Emergency Department Cardiac Risk Stratification With High-Sensitivity vs Conventional Troponin HEART Pathway

Key Points Question Are 30-day health outcomes and health service use different among emergency department (ED) patients undergoing cardiac risk stratification with a HEART pathway using conventional vs high-sensitivity troponin level? Findings In a multicenter cohort study of 17 384 ED patients with chest pain risk stratified with a HEART pathway incorporating conventional vs high-sensitivity troponin level, the high-sensitivity troponin group showed a higher rate of acute myocardial infarction (AMI) diagnosis during the ED visit and a lower rate of AMI diagnosis within 30 days after the visit, along with lower rates of admission, stress testing, and revascularization. Mortality was rare and similar across groups. Meaning This study suggests that a high-sensitivity troponin HEART pathway may be associated with earlier detection of AMI and improved resource use.


Introduction
Chest pain, a classic symptom of acute myocardial infarction (AMI), is the second most common reason for emergency department (ED) visits in the US. 1 Differentiating potentially life-threatening causes from other causes of chest pain is a common, high-stakes task.Historically, failure to detect AMI in the ED has been associated with increased mortality risk. 2 Consequently, clinicians frequently overuse advanced testing for patients with chest pain, resulting in unnecessary hospitalizations, billions of dollars in avoidable annual costs, and iatrogenic harm without clinical benefits. 3,4ergency departments in the US commonly use the HEART (History, Electrocardiogram, Age, Risk factors, and Troponin) score for cardiac risk stratification, combining patient history, risk factors, electrocardiography findings, and serum troponin. 5Some hospitals formally incorporate HEART scores into diagnostic and management pathways for patients presenting with chest pain.In 2017, the US Food and Drug Administration approved a higher-sensitivity troponin (hsTn) assay capable of detecting cardiac-specific serum troponin at lower levels than conventional troponin (cTn), 6 potentially enabling earlier rule-in or rule-out of AMI.Previous studies on the association of hsTn with patient care are mostly from outside the US and present equivocal results on clinical outcomes and service use [7][8][9][10][11][12] ; to our knowledge, little is known about the association of switching to an hsTn-based HEART pathway with patient outcomes and service use in clinical practice in the US.We conducted the first large, US multicenter study comparing health outcomes and care use among adults presenting to the ED with potential AMI before and after implementation of a new hsTn-based HEART pathway.

Study Design and Participants
Kaiser Permanente Southern California (KPSC) serves 4.6 million Kaiser members across 21 hospitals; KPSC data include outcomes and service use at KPSC and non-KPSC facilities and deaths anywhere.
In 2021, a total of 16 KPSC EDs phased in hsTn, providing an opportunity to examine differences in 30-day health outcomes and health care use among patients receiving risk stratification with the cTn HEART pathway (cTn group) vs the hsTn HEART pathway (hsTn group) in a multicenter pre-post cohort study.Our sample included adults (Ն18 years) presenting to the ED from January 1 to September 6, 2021, with chest pain who were risk stratified with a HEART pathway.We excluded transferred patients, those receiving hospice care, anyone with a COVID-19 diagnosis or positive test result during or within 1 month prior to the index ED visit, and anyone with do-not-resuscitate status.
All research procedures were conducted in accordance with the KPSC institutional review board, which granted a waiver of informed consent for this study because this was a quality improvement initiative and it was not feasible to collect informed consent on all eligible patients receiving care for suspected acute coronary syndrome.This study followed the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) reporting guideline.

Laboratory Troponin Testing and HEART Pathways
At least 2 months after the study's start and 5 weeks before its end, each hospital completely transitioned from cTn (AccuTnI+3 assay; Beckman Coulter Inc) and the conventional HEART pathway Coulter Diagnostics consultants that builds on the RAPID-TnT (Rapid Assessment of Possible Acute Coronary Syndrome in the Emergency Department With High-Sensitivity Troponin T) trial, 11 incorporating Beckman Coulter validation data and previously reported sex-specific AMI rule-in or rule-out thresholds. 13,14The main differences between the HEART pathways are (1) the new HEART pathway includes a branch point for presentation within or after 3 hours after symptom onset, (2)

Outcomes
Primary outcomes were all-cause mortality and AMI detection, both during the ED visit and within 30 days after.We ascertained AMI outcomes from International Statistical Classification of Diseases and Related Health Problems, Tenth Revision codes in billing documentation. 15Secondary outcomes were hospital admission, stress testing within 7 days, and coronary revascularization within 30 days.

Statistical Analysis
We compared baseline characteristics and outcomes between the cTn and hsTn groups using 2-sided  16 Race and ethnicity were included as proxies of social determinants of health and as potential confounding variables.Race and ethnicity were categorized as per National Institutes of Health standards for classification and obtained from the health plan's administrative records.

Cohort Characteristics
Of the total 28 258 adults who presented to a KPSC ED for chest pain and underwent troponin

Adjusted Mean Probability of AMI After the Index ED Visit and Within 30 Days
The adjusted odds ratio of AMI diagnosis after the index ED visit and within 30 days with risk stratification using the hsTn-based HEART pathway vs the cTn-based pathway was 0.54 (95% CI, 0.40-0.72)(eTable 3 in Supplement 1).The difference in marginal mean probability of AMI within 30 days between patients stratified by the hsTn pathway (0.014) patients stratified by the cTn pathway (0.024) was −0.010 (95% CI, −0.005 to −0.014), suggesting reduction of 1 AMI diagnosis within 30 days after the index visit for every 100 patients stratified with the hsTn vs cTn pathway.

Discussion
Compared with the cTn pathway, we found a small statistically significant increase in the percentage of patients risk stratified with the hsTn pathway who received a diagnosis of AMI at the index ED visit, coupled with a small statistically significant decrease in the percentage of patients who received a diagnosis of AMI in the 30 days after the index visit.Rates of admission, stress testing, and coronary revascularization were lower after implementation of the hsTn pathway; these reductions were small, although given the number of ED visits for chest pain, they may be meaningful.Our results suggest a useful role for an hsTn pathway for helping assess which patients have a lower risk of nearterm major adverse cardiac events, allowing them to be safely managed without hospital admission or invasive testing and also perhaps allowing for earlier diagnosis of patients with AMI (eg, during the ED visit).Further analyses are needed to determine whether, and to what extent, the small decreases in health care use seen using the more involved hsTn HEART pathway translate to meaningful cost savings with more widespread implementation.Although we found no difference in mortality between the hsTn and cTn groups, death was a rare event in both groups, and larger studies with a longer follow-up period are needed to further evaluate potential differences in mortality and morbidity.Furthermore, to address the question of the specificity of hsTn for detecting true AMI, future studies tracking individual patient results from subsequent tests (eg, angiography) after an elevated hsTn level could further differentiate the true-vs false-positive rate of AMI detection with hsTn.
Nearly one-third of patients (7963 of 28 258 [28.2%]) received troponin testing but were not assigned HEART scores and were therefore excluded from this analysis.Similar fractions of patients assigned to the cTn and hsTn pathways lacked HEART scores, suggesting that HEART scores were missing due to patient characteristics or clinician decisions rather than differences in pathway complexity.We found that the group of patients lacking HEART scores had increased cardiovascular risk (eg, higher mean Elixhauser comorbidity index score and higher prevalence of congestive heart failure) and poorer health outcomes (eg, greater rates of all-cause mortality, AMI diagnosis within 30 days, and coronary revascularization within 30 days).Therefore, it is possible that clinicians sometimes dismissed the HEART score-generating prompts for patients with many known cardiovascular risk factors and managed these patients as if they had high-risk scores.Overall, our finding that patients with missing HEART scores had poorer cardiovascular health and worse health outcomes reenforces the usefulness of the HEART pathway for patients at lower risk, for whom clinicians may rely more on a decision-making pathway than they would for patients with obvious AMI or with sufficiently high risk for AMI that they require admission regardless of HEART score.
There was 1 previous US study comparing the use of hsTn and cTn.In a small secondary subset analysis of patients enrolled in the HEART pathway randomized clinical trial who were risk stratified with hsTn rather than cTn, there was a similar detection of major adverse cardiac events in both groups 12 ; however, the study's small size limited its ability to have detected any differences.
European studies have shown no change in mortality or AMI when switching to hsTn, 11,17 while our study, embedding hsTn within a modified version of a widely used diagnostic algorithm, showed an increase in sensitivity for detecting AMI.Furthermore, our study showed a decrease in the admission rate from 15.0% to 12.2%, an 18.7% relative decrease vs other studies showing a 0% to 13% decrease. 11,17Given the potential benefits of hsTn to the health system demonstrated in our study, including fewer missed AMI diagnoses in the ED and decreased resource use, an important next step will be to test the replication in other US health systems and to examine outcomes of hsTn implementation specific to the ED (eg, length of ED stay) and other areas within the health care system.

Limitations
This study has some limitations.First, KPSC is a large, integrated health care system serving patients with health insurance, where prompt outpatient follow-up and testing may be challenging to replicate in other settings.Second, while a 30-day follow-up interval shows outcomes useful to ED clinicians, 1-year data would allow for a more comprehensive evaluation of the public health effect of hsTn.Third, this study occurred during the COVID-19 pandemic; although we excluded patients with active COVID-19, health and/or health care use outcomes may reflect increased ED discharge and decreased follow-up testing to protect patients from COVID-19.

Conclusions
In this multicenter pre-post cohort study of an hsTn-vs cTn-based HEART pathway as part of ED evaluation for AMI, the hsTn HEART pathway was associated with higher rates of ED AMI diagnoses and lower rates of AMI diagnoses after the index ED visit and within 30 days.In addition, we found less health service use for the hsTn group compared with the cTn group.An hsTn algorithm may

JAMA Network Open | Cardiology
High-Sensitivity vs Conventional Troponin HEART Pathway and health service use different among emergency department (ED) patients undergoing cardiac risk stratification with a HEART pathway using conventional vs high-sensitivity troponin level?Findings In a multicenter cohort study of 17 384 ED patients with chest pain risk stratified with a HEART pathway incorporating conventional vs highsensitivity troponin level, the highsensitivity troponin group showed a higher rate of acute myocardial infarction (AMI) diagnosis during the ED visit and a lower rate of AMI diagnosis within 30 days after the visit, along with lower rates of admission, stress testing, and revascularization.Mortality was rare and similar across groups.Meaning This study suggests that a high-sensitivity troponin HEART pathway may be associated with earlier detection of AMI and improved resource use.

JAMA
Network Open | Cardiology High-Sensitivity vs Conventional Troponin HEART Pathway (Figure, A) to hsTn (Access hsTnI assay; Beckman Coulter Inc) and a new HEART pathway (Figure, B) developed by KPSC emergency medicine, cardiology, and hospitalist physicians with Beckman

Figure .
Figure.Kaiser Permanente Southern California (KPSC) HEART Pathway Using Conventional Troponin and High-Sensitivity Troponin rule-in hsTn levels for late-presenting patients, (3) incorporation of repeat troponin values 2 hours apart, and (4) updated numerical scores given the increased sensitivity of hsTn.Ordering a cTn or hsTn test auto-prompted clinicians to document answers to HEART score questions to generate the HEART score used in the relevant HEART pathway.Cardiac risk stratification with either the cTnor hsTn-based HEART pathway was considered the intervention in this study.
Indeterminate: troponin change >5 to <11 pg/mL after 2 h Rule in: troponin change ≥11 pg/mL after 2 h Rule out: troponin change ≤5 pg/mL after 2 h No No No No Yes Yes Yes Yes The main differences between the HEART pathways are as follows: (1) the high-sensitivity troponin HEART pathway includes a branch point for presentation within or after 3 hours after symptom onset, (2) acute myocardial infarction rule-in high-sensitivity troponin levels for late-presenting patients, (3) incorporation of repeat troponin values 2 hours apart, and (4) updated numerical scores given high-sensitivity troponin's increased sensitivity.CAD indicates coronary artery disease; ECG, electrocardiography; NSTEMI, non-STsegment elevation myocardial infarction; and STEMI, ST-segment elevation myocardial infarction.SI conversion factor: To convert troponin to micrograms per liter, multiply by 0.001.aHigh-risk features: middle or left-sided pain, heavy chest pain, diaphoresis, relief of symptoms by sublingual nitrates, nausea or vomiting, radiating pain, and exertional pain.Low-risk features: welllocalized pain, sharp pain, nonexertional pain, no diaphoresis, and no nausea or vomiting.bNew ischemic changes: ischemic ST-segment depression and new ischemic T-wave changes.Nonspecific changes: repolarization abnormality, nonspecific T-wave changes, bundle-branch block, digoxin effect, pacemaker rhythms, left ventricular hypertrophy, early repolarization, and nonspecific ST-segment depression or elevation.cRisk factors: obesity (body mass index [calculated as e Percentages refer to 30-day mortality or acute myocardial infarction risk.fHEART score calculated as in A with substitution of assigned point values for high-sensitivity troponin levels by sex (women, >36 = 2 points, 12-36 = 1 point, <12 = 0 points; men, >60 = 2 points, 20-60 = 1 point, <20 = 0 points).AMI Wilcoxon rank sum tests (continuous variables) and χ 2 tests (categorical variables) with SAS, version 9.4 (SAS Institute Inc), with a significance level of P < .05.Supplemental analyses compared baseline characteristics and outcomes of patients with and patients without documented HEART scores using proportions testing (categorical variables) and t tests (continuous variables) with STATA SE, version 14.1 (StataCorp LP), with a significance level of P < .05.To account for potential differences in baseline characteristics between groups risk stratified with cTn vs hsTn HEART pathway, we used multivariable logistic regression to evaluate the odds ratio of AMI diagnosis after ED discharge but within 30 days of the index ED visit associated with the exposure of the cTn-vs hsTn-based HEART pathway for cardiac risk stratification.The regression model adjusted for the following potential confounders: age, sex, race (Asian, Black, or other), Hispanic ethnicity, HEART score category (low, moderate, or high risk), obesity (body mass index Ն30.0[calculated as weight in kilograms divided by height in meters squared]), prior diagnosis of coronary artery disease or stroke, prior coronary revascularization, family history of coronary artery disease, and Elixhauser comorbidity index.
a Fisher exact test.Compared

Table 2 .
Health Outcomes and Health Care Use a Fisher exact test.b Defined by diagnosis code and troponin value.c From ED arrival.