Abstract
Background: Patients with atrial fibrillation (AF) often present with symptoms similar to acute coronary syndrome (ACS), including chest pain and elevated levels of high-sensitivity cardiac troponin (hs-cTn). The 0/1-hour algorithm using hs-cTn is a rapid diagnostic tool endorsed by the European Society of Cardiology to rule out myocardial infarction (MI). However, because its effectiveness in patients with AF remains unclear, in this study we assessed the diagnostic accuracy of the 0/1-hour algorithm in patients with and without AF presenting with chest pain in the emergency department.
Methods and Results: We conducted a secondary analysis of the DROP-ACS cohort, including 1,333 patients from Japan and Taiwan, with AF in 10.3% of cases. We examined the algorithm’s negative predictive value (NPV), sensitivity, positive predictive value (PPV), and specificity for ruling MI in or out. Patients with AF were more frequently placed in the observe group (54% vs. 34.9%, P<0.05) and less often in the rule-out group (24.1% vs. 44.6%, P<0.05). The NPV and sensitivity for ruling out MI were 100%, while the PPV and specificity were lower in patients with AF (60% and 89.7%, respectively).
Conclusions: The 0/1-hour algorithm effectively ruled out MI in patients with AF, with high safety and accuracy. However, patients with AF are more likely to be stratified into the observe group, requiring further examination for final diagnosis.
Patients with symptoms suggestive of acute myocardial infarction (AMI) account for approximately 10% of emergency department (ED) visits. Current guidelines recommend measuring serial cardiac troponin (cTn) levels to safely rule out AMI in the ED.1 Atrial fibrillation (AF), the most common arrhythmia, is more prevalent in patients with atherosclerosis.2,3 AF can present with symptoms similar to those of acute coronary syndrome (ACS), such as acute chest pain, dyspnea, electrocardiographic (ECG) changes indicative of ischemia, and elevated cTn levels, even in the absence of acute or chronic coronary syndrome.4–6 It is well-established that the presence of AF in patients with ACS results in higher rates of in-hospital complications, rehospitalizations, and deaths, as well as poorer short- and long-term outcomes.7–9 However, the underlying cause of elevated cTn levels in patients with AF, whether caused by ACS or other mechanisms, remains unclear. Therefore, accurately diagnosing ACS in patients with AF is crucial. In 2015, the European Society of Cardiology (ESC) recommended early risk stratification of AMI using the 0/1-hour algorithm with high-sensitivity cTn (hs-cTn) as a Class I, Level of Evidence B guideline.10 This algorithm was designed to effectively rule out ACS early.11 However, evidence regarding the effectiveness of the 0/1-hour algorithm in patients with AF is limited, so in this study, we investigated the effect of AF on the utility of the 0/1-hour algorithm in patients presenting to the ED with chest pain. We specifically examined how the presence or absence of AF affects the diagnostic accuracy for AMI and subsequent death.
Methods
Study Design and Patients
This study was a secondary analysis of a cohort study, the “Diagnostics and Reduction of Asian Patients with Acute Coronary Syndrome Cost Analysis Based on the 0/1-Hour Algorithm Using High-sensitivity Cardiac Troponin (DROP-ACS) (UMIN000030668).” It is a prospective international multicenter diagnostic study that has been conducted at 5 sites in 2 countries (Japan and Taiwan) since November 2014.12,13 The inclusion criteria were adults (aged 30–89 years) who presented with chest pain related to a suspected cardiac cause. The 0/1-hour algorithm was implemented at the discretion of the attending physicians. The exclusion criteria included ST-elevation MI (STEMI), chronic kidney disease (e.g., serum creatinine >3 mg/dL), congestive heart failure (i.e., the presence of hypoxia and typical pulmonary congestion, as confirmed with chest radiography), ventricular tachycardia, and arrival in the ED >24 h after symptom onset. Informed consent was given by all patients. The study was approved by the ethics committees of the participating hospitals and was conducted in accordance with the Declaration of Helsinki of 1975, as revised in 1983.
Clinical Evaluation
All study patients underwent an initial clinical assessment that included physical examination, medical history, smoking status, 12-lead ECG, blood pressure measurement, heart rate, and standard blood tests. The Emergency Department Assessment of Chest Pain Score (EDACS) was calculated from prospectively collected data, without interpretation by the investigators. The History, Electrocardiogram, Age, Risk Factors, and Troponin (HEART) score, history, and ECG components were retrospectively calculated by cardiologists, who were blinded to the blood test results, including hs-cTnT, as well as patient outcomes.
ESC 0/1-Hour Algorithm
The 0/1-hour algorithm uses blood concentrations of hs-cTnT (5th generation; Roche Diagnostics, Basel, Switzerland) at admission and their absolute changes within 1 h to classify patients into the “rule-out,” “observe,” or “rule-in” group.14 Patients with hs-cTnT concentration <12 ng/L and a delta at 1 h of <3 ng/L were stratified into the rule-out group; patients with hs-cTnT levels of ≥52 ng/L or a delta at 1 h of ≥5 ng/L were stratified into the rule-in group; the remaining patients were assigned to the observe group. We did not exclude patients solely based on the concentrations obtained at 0 h, even if the troponin concentration was <5 ng/L.
Determination of Final Diagnosis
All attending physicians made a tentative diagnosis, based on the algorithm, and further confirmed a diagnosis at the 30-day follow-up via cardiac stress ECG, stress scintigraphy, coronary computed tomography angiography (CTA), or coronary angiography (CAG). Two independent cardiologists reviewed all of the available medical records. AMI was defined, based on the Fourth Universal Definition of Myocardial Infarction,15 which requires evidence of myocardial necrosis in addition to ischemia (as shown by ECG changes and positive findings on stress ECG, myocardial perfusion scan, coronary CTA, or CAG during catheterization). Overall, necrosis was diagnosed, based on an increase or decrease (20% relative change) in hs-cTnT concentrations, with at least 1 measurement above the 99th percentile of the normal reference range at an assay imprecision level of approximately 10%. Type 1 MI is characterized by atherosclerotic plaque rupture, ulceration, erosion, or dissection, leading to intraluminal thrombus formation in ≥1 coronary arteries, which results in myocardial necrosis (hs-cTnT concentration >14 ng/L, with an increase or decrease in hs-cTnT concentrations when serial testing is available). Type 2 MI is defined as myocardial necrosis in which a condition other than coronary plaque instability contributes to an imbalance between myocardial oxygen supply and demand (e.g., coronary artery spasm, coronary embolism, hypertension, or hypotension). All other patients were classified as not having AMI. Myocardial injury is defined by an elevated cTn value above the 99th percentile of the upper reference limit, with a rise and/or fall in cTn values indicating acute myocardial injury. However, myocardial injury encompasses both MI and other forms of myocardial damage. To specifically differentiate MI from myocardial injury in this study, we defined myocardial injury as follows: (1) elevated cTn level exceeding the 99th percentile at least once, (2) observable rise and/or fall in cTn levels and (3) no cardiovascular death or MI (including the index visit) occurring within 30 days of follow-up. Unstable angina was diagnosed in patients with ischemic symptoms at rest or with minor exertion, but with no evidence of necrosis; cardiac catheterization revealed the presence of a thrombus in the culprit lesion of the coronary artery. Vasospastic angina was diagnosed in patients with no obvious coronary stenosis, in whom the ECG findings revealed ST-elevation, or who had angiographic evidence of coronary artery spasm, which was relieved after intracoronary administration of 1 mg of nitroglycerin. Noncardiac chest pain was diagnosed when no significant abnormal findings were observed on ECG, blood examination, or chest radiography at the index visit.
Endpoints
The primary outcome measures were the negative predictive value (NPV) and sensitivity of the 0/1-hour algorithm for ruling out AMI. The secondary outcome measures included the positive predictive value (PPV) and specificity of the 0/1-hour algorithm for ruling in AMI, as well as the proportion of patients assigned to the observe group. During follow-up, the patients were monitored by a cardiologist at the same hospital. In the absence of a cardiologist, patients were followed up using hospital records and telephone interviews to determine adverse events 30 days after the initial visit.
Statistical Analysis
Continuous variables are expressed as the mean (±standard deviation) or as the median (interquartile range [IQR]). Categorical variables are expressed as a number and percentages. Differences were statistically significant at P<0.05. Differences in baseline characteristics between patients with and without AF were assessed using the Mann-Whitney U test for continuous variables and Fisher’s exact test for categorical variables. A one-way analysis of variance was used to compare the 3 groups. Statistical analyses were conducted using SPSS (version 16.0; IBM Corp., Armonk, NY, USA), R version 3.5.0 (R Foundation for Statistical Computing, Vienna, Austria), and JMP version 9.0.0 (SAS Institute Inc., Cary, NC, USA). One author (K.I.) had access to all data in this study and takes full responsibility for its completeness and analysis.
Results
Patients’ Characteristics
A total of 1,333 individuals were eligible for the analysis between November 2014 and February 2022. The baseline characteristics are shown in Table 1. The median age was 70 years, men constituted 59.1% of the patients, and the prevalence of AF at presentation was 10.3%. Paroxysmal AF accounted for 57.6%, and persistent AF comprised 42.3%. The median IQR of the HEART score was 4 (3–5), and the median IQR of the EDACS score was 16 (12–20). The final adjudicated diagnoses were type 1 MI in 142 (10.7%) patients, type 2 MI in 48 (3.6%) patients, unstable angina in 74 (5.6%) patients, and vasospastic angina in 23 (1.7%) patients. Of the 1,333 patients, 10.3% (n=137) had AF and tended to have higher hs-cTnT levels and NT-pro-B-type natriuretic peptide levels, higher pulse rate, lower low-density lipoprotein cholesterol levels, and lower estimated glomerular filtration rate than patients without AF. Oral anticoagulants were used among 53.3% of patients with AF. The hs-cTnT levels were higher in patients with AF compared with those without AF in the rule-out and observe groups (P<0.05). However, delta hs-cTnT showed no significant difference (P=0.69 in the rule-out group, P=0.36 in the observe group). In the rule-in group, hs-cTnT levels were higher in patients without AF, but the differences were not statistically significant (P=0.41 for the initial, P=0.11 for the second measurement). Delta hs-cTnT did not show significant differences either (P=0.23) (Table 2).
Table 1.
Baseline Characteristics of the Patients
Characteristic
(N=1,333) |
All patients
(N=1,333) |
AF
(n=137; 10.3%) |
No AF
(n=1,196; 89.7%) |
P value |
| Age, years, median (IQR) |
70 (58–80) |
76 (71–83) |
69 (57–79) |
<0.05 |
| Male sex, n (%) |
788 (59.1) |
85 (62.0) |
703 (58.7) |
0.46 |
| BMI, kg/m2, median (IQR) |
23.2 (21.0–25.7) |
22.6 (20.3–25.4) |
23.2 (21.1–25.9) |
0.13 |
| Risk factors, n (%) |
| Hypertension |
805 (60) |
90 (66) |
715 (60) |
0.19 |
| Dyslipidemia |
961 (72) |
88 (64) |
873 (73) |
<0.05 |
| Diabetes mellitus |
372 (28) |
38 (28) |
334 (28) |
0.90 |
| History of smoking |
666 (50) |
61 (45) |
605 (51) |
0.18 |
| History, n (%) |
| Coronary artery disease |
350 (26) |
56 (41) |
294 (25) |
<0.05 |
| Cerebral vascular disorder |
136 (10) |
22 (16) |
114 (10) |
<0.05 |
| Laboratory data, median (IQR) |
| Initial hs-cTnT, ng/L |
13 (6–24) |
17 (11–29) |
13 (6–23) |
<0.05 |
| LDL-C, mg/dL |
105 (84–132) |
98 (73–124) |
107 (85–133) |
<0.05 |
| Hemoglobin A1c, % |
5.9 (5.6–6.4) |
6.0 (5.7–6.4) |
5.9 (5.6–6.4) |
0.18 |
| eGFR, mL/min/1.73 m2 |
69 (57–81) |
61 (48–71) |
70 (58–82) |
<0.05 |
| NT-proBNP, pg/mL |
117 (40–373) |
540 (168–1,441) |
101 (34–292) |
<0.05 |
| SBP, mmHg |
147 (130–166) |
146 (126–161) |
147 (130–166) |
0.15 |
| DBP, mmHg |
84 (75–96) |
83 (75–95) |
84 (74–96) |
0.97 |
| Pulse rate, beats/min |
78 (68–90) |
84 (71–104) |
78 (68–89) |
<0.05 |
| Score, median (IQR) |
| HEART score |
4 (3–5) |
4 (3–5) |
4 (3–5) |
<0.05 |
| EDACS score |
16 (12–20) |
20 (16–22) |
16 (12–20) |
<0.05 |
| Medications, n (%) |
| Anticoagulant |
N/A |
73 (53.3) |
N/A |
– |
| Warfarin |
N/A |
19 (13.9) |
N/A |
– |
| DOAC |
N/A |
54 (39.4) |
N/A |
– |
| Dabigatran |
N/A |
9 (12.3) |
N/A |
– |
| Edoxaban |
N/A |
11 (15.1) |
N/A |
– |
| Rivaroxaban |
N/A |
15 (20.5) |
N/A |
– |
| Apixaban |
N/A |
19 (26.0) |
N/A |
– |
| Adjudicated final diagnosis, n (%) |
| Myocardial injury |
75 (5.6) |
12 (8.8) |
63 (5.3) |
0.09 |
| Non-cardiac chest pain |
959 (71.9) |
99 (72.3) |
860 (71.9) |
0.93 |
| ACS |
264 (19.8) |
23 (16.8) |
241 (20.2) |
0.35 |
| MI |
190 (14.3) |
20 (14.6) |
170 (14.2) |
0.90 |
| Type 1 MI |
142 (10.7) |
6 (4.4) |
136 (11.4) |
<0.05 |
| Type 2 MI |
48 (3.6) |
14 (10.2) |
34 (2.8) |
<0.05 |
| Unstable angina |
74 (5.6) |
3 (2.2) |
71 (5.9) |
0.13 |
| Takotsubo cardiomyopathy |
12 (0.9) |
2 (1.5) |
10 (0.8) |
0.46 |
| Vasospastic angina |
23 (1.7) |
1 (0.7) |
22 (1.8) |
0.34 |
ACS, acute coronary syndrome; AF, atrial fibrillation; BMI, body mass index; DBP, diastolic blood pressure; DOAC, direct oral anticoagulant; eGFR, estimated glomerular filtration rate; hs-cTnT, high-sensitivity cardiac troponin T; IQR, interquartile range; LDL-C, low-density lipoprotein cholesterol; MI, myocardial infarction; NT-proBNP, N-terminal pro-B-type natriuretic peptide; SBP, systolic blood pressure; SD, standard deviation.
Table 2.
hs-cTnT Levels of Patients in Each Group of the 0/1-Hour Algorithm
hs-cTnT, ng/L,
median (IQR) |
Rule-out |
Observe |
Rule-in |
| AF |
No AF |
AF |
No AF |
AF |
No AF |
| Initial |
8 (6.5–9.5) |
6 (5–8) |
19 (15–28) |
17 (13–25) |
44 (16–89) |
53 (19–162) |
| P value |
<0.05 |
0.06 |
0.41 |
| Second |
8 (6, 10) |
6 (5–8) |
19 (16–30) |
18 (13–25) |
57 (29–91) |
74 (40–228) |
| P value |
<0.05 |
<0.05 |
0.11 |
| Delta |
0 (0–1) |
0 (0–0) |
0 (−1 to 1) |
0 (−1 to 1) |
9 (5–20) |
13 (5–45) |
| P value |
0.69 |
0.36 |
0.23 |
AF, atrial fibrillation; hs-cTnT, high-sensitivity cardiac troponin T; IQR, interquartile range.
Distribution Pattern of the 0/1-Hour Algorithm
The Figure shows the distribution patterns of patients with and without AF, based on the results of the 0/1-hour algorithm. The number of patients with and without AF in the rule-in group was not significantly different (21.9% vs. 20.5%, P=0.7). A significantly higher proportion of patients with AF was classified into the observe group (54.0% vs. 34.9%; P<0.05), whereas a significantly lower proportion was classified into the rule-out group than among patients without AF (24.1% vs. 44.6%; P<0.05). Among patients classified in the rule-out group, none were diagnosed with MI (NPV and sensitivity for AMI were 100% [95% confidence interval (CI), 84.7–100.0%] and 100% [95% CI, 76.2–100.0%], respectively, for patients with AF and 100% [95% CI, 98.7–100.0%] and 100% [95% CI, 96.8–100.0%], respectively, for patients without AF). Among patients in the rule-in group, 18 of 30 patients with AF were diagnosed with AMI, compared with 168 of 245 patients without AF (PPV and specificity were 60.0% [95% CI, 40.6–77.3%] and 89.7% [95% CI, 82.8–94.6%], respectively, for patients with AF, and 68.6% [95% CI, 62.4–74.3%] and 92.5% [95% CI, 90.7–94.0%], respectively, for patients without AF) (Table 3). Interestingly, none of the patients had a HEART score MI rating below 4 (Supplementary Table). As highlighted in the 2022 ACC Expert Consensus Decision Pathway on the Evaluation and Disposition of Acute Chest Pain in the Emergency Department,16 the integration of a low-risk HEART score into the 0/1-hour algorithm may be particularly useful for identifying AF patients with ACS. Patients with AF, compared with patients without AF, had significantly fewer type 1 MI (20.0% vs. 54.7%, P<0.05) and significantly more type 2 MI (40.0% vs. 13.9%, P<0.05) (Table 4). Among patients with paroxysmal AF, 45.6% were categorized into the observe group, whereas 65.5% of those with persistent AF fell into this category (Supplementary Figure).
Table 3.
2×2 Tables and Calculation of Negative and Positive Predictive Values, as Well as Sensitivity and Specificity for the Rule-Out and Rule-in of Myocardial Infarction
| A. Algorithm classification vs. adjudicated diagnosis |
| Algorithm classification |
AF |
No AF |
| AMI |
Non-AMI |
Total |
AMI |
Non-AMI |
Total |
| Rule-out |
0 |
33 |
33 |
0 |
533 |
533 |
| Observe |
2 |
72 |
74 |
2 |
416 |
418 |
| Rule-in |
18 |
12 |
30 |
168 |
77 |
245 |
| Total |
20 |
117 |
137 |
170 |
1,026 |
1,196 |
| B. Negative and positive predictive value |
| Diagnostic test performance measures |
AF |
No AF |
| Estimate, % |
95% CI |
Counts |
Estimate, % |
95% CI |
Count |
| NPV |
100.0 |
84.7–100 |
33/33 |
100.0 |
98.7–100 |
533/533 |
| PPV |
60.0 |
40.6–77.3 |
18/30 |
68.6 |
62.4–74.3 |
168/245 |
| C. Sensitivity and specificity |
| Diagnostic test performance measures |
AF |
No AF |
| Estimate, % |
95% CI |
Counts |
Estimate, % |
95% CI |
Count |
| Sensitivity in the rule-out |
100.0 |
76.2–100 |
20/20 |
100.0 |
96.8–100 |
170/170 |
| Specificity in the rule-in |
89.7 |
82.8–94.6 |
105/117 |
92.5 |
90.7–94.0 |
949/1,026 |
*Sensitivity: true positive/diseased (AMI). The rule-out group defines patients with no AMI, based on the 0/1-hour algorithm. Only patients in this group are ruled out. Thus, for the rule-out, whether patients are in the observe group or the rule-in group is irrelevant, and both groups are combined. Specificity: true negative/nondiseased (non-AMI). The rule-in group defines patients with AMI, based on the 0/1-hour algorithm. Only patients in this group are ruled in. Thus, for the rule-in, whether patients are in the observe group or the rule-out group is irrelevant, and both groups are combined. AF, atrial fibrillation; AMI, acute myocardial infarction; CI, confidence interval; NPV, negative predictive value; PPV, positive predictive value.
Table 4.
Ratio of Cardiovascular Disease in the 0/1-Hour Algorithm
| Adjudicated final diagnosis, n (%) |
AF |
No AF |
P value |
| 0/1-hour algorithm “rule-out” |
33 (24.1) |
533 (44.6) |
<0.05 |
| Myocardial injury |
0 (0.0) |
0 (0) |
– |
| Noncardiac chest pain |
32 (97.0) |
498 (93.4) |
0.42 |
| ACS |
0 (0.0) |
25 (4.7) |
0.20 |
| MI |
0 (0.0) |
0 (0.0) |
– |
| Type 1 MI |
0 (0.0) |
0 (0.0) |
– |
| Type 2 MI |
0 (0.0) |
0 (0.0) |
– |
| Unstable angina |
0 (0.0) |
25 (4.7) |
0.20 |
| Takotsubo cardiomyopathy |
0 (0.0) |
0 (0.0) |
– |
| Vasospastic angina |
1 (3.0) |
10 (1.9) |
0.64 |
| 0/1-hour algorithm “observe” |
74 (54.0) |
418 (34.9) |
<0.05 |
| Myocardial injury |
3 (4.1) |
4 (1.0) |
<0.05 |
| Noncardiac chest pain |
66 (89.2) |
357 (85.4) |
0.39 |
| ACS |
5 (6.8) |
44 (10.5) |
0.32 |
| MI |
2 (2.7) |
2 (0.5) |
<0.05 |
| Type 1 MI |
0 (0.0) |
2 (0.5) |
0.55 |
| Type 2 MI |
2 (2.7) |
0 (0.0) |
<0.05 |
| Unstable angina |
3 (4.1) |
42 (10.0) |
0.10 |
| Takotsubo cardiomyopathy |
0 (0.0) |
3 (0.7) |
0.46 |
| Vasospastic angina |
0 (0.0) |
10 (2.4) |
0.18 |
| 0/1-hour algorithm “rule-in” |
30 (21.9) |
245 (20.5) |
0.70 |
| Myocardial injury |
9 (30.0) |
59 (24.1) |
0.48 |
| Noncardiac chest pain |
1 (3.3) |
5 (2.0) |
0.65 |
| ACS |
18 (60.0) |
172 (70.2) |
0.25 |
| MI |
18 (60.0) |
168 (68.6) |
0.34 |
| Type 1 MI |
6 (20.0) |
134 (54.7) |
<0.05 |
| Type 2 MI |
12 (40.0) |
34 (13.9) |
<0.05 |
| Unstable angina |
0 (0.0) |
4 (1.6) |
0.77 |
| Takotsubo cardiomyopathy |
2 (6.7) |
7 (2.9) |
0.27 |
| Vasospastic angina |
0 (0.0) |
2 (0.8) |
0.62 |
ACS, acute coronary syndrome; AF, atrial fibrillation; MI, myocardial infarction.
Endpoints
The cumulative 30-day mortality rates for patients without AF were 0.4% (n=1) in the rule-in group, 1.0% (n=4) in the observe group, and 0.2% (n=1) in the rule-out group. However, no events occurred within 30 days in patients with AF (Table 5). Among patients with AF classified as “observe”, but ultimately diagnosed with noncardiac chest pain, 15 cases were associated with inflammatory conditions, including infections, whereas the remaining 51 cases showed no recurrence of symptoms during the 30-day follow-up.
Table 5.
Major Adverse Cardiovascular Events and Deaths
| 30-Day outcome, n (%) |
AF |
No AF |
| 0/1-hour algorithm “rule-out” |
33 |
533 |
| CVA |
0 (0.0) |
0 (0.0) |
| CV death |
0 (0.0) |
0 (0.0) |
| Non-CV death |
0 (0.0) |
1 (0.2) |
| No event |
33 (100.0) |
532 (99.8) |
| 0/1-hour algorithm “observe” |
74 |
418 |
| CVA |
0 (0.0) |
1 (0.2) |
| CV death |
0 (0.0) |
1 (0.2) |
| Non-CV death |
0 (0.0) |
3 (0.7) |
| No event |
74 (100.0) |
413 (98.8) |
| 0/1-hour algorithm “rule-in” |
30 |
245 |
| CVA |
0 (0.0) |
0 (0.0) |
| CV death |
0 (0.0) |
1 (0.4) |
| Non-CV death |
0 (0.0) |
0 (0.0) |
| No event |
30 (100.0) |
244 (99.6) |
AF, atrial fibrillation; CV, cardiovascular; CVA, cerebrovascular accident.
Discussion
In this study, a subanalysis of a multicenter international study was conducted to address recent concerns regarding the management of patients with AF by using the 0/1-hour algorithm. We report several key findings from the current study. First, the safety of triage for non-STEMI (NSTEMI) rule-out was 100% for NPV [95% CI (84.7–100%)] and sensitivity [95% CI (76.2–100%)], even in the presence of AF. Given the HEART or EDACS risk scores, this cohort was rated as an intermediate-risk group, although the algorithm worked safely. Second, the PPV for the NSTEMI rule-in in patients with AF was 60% [95% CI (40.6–77.3%)] and was lower in patients with AF than in patients without AF. Although the PPV and specificity were slightly lower in patients with AF, their 95% CIs overlapped widely, suggesting an acceptable performance in clinical practice. Third, the percentage of patients triaged in the observe group was substantially higher among patients with AF than among patients without AF. This can be attributed to the known phenomenon of troponin elevation in AF, even in the absence of myocardial necrosis. Although hs-cTn assays are indispensable for diagnosing MI, as demonstrated by animal studies showing that troponin T and I release correlate with necrotic tissue damage,17 Redfearn et al. reported that patients with supraventricular tachycardia, who did not have cardiac necrosis, showed rapid troponin elevations.18 This can be explained by the composition of troponin, which includes an unbound pool of troponin I and T in the cytosol (5–8%) that is quickly released during myocyte stress.19 This unbound pool is released rapidly and cleared from the blood with a half-life of approximately 2 h, indicating reversible myocyte damage rather than necrosis.20,21 Studies, such as those by Turer et al., have shown that reversible myocyte damage due to rapid atrial pacing can elevate hs-cTnT levels in the coronary sinus and peripheral blood without indicating necrosis.22 These findings provide a plausible explanation for the higher proportion of patients with AF in the observe group. Fourth, the 0/1-hour algorithm enabled powerful and reliable risk stratification of short-term deaths and major adverse cardiac events. By leveraging Bayes’ theorem and hs-cTn assays, the algorithm effectively identifies patients with low-risk and accelerates clinical decision-making, reducing unnecessary ED stays and associated costs. Its application in patients with AF supports its utility as a robust framework for risk assessment. Despite these strengths, the high proportion of patients with AF classified in the observe group underscores the need for additional stratification tools. Elevated hs-cTn levels in AF often reflect myocardial stress or type 2 MI rather than type 1 MI, complicating the use of biomarker-based algorithms. The observe group, however, remains a recognized challenge in acute chest pain management, irrespective of the algorithm used. Strategies such as incorporating coronary CTA, stress myocardial scintigraphy, or additional hs-cTn testing at 3 h have been explored as potential solutions.23 Notably, Lopez-Ayala et al. proposed a modified approach that set new cutoff values for hs-cTn testing at 3 h, aiming for a negative NPV >99% for rule-out, specificity >95%, and PPV >75%.24 Despite these efforts, >60% of patients remained in the observe group, thereby highlighting the persistent difficulty in effectively reclassifying patients. This underscores the ongoing challenge of resolving the issue, even with advanced diagnostic strategies. Given this context, we believe that the observe group represents an unresolved challenge that is not unique to the 0/1-hour algorithm or patients with AF. Instead, it reflects a broader limitation of current diagnostic strategies for NSTEMI. Future studies focusing on innovative modalities or algorithm refinements may help address this issue.
Study Limitations
First, the sample size was small; however, it included data from 5 hospitals across 2 countries, thereby supporting its universality. Second, the increase in the observe group reduced the proportion of patients discharged directly from the ED, raising concerns about prolonged ED stays. Third, patients on dialysis were excluded, thus limiting the generalizability of the findings. Fourth, the exclusion criterion based on undetectable hs-cTnT (<5 ng/L) at 0 h was not applied because the lowest measurable concentration of hs-cTnT at the participating facilities was recorded as 5 ng/L, even when actual values fell below this threshold.
Conclusions
The 0/1-hour algorithm safely stratified the patients, including patients with AF. However, challenges such as increased ED congestion and hospital admissions due to the expanded observe group warrant further investigation in future studies.
Acknowledgments
This work was reported at the 88th
Annual Scientific Meeting of the Japanese Circulation Society, held on March 8–10, 2024.
Disclosures
K.I. received a Grant-in-Aid for Scientific Research C (No. 18K09954) and research grants from Roche Diagnostics, SB Bioscience Co., Ltd., Fujirebio Inc., Bayer Sysmex, and Kanto Chemical Co. Inc. T.M., PhD received a research grant from Roche Diagnostics and is a member of Circulation Journal’s Editorial Team. The remaining authors have no conflicts of interest to declare.
IRB Information
This study was approved by the Ethical Review Board of Juntendo University Nerima Hospital (Reference no. 17–16).
Data Availability
All deidentified participant data, ending 10 years after the publication of this article, will be shared upon request for any type of analysis. The data will be provided as Excel files via email. Please contact the corresponding author directly to request data sharing.
Supplementary Files
Please find supplementary file(s);
https://doi.org/10.1253/circj.CJ-24-0811
References
- 1.
Kikuchi M, Tahara Y, Yamaguchi J, Nakashima T, Nomura O, Tanaka A, et al. Executive Summary: Acute Coronary Syndrome in the Japan Resuscitation Council Guidelines for Resuscitation 2020. Circ J 2023; 87: 866–878.
- 2.
Abolbashari M. Atherosclerosis and atrial fibrillation: Double trouble. Curr Cardiol Rep 2022; 24: 67–73.
- 3.
da Silva RM. Influence of inflammation and atherosclerosis in atrial fibrillation. Curr Atheroscler Rep 2017; 19: 2.
- 4.
Liebetrau C, Weber M, Tzikas S, Palapies L, Mollmann H, Pioro G, et al. Identification of acute myocardial infarction in patients with atrial fibrillation and chest pain with a contemporary sensitive troponin I assay. BMC Med 2015; 13: 169.
- 5.
Bukkapatnam RN, Robinson M, Turnipseed S, Tancredi D, Amsterdam E, Srivatsa UN. Relationship of myocardial ischemia and injury to coronary artery disease in patients with supraventricular tachycardia. Am J Cardiol 2010; 106: 374–377.
- 6.
Kochiadakis GE, Skalidis EI, Kalebubas MD, Igoumenidis NE, Chrysostomakis SI, Kanoupakis EM, et al. Effect of acute atrial fibrillation on phasic coronary blood flow pattern and flow reserve in humans. Eur Heart J 2002; 23: 734–741.
- 7.
Jabre P, Jouven X, Adnet F, Thabut G, Bielinski SJ, Weston SA, et al. Atrial fibrillation and death after myocardial infarction: A community study. Circulation 2011; 123: 2094–2100.
- 8.
Kundu A, O’Day K, Shaikh AY, Lessard DM, Saczynski JS, Yarzebski J, et al. Relation of atrial fibrillation in acute myocardial infarction to in-hospital complications and early hospital readmission. Am J Cardiol 2016; 117: 1213–1218.
- 9.
Masunaga N, Ogawa H, Minami K, Ishigami K, Ikeda S, Doi K, et al. Association of concomitant coronary artery disease with cardiovascular events in patients with atrial fibrillation: The Fushimi AF Registry. Circ J 2022; 86: 1252–1262.
- 10.
Roffi M, Patrono C, Collet JP, Mueller C, Valgimigli M, Andreotti F, et al. 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: 267–315.
- 11.
Nomura O, Hashiba K, Kikuchi M, Kojima S, Hanada H, Mano T, et al. Performance of the 0-hour/1-hour algorithm for diagnosing myocardial infarction in patients with chest pain in the emergency department: A systematic review and meta-analysis. Circ Rep 2022; 4: 241–247.
- 12.
Shiozaki M, Inoue K, Suwa S, Lee CC, Chikata Y, Ishiura J, et al. Utility of the 0-hour/1-hour high-sensitivity cardiac troponin T algorithm in Asian patients with suspected non-ST elevation myocardial infarction. Int J Cardiol 2017; 249: 32–35.
- 13.
Shiozaki M, Inoue K, Suwa S, Lee CC, Chiang SJ, Sato A, et al. Implementing the European Society of Cardiology 0-h/1-h algorithm in patients presenting very early after chest pain. Int J Cardiol 2020; 320: 1–6.
- 14.
Reichlin T, Schindler C, Drexler B, Twerenbold R, Reiter M, Zellweger C, et al. One-hour rule-out and rule-in of acute myocardial infarction using high-sensitivity cardiac troponin T. Arch Intern Med 2012; 172: 1211–1218.
- 15.
Thygesen K, Alpert JS, Jaffe AS, Chaitman BR, Bax JJ, Morrow DA, et al. Fourth universal definition of myocardial infarction (2018). Circulation 2018; 138: e618–e651.
- 16.
Kontos MC, de Lemos JA, Deitelzweig SB, Diercks DB, Gore MO, Hess EP, et al. 2022 ACC Expert Consensus Decision Pathway on the Evaluation and Disposition of Acute Chest Pain in the Emergency Department: A report of the American College of Cardiology Solution Set Oversight Committee. J Am Coll Cardiol 2022; 80: 1925–1960.
- 17.
Fishbein MC, Wang T, Matijasevic M, Hong L, Apple FS. Myocardial tissue troponins T and I: An immunohistochemical study in experimental models of myocardial ischemia. Cardiovasc Pathol 2003; 12: 65–71.
- 18.
Redfearn DP, Ratib K, Marshall HJ, Griffith MJ. Supraventricular tachycardia promotes release of troponin I in patients with normal coronary arteries. Int J Cardiol 2005; 102: 521–522.
- 19.
Katus HA, Remppis A, Scheffold T, Diederich KW, Kuebler W. Intracellular compartmentation of cardiac troponin T and its release kinetics in patients with reperfused and nonreperfused myocardial infarction. Am J Cardiol 1991; 67: 1360–1367.
- 20.
White HD. Pathobiology of troponin elevations: Do elevations occur with myocardial ischemia as well as necrosis? J Am Coll Cardiol 2011; 57: 2406–2408.
- 21.
Costabel JP, Burgos LM, Trivi M. The significance of troponin elevation in atrial fibrillation. J Atr Fibrillation 2017; 9: 1530.
- 22.
Turer AT, Addo TA, Martin JL, Sabatine MS, Lewis GD, Gerszten RE, et al. Myocardial ischemia induced by rapid atrial pacing causes troponin T release detectable by a highly sensitive assay: Insights from a coronary sinus sampling study. J Am Coll Cardiol 2011; 57: 2398–2405.
- 23.
Nestelberger T, Wildi K, Boeddinghaus J, Twerenbold R, Reichlin T, Gimenez MR, et al. Characterization of the observe zone of the ESC 2015 high-sensitivity cardiac troponin 0 h/1 h-algorithm for the early diagnosis of acute myocardial infarction. Int J Cardiol 2016; 207: 238–245.
- 24.
Lopez-Ayala P, Nestelberger T, Boeddinghaus J, Koechlin L, Ratmann PD, Strebel I, et al. Novel criteria for the observe-zone of the ESC 0/1 h-hs-cTnT algorithm. Circulation 2021; 144: 773–787.
