P, et al. A 1-h combination algorithm allows fast rule-out and rule-in of major adverse cardiac events. J Am Coll Cardiol. 2016;67(13):1531-1540. doi: 10.1016/j.jacc.2016.01.059 [DOI] [PubMed] [Google Scholar]
- 50.Jaeger C, Wildi K, Twerenbold R, et al. One-hour rule-in and rule-out of acute myocardial infarction using high-sensitivity cardiac troponin I. Am Heart J. 2016;171(1):92-102.e1-5. doi: 10.1016/j.ahj.2015.07.022 [DOI] [PubMed] [Google Scholar]
- 51.deFilippi C, Seliger SL, Kelley W, et al. Interpreting cardiac troponin results from high-sensitivity assays in chronic kidney disease without acute coronary syndrome. Clin Chem. 2012;58(9):1342-1351. doi: 10.1373/clinchem.2012.185322 [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
eTable 1. Contemporary troponin analytical data by clinical site
eTable 2. Adjusted odds ratios (aOR) for safety events among ESC 0/1-hour hs-cTnT algorithm patients with known CAD vs. no known CAD
eTable 3. Safety events when using the ESC 0/1-hour hs-cTnT algorithm combined with the HEART score (rule-out ≤ 3, observation 4-6, rule-in ≥ 7)
eTable 4. Test characteristics of the ESC 0/1-hour hs-cTnT algorithm combined with the HEART score (rule-out ≤ 3, observation 4-6, rule-in ≥7)
eTable 5. Subgroup analysis for early presenters (≤ 3 hours of chest pain until ED evaluation)
eTable 6. Subgroup analysis for late presenters (> 3 hours of chest pain until ED evaluation)
eFigure. Study flow diagram
Data sharing statement