ABSTRACT
Background
This study aims to determine if differences in age affect safety and healthcare utilization among patients with chest pain in a multisite US ED cohort, after accounting for comorbidities and high‐sensitivity troponin (hs‐cTn).
Methods
We conducted a multisite observational study using the Wake Forest Chest Pain Registry, which included patients ≥ 18 years old with chest pain across 25 EDs (01/01/2021 to 12/31/2021). Each site used an hs‐cTn chest pain protocol. Patients were categorized as older (≥ 65 years), middle‐aged (46−64 years), or young (18−45 years). The primary safety outcome was death or MI at 30 days. The primary healthcare utilization outcome was 30‐day hospitalization. Multivariable logistic regression models assessed the association between age and outcomes, adjusting for sex, race, ethnicity, obesity, smoking, rurality, coronary disease, diabetes, hypertension, hyperlipidemia, insurance, site, and hs‐cTn, with young patients as the reference.
Results
Among 40 979 patients, 25.1% were older, 39.7% middle‐aged, and 35.2% young. At 30 days, death or MI occurred in 7.3% (750/10 298) of older, 3.8% (611/16 260) of middle‐aged, and 0.8% (108/14 421) of young patients. After adjustment, death or MI at 30 days was more common among older (aOR 1.57, 95% CI 1.17−2.12) and middle‐aged (aOR 1.57, 95% CI 1.22‐2.02) patients. Hospitalization at 30 days occurred in 56.3% (5799/10 298) of older, 35.4% (5761/16 260) of middle‐aged, and 12.8% (1849/14 421) of young patients. With adjustment, hospitalization remained more common among older (aOR 2.51, 95% CI 2.27−2.78) and middle‐aged (aOR 1.93, 95% CI 1.80−2.07) patients.
Conclusion
After adjusting for comorbidities and hs‐cTn results, older adults had higher rates of death or MI and hospitalization at 30 days.
Keywords: age differences, chest pain, high‐sensitivity cardiac troponin, older adults
Summary
This study of > 40 000 adult patients across 25 emergency department (ED) sites aims to determine if differences in age affect safety and healthcare utilization among patients presenting to the ED with chest pain, even after accounting for comorbidities and high‐sensitivity troponin (hs‐cTn).
After adjusting for comorbidities and hs‐cTn results, older adults had higher rates of death or MI and hospitalization at 30 days.
These findings suggest that accelerated diagnostic protocols or algorithms for risk stratification of ED patients with chest pain should strongly consider including age as a variable.
1. Introduction
Each year, more than 7 million patients present to an emergency department (ED) in the US with acute chest pain [1, 2]. Older adults are at particularly high risk for acute coronary syndrome (ACS) as age is a key risk marker for coronary heart disease [3, 4, 5]. It is well known that atherosclerotic plaque burden, thromboembolic risk, multivessel coronary artery disease, and coronary calcifications increase with age [6, 7]. Increasing age is also associated with functional decline, frailty, and dementia, which complicate providing emergency care to older adults with chest pain [7, 8]. Given that older adults may present with atypical symptoms for ACS and that they account for nearly 40% of ACS admissions [5, 9, 10], it is key for emergency clinicians to appreciate the risk involved when caring for older adults presenting with acute chest pain.
Despite the increased risk and challenging nature of caring for older patients with acute chest pain, most contemporary accelerated diagnostic protocols (ADPs) are troponin‐only in nature and do not consider age or other clinical variables. For example, the European Society of Cardiology 0/1‐h algorithm, the High‐STEACS Pathway, and the American College of Cardiology's Expert Consensus Decision Pathway do not consider age when risk‐stratifying patients for ACS [11, 12, 13, 14, 15, 16]. Our team has previously shown that troponin‐only ADPs place older adults at increased risk for missed myocardial infarction (MI) compared to their middle‐aged and young counterparts [17, 18]. However, our prior studies on this topic are limited by small sample size and inclusion of a limited number of ED sites [17, 18, 19].
To better understand the role of age in adverse cardiac events and healthcare utilization, we conducted a large, prospective observational cohort study of patients presenting with acute chest pain across 25 ED sites. The primary objective was to determine and compare the rates of all‐cause death and MI as well as hospitalization at 30 days for older, middle‐aged, and young patients while adjusting for known cardiovascular disease confounders and for index visit hs‐cTnI measures. Additional objectives included examining rates of major adverse cardiovascular events (MACE: all‐cause death, MI, and revascularization) and objective cardiac testing (OCT: stress testing and coronary angiography) at 30 days across age groups. We hypothesized that older adults would have increased 30‐day cardiac events, hospitalizations, and OCT compared to young patients, even after adjusting for their comorbidities and initial high‐sensitivity cardiac troponin (hs‐cTn) measure.
2. Methods
2.1. Study Design and Oversight
We conducted a multisite observational cohort study of patients presenting to the ED with acute chest pain. Patients were prospectively accrued to the Wake Forest Chest Pain Registry from January 1, 2021, to December 31, 2021, under a waiver of informed consent. This study was approved by the institutional review board. The methods of the Wake Forest Chest Pain Registry have been previously described [13, 20, 21]. The Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) guidelines helped direct the research and reporting processes [22].
2.2. Study Setting and Population
The Wake Forest Chest Pain Registry accrued patients from 25 EDs in North Carolina. These sites included two academic tertiary care centers and 23 community hospitals. The sites were geographically diverse and included urban, rural, and suburban areas. Across all sites, there were approximately 1 million ED encounters during the study year. Supporting Information S1: Table 1 describes each ED site. Each ED used a History, Electrocardiogram, Age, Risk factor, and Troponin (HEART) Pathway‐based ADP with high‐sensitivity cardiac troponin (hs‐cTn) for risk stratification purposes [23, 24]. Each site used the Beckman Coulter Access hs‐cTnI assay (Brea, CA), which has manufacturer reported sex‐specific 99th percentile URLs of 15 ng/L for women on the DXI platform and 12 ng/L on the Access 2 platform and 20 ng/L for men regardless of platform, with 10% coefficient of variation (CV) at 4 ng/L, and a limit of quantitation (LoQ) at 2 ng/L [25, 26, 27].
Patients ≥ 18 years old being evaluated for ACS who had a chief complaint of chest pain and at least one troponin ordered were included in the analysis. Patients with evidence of ST‐segment elevation myocardial infarction (STEMI) were excluded. Repeat visits for chest pain were identified and considered recurrent care. Patients transferred within the network were classified based on their initial ED visit, with care at the receiving hospital considered part of their index encounter. Consistent with prior cardiovascular studies, patients were classified as older (≥ 65 years), middle‐aged (46−64 years), or young (18‐45 years) [17, 18, 19].
2.3. Data Collection and Variables
Index encounter data (including ED, observation unit, and hospitalization data) through 30 days of follow‐up were extracted from the EHR (Cerner, Kansas City, MO and Clarity‐Epic Systems Corporation, Verona, WI). Pre‐validated, structured EHR variables and diagnosis and procedure codes (Current Procedural Technology [CPT] and International Classification of Diseases [ICD] 9 and 10 codes) were used to obtain patient demographics, comorbidities, cardiovascular risk factors, troponin results, dispositions, diagnoses, and vital status [24, 28, 29, 30, 31]. Age was determined in the EHR using the patient's self‐reported birthdate. The North Carolina State Center for Health Statistics was queried for death data not captured in our EHR. As with prior EHR‐based cardiovascular cohort studies, we assumed complete follow‐up [13, 17, 20, 24, 32, 33]. We previously found that these methods missed very few outcome events [24, 34].
2.4. Outcomes
The primary safety outcome was the composite of all‐cause death or MI at 30 days, inclusive of the index encounter. Secondary safety outcomes included major adverse cardiovascular events (MACE: the composite of all‐cause death, MI, and coronary revascularization) at 30 days as well as the individual MACE subcomponents at index through 30 days. Coronary revascularization included any percutaneous coronary intervention (PCI) with or without stent placement and coronary artery bypass grafting (CABG).
The primary healthcare utilization outcome was hospitalization at 30 days, which included inpatient and observation unit admissions, regardless of cause. Secondary healthcare utilization outcomes included OCT at 30 days, which was defined as non‐invasive testing (NIT: stress testing and coronary computed tomography angiography [CCTA]) and invasive coronary angiography.
2.5. Statistical Analysis
Frequencies and percentages or medians and interquartile ranges (IQRs) were used to describe the characteristics of patients within each age group. To assess the association between age group and each outcome, multivariable logistic regression was performed. Models were adjusted for sex, race, ethnicity, obesity, smoking, rurality, coronary artery disease, diabetes, hypertension, hyperlipidemia, insurance status, ED site, and initial hs‐cTn. Rurality was determined as either urban or rural using the Federal Office of Rural Health Policy definition [35]. Hs‐cTn was treated as a binary variable using the sex‐specific 99th percentile URL hs‐cTnI cut‐points. These variables were selected a priori due to their relevance and inclusion in previous cardiovascular risk stratification studies [13, 15, 17, 19, 20, 21]. Missing obesity and smoking status were imputed by predictive mean matching using all predictors and outcome variables to create 10 data sets with complete obesity and smoking status data. No other covariates required imputation. Separate, unadjusted logistic models were also fit to evaluate the association between age and each outcome. Young patients were considered the reference group for the primary analyses. Unadjusted odds ratios (ORs) or adjusted ORs (aORs) with corresponding 95% confidence intervals (95% CI) were calculated as appropriate for each model. We conducted two additional pre‐specified analyses. In the first, we assessed the association between age group and the study outcomes using the methods above, except that we compared older patients to middle‐aged patients. In the second analysis, we examined safety outcomes and healthcare utilization event rates by decade of life (e.g., 18−29 years, 30−39 years, 40−49 years, etc.).
3. Results
During the 12‐month study period, 40 979 patients were evaluated for acute chest pain. The cohort was 56.6% (23 188/40 979) female, 41.3% non‐White, (16 929/40 979), and had a median age of 52 (IQR 40−65) years. Older patients accounted for 25.1% (10 298/40 979), middle‐aged patients 39.7% (16 260/40 979), and young patients 35.2% (14 421/40 979) of the sample. At 30 days, the primary safety outcome of death or MI occurred in 3.6% (1469/40 979) and MACE in 4.5% (1828/40 979) of the overall cohort. Among all patients, 32.7% (13409/40 979) were hospitalized and 16.9% (6940/40 979) received OCT at 30 days. Table 1 describes the cohort. Figure 1 shows the study flow diagram.
Table 1.
Description of the cohort.
| Characteristic | Older | Middle‐aged | Young |
|---|---|---|---|
| (≥ 65 years) | (46–64 years) | (18−45 years) | |
| n = 10 298 | n = 16 260 | n = 14 421 | |
| n (%) | n (%) | n (%) | |
| Age, median (IQR), years | 74 (69‐79) | 54 (50‐59) | 35 (29‐41) |
| Female | 5838 (56.7) | 8951 (55.0) | 8399 (58.2) |
| Race/Ethnicity | |||
| Black, Not Hispanic | 2354 (22.9) | 5620 (34.6) | 5506 (38.2) |
| Hispanic or Latino | 276 (2.7) | 755 (4.6) | 1164 (8.1) |
| Other | 232 (2.2) | 495 (3.0) | 514 (3.6) |
| White, Not Hispanic | 7436 (72.2) | 9390 (57.8) | 7237 (50.2) |
| Insurance status | |||
| Other | 358 (3.5) | 3104 (19.1) | 4007 (27.8) |
| Private | 702 (6.8) | 8065 (49.6) | 6558 (45.5) |
| Public | 9238 (89.7) | 5091 (31.3) | 3856 (26.7) |
| Rurality | |||
| Rural | 2651 (25.7) | 3741 (23.0) | 2798 (19.4) |
| Urban | 7647 (74.3) | 12519 (77.0) | 11 623 (80.6) |
| Risk factors | |||
| BMI ≥ 30 kg/m2* | 5705 (54.1) | 7175 (55.7) | 3456 (40.8) |
| Smoking± | 4191 (40.8) | 6789 (42.0) | 4757 (33.4) |
| Diabetes | 2502 (24.3) | 2625 (16.1) | 654 (4.5) |
| Coronary artery disease | 2897 (28.1) | 2050 (12.6) | 186 (1.3) |
| Hypertension | 7299 (70.9) | 8175 (50.3) | 2675 (18.6) |
| Hyperlipidemia | 6910 (67.1) | 7379 (45.4) | 2132 (14.8) |
| Elevated Initial hs‐cTn | 2254 (21.9) | 1785 (11.0) | 610 (4.2) |
Note: *1823 older, 3375 middle‐aged, and 3883 young patients were missing BMI; ±25 older, 97 middle‐aged, and 166 young patients were missing smoking status.
Abbreviations: BMI, body mass index; hs‐cTn, high‐sensitivity cardiac troponin; IQR, interquartile range.
Figure 1.
Study flow diagram.
The primary safety outcome of death or MI at 30 days occurred in 7.3% (750/10 298) of older, 3.8% (611/16 260) of middle‐aged, and 0.8% (108/14 421) of young patients. At 30 days, MACE occurred in 9.2% (943/10 298) of older, 4.7% (765/16 260) of middle‐aged, and 0.8% (120/14 421) of young patients. After adjusting for comorbidities and hs‐cTn results, older adults still had higher rates of 30‐day death or MI (aOR 1.57, 95% CI 1.17−2.12) and 30‐day MACE (aOR 1.79, 95% CI 1.36−2.36) than young patients. Similarly, 30‐day death or MI (aOR 1.57, 95% CI 1.22−2.02) and 30‐day MACE (aOR 1.63, 95% CI 1.29−2.06) rates were higher among middle‐aged patients than young patients. Table 2 describes the safety events by age group. Table 3 presents the unadjusted and aORs for safety events at index through 30 days.
Table 2.
Safety and healthcare utilization outcomes from index through 30 days by age group.
| Older | Middle‐aged | Young | |
|---|---|---|---|
| (≥ 65 years) | (46−64 years) | (18−45 years) | |
| n = 10298 | n = 16 260 | n = 14 421 | |
| n (%) | n (%) | n (%) | |
| Safety | |||
| Index | |||
| MI | 632 (6.1) | 511 (3.1) | 83 (0.6) |
| Death | 75 (0.7) | 37 (0.2) | 7 (0.1) |
| Revascularization | 445 (4.3) | 391 (2.4) | 29 (0.2) |
| Death or MI | 645 (6.3) | 521 (3.2) | 88 (0.6) |
| MACE | 805 (7.8) | 656 (4.0) | 94 (0.7) |
| 30‐day follow‐up | |||
| MI | 224 (2.2) | 162 (1.0) | 29 (0.2) |
| Death | 163 (1.6) | 66 (0.4) | 9 (0.1) |
| Revascularization | 90 (0.9) | 68 (0.4) | 13 (0.1) |
| Death or MI | 267 (2.6) | 184 (1.1) | 35 (0.2) |
| MACE | 320 (3.1) | 219 (1.4) | 42 (0.3) |
| 30 days (index + follow‐up) | |||
| MI | 702 (6.8) | 580 (3.6) | 97 (0.7) |
| Death | 198 (1.9) | 88 (0.5) | 15 (0.1) |
| Revascularization | 523 (5.1) | 453 (2.8) | 40 (0.3) |
| Death or MI | 750 (7.3) | 611 (3.8) | 108 (0.8) |
| MACE | 943 (9.2) | 765 (4.7) | 120 (0.8) |
| Healthcare utilization | |||
| Index | |||
| Hospitalization | 5445 (52.9) | 5257 (32.3) | 1492 (10.4) |
| Objective cardiac testing | 2552 (24.8) | 2961 (18.2) | 499 (3.5) |
| Non‐invasive testing | 1710 (16.6) | 2244 (13.8) | 415 (2.9) |
| Invasive coronary angiography | 915 (8.9) | 780 (4.8) | 94 (0.7) |
| Follow‐up | |||
| Hospitalization | 953 (9.6) | 943 (5.8) | 498 (3.5) |
| Objective cardiac testing | 431 (4.2) | 532 (3.3) | 190 (1.3) |
| Non‐invasive testing | 377 (3.7) | 494 (3.0) | 180 (1.3) |
| Invasive coronary angiography | 201 (2.0) | 197 (1.2) | 43 (0.3) |
| 30 days (index + follow‐up) | |||
| Hospitalization | 5799 (56.3) | 5761 (35.4) | 1849 (12.8) |
| Objective cardiac testing | 2887 (28.0) | 3387 (20.8) | 666 (4.6) |
| Non‐invasive testing | 2018 (19.6) | 2,652 (16.3) | 576 (4.0) |
| Invasive coronary angiography | 964 (9.4) | 815 (5.0) | 103 (0.7) |
Abbreviations: MACE, major adverse cardiovascular event (death, MI, or revascularization); MI, myocardial infraction.
Table 3.
Unadjusted and adjusted odds ratios for index and 30‐day study outcomes for older and middle‐aged patients, with young patients being the reference group.
| Older: Young | Middle‐aged: Young | |||
|---|---|---|---|---|
| Outcome | OR (95% CI) | aOR (95% CI) | OR (95% CI) | aOR (95% CI) |
| Safety | ||||
| Index | ||||
| MI | 11.29 (8.97, 14.22) | 1.44 (1.03, 2.02) | 5.61 (4.44, 7.08) | 1.55 (1.16, 2.07) |
| Death or MI | 10.88 (8.7, 13.62) | 1.40 (1.01, 1.95) | 5.39 (4.3, 6.77) | 1.52 (1.15, 2.01) |
| MACE | 12.92 (10.42, 16.03) | 1.62 (1.19, 2.21) | 6.41 (5.16, 7.96) | 1.62 (1.24, 2.12) |
| 30 days (index + follow‐up) | ||||
| MI | 10.80 (8.72, 13.38) | 1.52 (1.11, 2.08) | 5.46 (4.4, 6.78) | 1.56 (1.20,2.03) |
| Death or MI | 10.41 (8.49, 12.76) | 1.57 (1.17, 2.12) | 5.17 (4.21, 6.36) | 1.57 (1.22, 2.02) |
| MACE | 12.01 (9.92, 14.55) | 1.79 (1.36, 2.36) | 5.88 (4.85, 7.14) | 1.63 (1.29, 2.06) |
| Healthcare utilization | ||||
| Index | ||||
| Hospitalization | 9.72 (9.1, 10.39) | 2.66 (2.39, 2.96) | 4.14 (3.89, 4.41) | 2.03 (1.88, 2.19) |
| Objective cardiac testing | 9.19 (8.32, 10.16) | 3.07 (2.64, 3.57) | 6.21 (5.63, 6.85) | 2.88 (2.58, 3.21) |
| Non‐invasive testing | 6.72 (6.02, 7.51) | 3.46 (2.93, 4.07) | 7.68 (6.19, 9.52) | 1.82 (1.42, 2.35) |
| Invasive coronary angiography | 14.86 (12.0, 18.41) | 1.71 (1.27, 2.32) | 5.40 (4.85, 6.02) | 3.00 (2.67, 3.38) |
| 30 Days (Index + Follow‐up) | ||||
| Hospitalization | 8.76 (8.23, 9.33) | 2.51 (2.27, 2.78) | 3.73 (3.52, 3.96) | 1.93 (1.8, 2.07) |
| Objective cardiac testing | 8.05 (7.36, 8.79) | 3.03 (2.64, 3.47) | 5.43 (4.98, 5.92) | 2.68 (2.43, 2.95) |
| Non‐invasive testing | 5.86 (5.32, 6.45) | 3.35 (2.89, 3.87) | 4.68 (4.27, 5.14) | 2.77 (2.5, 3.07) |
| Invasive coronary angiography | 14.36 (11.7, 17.62) | 1.73 (1.29, 2.31) | 7.34 (5.97, 9.02) | 1.77 (1.38, 2.25) |
Abbreviations: aOR, adjusted odds ratio; CI, confidence interval, MACE, major adverse cardiovascular event (death, MI, or revascularization); OR, odds ratio.
The primary healthcare utilization outcome of hospitalization at 30 days occurred in 56.3% (5799/10 298) of older, 35.4% (5761/16 260) of middle‐aged, and 12.8% (1849/14 421) of young patients. OCT at 30 days was obtained in 28.0% (2887/10 298) of older, 20.8% (3387/16 260) of middle‐aged, and 4.6% (666/14 421) of young patients. With adjustment, hospitalization at 30 days remained more common among older (aOR 2.51, 95% CI 2.27−2.78) and middle‐aged (aOR 1.93, 95% CI: 1.80−2.07) patients than young patients. OCT was also more common among older (aOR 3.03, 95% CI 2.64−3.47) and middle‐aged (aOR 2.68, 95% CI 2.43−2.95) patients compared to young patients. Table 2 describes healthcare utilization by age group. Table 3 shows the unadjusted and aORs for healthcare utilization outcomes at index through 30 days. Supporting Information S1: Table 2 presents healthcare utilization outcomes by decade of life.
In our pre‐specified analysis evaluating the association between age with the study outcomes, where middle‐aged patients were the reference group, we found that older adults had similar rates of 30‐day death or MI (aOR 0.99, 95% CI 0.84−1.15) and 30‐day MACE (aOR 1.05, 95% CI 0.91−1.21). However, older adults had increased hospitalizations (aOR 1.36, 95% CI 1.26−1.47) and OCT (aOR 1.17, 95% CI 1.08−1.27) compared to middle‐aged patients. When examining outcomes by each decade of life, we found that 30‐day safety events and healthcare utilization tended to become more prevalent with each decade of life. However, OCT rates trended down among the oldest adults (≥ 90 years). Supporting Information S1: Table 2 presents study outcomes by decade of life. Supporting Information S1: Table 3 presents the association between age group with study outcome among older adults with middle‐aged adults as the reference group.
4. Discussion
In this large multisite U.S. ED cohort study, we found that older and middle‐aged patients were more likely to have death or MI at 30 days compared to young patients after adjustment for key comorbidities associated with cardiovascular disease and index visit hs‐cTn measures. We also observed higher adjusted odds of healthcare utilization among older and middle‐aged adults for hospitalization and OCT. Thus, this study provides further evidence that age remains a key independent cardiovascular risk factor for ED patients with acute chest pain, even in the era of hs‐cTn. This suggests that future ADPs should incorporate age when risk‐stratifying patients for possible ACS and that many existing hs‐cTn ADPs, which do not include age as a variable, may underestimate risk in older patients.
While adjusting for well‐known cardiovascular comorbidities and hs‐cTn results did attenuate the effect size, age remained strongly associated with death, MI, and MACE at 30‐days. It is well known that increasing age is associated with greater atherosclerotic plaque burden, thromboembolic risk, and multivessel coronary calcifications, leading to increased cardiovascular events [6, 7]. However, our findings suggest that hs‐cTn and other cardiovascular risk factor phenotypes (e.g., hyperlipidemia, diabetes, etc.) insufficiently explain this disease and risk progression. While hs‐cTn is the gold‐standard biomarker for myocardial injury and does have important prognostic value for downstream cardiovascular risk, it is not a marker of atherosclerosis burden or even myocardial ischemia in the absence of infarction. Thus, unless novel biomarkers or clinical variables are uncovered that more fully explain risk progression with aging, incorporating patient age into future ADPs may be necessary for accurate risk stratification. Given the widespread use of troponin‐only ADPs, which risk‐stratify patients solely based on hs‐cTn measures, these findings are highly clinically relevant. Further, these data are consistent with prior data demonstrating that troponin‐only ADPs, such as the European Society of Cardiology 0/1‐h algorithm, have high rates of missed cardiac events among older and middle‐aged patients in the United States [19].
In addition to demonstrating an increased adjusted odds of cardiac events with increasing age, our study also demonstrated that healthcare resource use, in the form of hospitalizations and OCT, increased with age after adjusting for comorbidities and hs‐cTn measures. These findings were not consistent with prior literature, as numerous prior studies have reported cardiovascular care disparities and under‐testing among older adults [7, 10, 36]. In fact, it seems likely that older adults were actually being over‐tested in our cohort. For example, 9.4% of older adults had invasive coronary angiography, but just 5.1% had a revascularization procedure performed at 30 days. In light of the ISCHEMIA and COURAGE trials and the potential for over‐testing among older adults, clinicians should thoughtfully consider the role of invasive angiography with potential PCI compared to medical management alone [37, 38, 39].
We also found a low prevalence of cardiac events among young patients in this cohort, with just 0.8% having 30‐day death or MI. Despite < 1% of young patients having an adverse event, more than 10% were hospitalized, and nearly 5% received OCT, further suggesting that clinicians are likely over‐testing in this age group. In contrast to the low prevalence of disease among young patients in this cohort, other chest pain cohorts report that 2%–7% of young patients have death or MI at 30 days [17, 18, 19, 40, 41]. Therefore, even though we found young patients to have a very low event rate in this cohort, clinicians must consider the pre‐test probability of disease in their patient population before determining that hospitalization and OCT are not indicated in young patients.
Future studies focused on cardiovascular care differences driven by age should aim to elucidate the mechanism driving the increased risk conferred by age. Ideally, future chest pain risk stratification work will account for frailty, which we are unable to adjust for in our current chest pain registry. Additionally, while we adjusted for initial hs‐cTnI using the established sex‐specific 99th percentile URL cut‐points, age‐specific hs‐cTn cut‐points are not presently recommended [42, 43]. It is unknown whether age‐specific hs‐cTn cut‐points would improve chest pain risk stratification, particularly among older adults. Finally, in the future, machine learning and artificial intelligence risk stratification tools, such as ARTEMIS and CoDE‐ACS, will likely leverage a large number of clinical variables, including age [44, 45]. We believe this will be important, especially as this study adds to a growing body of literature showing that age is a key risk marker, even after accounting for comorbidities and hs‐cTn.
5. Limitations
This study has limitations. Although this study was conducted in one state, we accrued patients from 25 diverse EDs that include rural, suburban, and urban areas. The EDs also include both academic tertiary care centers and community hospitals. While our use of the EHR to identify outcome events may be less reliable than traditional follow‐up methods, it is pragmatic and allowed us to evaluate more than 40 000 ED patients. Additionally, our prior studies suggest a high degree of validity with using the EHR for follow‐up, as we have traditionally missed very few events [13, 20, 24]. It is important to note that our outcomes are not adjudicated, thus risking misclassification bias. Furthermore, we are also unable to determine if death and hospitalization events were cardiac‐related or not.
6. Conclusion
In this large multisite US study, we observed that older and middle‐aged patients presenting to the ED with acute chest pain were more likely to have death or MI at 30 days than young patients, even after adjustment for comorbidities and hs‐cTn measures. Older and middle‐aged patients were also hospitalized more frequently and received more cardiac testing than young patients. While clinicians might expect older patients to have higher rates of death or MI and healthcare utilization due to their comorbidities, we found that older and middle‐aged patients were still at higher risk of adverse events and healthcare utilization even after adjusting for comorbidities and hs‐cTn. These findings suggest that ADPs or algorithms for risk stratification of ED patients with possible ACS should strongly consider including age as a variable.
Author Contributions
Nicklaus P. Ashburn, Simon A. Mahler, and Anna C. Snavely conceived the study idea. Lyle Paukner and Anna C. Snavely coordinated data management. Anna C. Snavely and Lyle Paukner performed the data analysis. David A. Pearson assisted with trial coordination. All authors contributed to the manuscript and its revision. Nicklaus P. Ashburn takes responsibility for the manuscript as a whole.
Conflicts of Interest
Dr. Ashburn receives funding from NHLBI (K23HL169929), AHRQ (R01HS029017), and the Emergency Medicine Foundation.
Dr. Snavely receives funding from NHLBI (K23HL169929), Abbott Laboratories, HRSA (1H2ARH399760100), AHRQ (R01HS029017 and R21HS029234), and the Emergency Medicine Foundation.
Dr. Supples receives funding from the NIH (UL1TR001420), AHRQ (R01HS029017), and the National Foundation of Emergency Medicine.
Dr. Hutchison receives funding from the Emergency Medicine Foundation.
Dr. Pearson is an investor and advisor for Power Medical Inc.
Dr. Mahler receives funding/support from Roche Diagnostics, Abbott Laboratories, QuidelOrtho, Siemens, Grifols, Polymedco, Beckman Coulter, Genetesis, Cytovale, National Foundation of Emergency Medicine, BlueJay Diagnostics, Reprieve Cardiovascular, Duke Endowment, Brainbox, HRSA (1H2ARH399760100), the Emergency Medicine Foundation, and AHRQ (R01HS029017 and R21HS029234). He is a consultant for Roche, QuidelOrtho, Abbott, Siemens, Polymedco, Inflammatix, and Radiometer, and is the Chief Medical Officer for Impathiq Inc. The other authors declare no conflicts of interest.
Supporting information
Supplemental Material ‐ Age ‐ CC.
Supplementary Information
Acknowledgments
This project was funded internally by Wake Forest University Health Sciences. We acknowledge the assistance of the Wake Forest Clinical and Translational Science Institute, supported by the National Center for Advancing Translational Sciences (NCATS), National Institutes of Health, through UL1TR001420 (Donald McClain, PI). The WF CTSI assisted with data extraction. Funding sources had no role in design or conduct of this investigation. This includes no role in the collection, management, analysis, and interpretation of data nor preparation, review, or approval of the manuscript.
Data Availability Statement
The data, analytic methods, and study materials will not be made available to other researchers.
References
- 1. Owens P. L., Barrett M. L., Gibson T. B., Andrews R. M., Weinick R. M., and Mutter R. L., “Emergency Department Care in the United States: A Profile of National Data Sources,” Annals of Emergency Medicine 56, no. 2 (2010): 150–165, 10.1016/j.annemergmed.2009.11.022. [DOI] [PubMed] [Google Scholar]
- 2. Virani S. S., Alonso A., Aparicio H. J., et al., “Heart Disease and Stroke Statistics‐2021 Update: A Report From the American Heart Association,” Circulation 143, no. 8 (2021): e254–e743, 10.1161/CIR.0000000000000950. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3. Duggan J. P., Peters A. S., Trachiotis G. D., and Antevil J. L., “Epidemiology of Coronary Artery Disease,” Surgical Clinics of North America 102, no. 3 (2022): 499–516, 10.1016/j.suc.2022.01.007. [DOI] [PubMed] [Google Scholar]
- 4. D'Agostino R. B., Vasan R. S., Pencina M. J., et al., “General Cardiovascular Risk Profile for Use in Primary Care: The Framingham Heart Study,” Circulation 117, no. 6 (2008): 743–753, 10.1161/CIRCULATIONAHA.107.699579. [DOI] [PubMed] [Google Scholar]
- 5. Gulati M., Levy P. D., Mukherjee D., et al., “2021 AHA/ACC/ASE/CHEST/SAEM/SCCT/SCMR Guideline for the Evaluation and Diagnosis of Chest Pain,” Journal of the American College of Cardiology 78, no. 22 (2021): e187–e285, 10.1016/j.jacc.2021.07.053. [DOI] [PubMed] [Google Scholar]
- 6. Singam N. S. V., Fine C., and Fleg J. L., “Cardiac Changes Associated With Vascular Aging,” Clinical Cardiology 43, no. 2 (2020): 92–98, 10.1002/clc.23313. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7. Damluji A. A., Forman D. E., Wang T. Y., et al., “Management of Acute Coronary Syndrome in the Older Adult Population: A Scientific Statement From the American Heart Association,” Circulation 147, no. 3 (2023): e32–e62, 10.1161/CIR.0000000000001112. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8. Damluji A. A., Forman D. E., van Diepen S., et al., “Older Adults in the Cardiac Intensive Care Unit: Factoring Geriatric Syndromes in the Management, Prognosis, and Process of Care: A Scientific Statement From the American Heart Association,” Circulation 141, no. 2 (2020): e6–e32, 10.1161/CIR.0000000000000741. [DOI] [PubMed] [Google Scholar]
- 9. De Luca L., Marini M., Gonzini L., et al., “Contemporary Trends and Age‐Specific Sex Differences in Management and Outcome for Patients With St‐Segment Elevation Myocardial Infarction,” Journal of the American Heart Association 5, no. 12 (2016): e004202, 10.1161/JAHA.116.004202. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10. Lowenstern A., Alexander K. P., Hill C. L., et al., “Age‐Related Differences in the Noninvasive Evaluation for Possible Coronary Artery Disease: Insights From the Prospective Multicenter Imaging Study for Evaluation of Chest Pain (PROMISE) Trial,” JAMA Cardiology 5, no. 2 (2020): 193–201, 10.1001/jamacardio.2019.4973. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11. Boeddinghaus J., Nestelberger T., Twerenbold R., et al., “Impact of Age on the Performance of the ESC 0/1h‐algorithms for Early Diagnosis of Myocardial Infarction,” European Heart Journal 39, no. 42 (2018): 3780–3794, 10.1093/eurheartj/ehy514. [DOI] [PubMed] [Google Scholar]
- 12. 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,” Archives of Internal Medicine 172, no. 16 (2012): 1211–1218, 10.1001/archinternmed.2012.3698. [DOI] [PubMed] [Google Scholar]
- 13. Mahler S. A., Ashburn N. P., Supples M. W., Hashemian T., and Snavely A. C., “Validation of the ACC Expert Consensus Decision Pathway for Patients With Chest Pain,” Journal of the American College of Cardiology 83, no. 13 (2024): 1181–1190, 10.1016/j.jacc.2024.02.004. [DOI] [PubMed] [Google Scholar]
- 14. Kontos M. C., de Lemos J. A., Deitelzweig S. B., et al., “2022 ACC Expert Consensus Decision Pathway on the Evaluation and Disposition of Acute Chest Pain in the Emergency Department,” Journal of the American College of Cardiology 80, no. 20 (2022): 1925–1960, 10.1016/j.jacc.2022.08.750. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15. Ashburn N. P., Snavely A. C., Supples M. W., et al., “Performance of the High‐STEACS Early Rule out Pathway Using hs‐cTnT at 30 Days in a Multisite US Cohort,” Circulation: Cardiovascular Quality and Outcomes 18, no. 2 (2025): e011084, 10.1161/CIRCOUTCOMES.124.011084. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16. Chapman A. R., Hesse K., Andrews J., et al., “High‐Sensitivity Cardiac Troponin I and Clinical Risk Scores in Patients With Suspected Acute Coronary Syndrome,” Circulation 138, no. 16 (2018): 1654–1665, 10.1161/CIRCULATIONAHA.118.036426. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17. Supples M., Smits S., Villenthi A., Snavely A. C., Ashburn N. P., and Mahler S. A., “Safety and Effectiveness of the High‐Sensitivity HEART Pathway Safely Across Age Groups in Patients With Suspected Acute Coronary Syndrome,” American Journal of Cardiology 250 (2025): 38–47, 10.1016/j.amjcard.2025.04.022. [DOI] [PubMed] [Google Scholar]
- 18. Ashburn N. P., Snavely A. C., Paradee B. E., O'Neill J. C., Stopyra J. P., and Mahler S. A., “Age Differences in the Safety and Effectiveness of the HEART Pathway Accelerated Diagnostic Protocol for Acute Chest Pain,” Journal of the American Geriatrics Society 70, no. 8 (2022): 2246–2257, 10.1111/jgs.17777. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19. Millard M. J., Ashburn N. P., Snavely A. C., et al., “European Society of Cardiology 0/1‐hour Algorithm (High‐Sensitivity Cardiac Troponin T) Performance Across Distinct Age Groups,” Heart 110, no. 12 (2024): 838–845, 10.1136/heartjnl-2023-323621. [DOI] [PubMed] [Google Scholar]
- 20. Mahler S. A., Ashburn N. P., Paradee B. E., Stopyra J. P., O'Neill J. C., and Snavely A. C., “Safety and Effectiveness of the High‐Sensitivity Cardiac Troponin HEART Pathway in Patients With Possible Acute Coronary Syndrome,” Circulation: Cardiovascular Quality and Outcomes 17, no. 3 (2024): e010270, 10.1161/CIRCOUTCOMES.123.010270. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21. Ashburn N. P., Snavely A. C., Villenthi A., Hashemian T., Supples M. W., and Mahler S. A., “Multisite Validation of a Strategy to Identify Very Low Risk Emergency Department Patients Without Troponin,” JACC: Advances 4, no. 7 (2025): 101913, 10.1016/j.jacadv.2025.101913. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.The Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) Statement: Guidelines for Reporting Observational Studies | The EQUATOR Network, http://www.equator-network.org/reporting-guidelines/strobe/. [DOI] [PMC free article] [PubMed]
- 23. Mahler S. A., Riley R. F., Hiestand B. C., et al., “The HEART Pathway Randomized Trial: Identifying Emergency Department Patients With Acute Chest Pain for Early Discharge,” Circulation: Cardiovascular Quality and Outcomes 8, no. 2 (2015): 195–203, 10.1161/CIRCOUTCOMES.114.001384. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24. Mahler S. A., Lenoir K. M., Wells B. J., et al., “Safely Identifying Emergency Department Patients With Acute Chest Pain for Early Discharge,” Circulation 138, no. 22 (2018): 2456–2468, 10.1161/circulationaha.118.036528. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25. Mark D. G., Huang J., Lee K. K., et al., “Implementation of a High‐Sensitivity Cardiac Troponin Assay and Diagnostic Protocol for Suspected Acute Coronary Syndrome,” American Journal of Cardiology 251 (2025): 25–33, 10.1016/j.amjcard.2025.05.005. [DOI] [PubMed] [Google Scholar]
- 26. Storti S., Masotti S., Prontera C., et al., “Evaluation of Analytical Performance and Comparison of Clinical Results of the New Generation Method AccuTnI+3 for the Measurement of Cardiac Troponin I Using Both Patients and Quality Control Plasma Samples,” Clinica Chimica Acta 451, no. Pt B (2015): 129–134, 10.1016/j.cca.2015.09.016. [DOI] [PubMed] [Google Scholar]
- 27. Nestelberger T., Boeddinghaus J., Greenslade J., et al., “Two‐Hour Algorithm for Rapid Triage of Suspected Acute Myocardial Infarction Using a High‐Sensitivity Cardiac Troponin I Assay,” Clinical Chemistry 65, no. 11 (2019): 1437–1447, 10.1373/clinchem.2019.305193. [DOI] [PubMed] [Google Scholar]
- 28. Quan H., Sundararajan V., Halfon P., et al., “Coding Algorithms for Defining Comorbidities in ICD‐9‐CM and ICD‐10 Administrative Data,” Medical Care 43, no. 11 (2005): 1130–1139, 10.1097/01.mlr.0000182534.19832.83. [DOI] [PubMed] [Google Scholar]
- 29. So L., Evans D., and Quan H., “ICD‐10 Coding Algorithms for Defining Comorbidities of Acute Myocardial Infarction,” BMC Health Services Research 6 (2006): 161, 10.1186/1472-6963-6-161. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30. Tonelli M., Wiebe N., Fortin M., et al., “Methods for Identifying 30 Chronic Conditions: Application to Administrative Data,” BMC Medical Informatics and Decision Making 15 (2015): 31, 10.1186/s12911-015-0155-5. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31. Riley R. F., Don C. W., Powell W., Maynard C., and Dean L. S., “Trends in Coronary Revascularization in the United States From 2001 to 2009: Recent Declines in Percutaneous Coronary Intervention Volumes,” Circulation: Cardiovascular Quality and Outcomes 4, no. 2 (2011): 193–197, 10.1161/CIRCOUTCOMES.110.958744. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32. Stewart J., Hashemian T., Snavely A. C., Ashburn N. P., Supples M. W., and Mahler S. A., “Performance of Sex‐Specific Vs Single High‐Sensitivity Troponin Thresholds in a Large Multisite Chest Pain Cohort,” Academic Emergency Medicine 31, no. S1 (2024): 255. [Google Scholar]
- 33. Veasey C. J., Snavely A. C., Kearns Z. L., Ashburn N. P., Hashemian T., and Mahler S. A., “The High‐Sensitivity HEART Pathway Safely Reduces Hospitalizations Regardless of Sex or Race in a Multisite Prospective Us Cohort,” Clinical Cardiology 47, no. 10 (2024): e70027, 10.1002/clc.70027. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34. Stopyra J. P., Snavely A. C., Lenoir K. M., et al., “HEART Pathway Implementation Safely Reduces Hospitalizations at One Year in Patients With Acute Chest Pain,” Annals of Emergency Medicine 76, no. 5 (2020): 555–565, 10.1016/j.annemergmed.2020.05.035. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 35.How We Define Rural, accessed July 9, 2025, https://www.hrsa.gov/rural-health/about-us/what-is-rural.
- 36. Diercks D. B. and Miller C. D., “Disparities in the Care of Chest Pain,” Canadian Medical Association Journal 179, no. 7 (2008): 631–633, 10.1503/cmaj.081190. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 37. Boden W. E., O'Rourke R. A., Teo K. K., et al., “Optimal Medical Therapy With or Without Pci for Stable Coronary Disease,” New England Journal of Medicine 356, no. 15 (2007): 1503–1516, 10.1056/NEJMoa070829. [DOI] [PubMed] [Google Scholar]
- 38. Maron D. J., Hochman J. S., Reynolds H. R., et al., “Initial Invasive or Conservative Strategy for Stable Coronary Disease,” New England Journal of Medicine 382, no. 15 (2020): 1395–1407, 10.1056/NEJMoa1915922. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 39. Jou S., Gelfman L. P., Alexander K. P., et al., “Clinical Practice Patterns Among Older Multimorbid Adults Presenting With Suspected Ischemic Symptoms: A Multi‐Center Survey,” American Heart Journal Plus: Cardiology Research and Practice 52, no. 100517 (2025): 100517, 10.1016/j.ahjo.2025.100517. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 40. Tungsubutra W., Tresukosol D., Buddhari W., et al., “Acute Coronary Syndrome in Young Adults: The Thai ACS Registry,” Journal of the Medical Association of Thailand 90, no. Suppl 1 (2007): 81–90, https://pubmed.ncbi.nlm.nih.gov/18431890/. [PubMed] [Google Scholar]
- 41. Avezum A., Makdisse M., Spencer F., et al., “Impact of Age on Management and Outcome of Acute Coronary Syndrome: Observations From the Global Registry of Acute Coronary Events (Grace),” American Heart Journal 149, no. 1 (2005): 67–73, 10.1016/j.ahj.2004.06.003. [DOI] [PubMed] [Google Scholar]
- 42. Vasile V. C. and Jaffe A. S., “High‐Sensitivity Cardiac Troponin for the Diagnosis of Patients With Acute Coronary Syndromes,” Current Cardiology Reports 19, no. 10 (2017): 92, 10.1007/s11886-017-0904-4. [DOI] [PubMed] [Google Scholar]
- 43. Sandoval Y. and Jaffe A. S., “Caveats of High‐Sensitivity Cardiac Troponin Rapid Risk Stratification Protocols: One Size Does Not Fit All,” Heart 110, no. 12 (2024): 821–822, 10.1136/heartjnl-2024-323885. [DOI] [PubMed] [Google Scholar]
- 44. Neumann J. T., Twerenbold R., Ojeda F., et al., “Personalized Diagnosis in Suspected Myocardial Infarction,” Clinical Research in Cardiology 112, no. 9 (2023): 1288–1301, 10.1007/s00392-023-02206-3. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 45. Doudesis D., Lee K. K., Boeddinghaus J., et al., “Machine Learning for Diagnosis of Myocardial Infarction Using Cardiac Troponin Concentrations,” Nature Medicine 29, no. 5 (2023): 1201–1210, 10.1038/s41591-023-02325-4. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Supplemental Material ‐ Age ‐ CC.
Supplementary Information
Data Availability Statement
The data, analytic methods, and study materials will not be made available to other researchers.