2021 AHA/ACC/ASE/CHEST/SAEM/SCCT/SCMR Guideline for the Evaluation and Diagnosis of Chest Pain: A Report of the American College of Cardiology/American Heart Association Joint Committee on Clinical Practice Guidelines

After injuries, chest pain is the second most common reason for adults to present to the emergency department (ED) in the United States and accounts for >6.5 million visits, which is 4.7% of all ED visits.¹ Chest pain also leads to nearly 4 million outpatient visits annually in the United States.² Chest pain remains a diagnostic challenge in the ED and outpatient setting and requires thorough clinical evaluation. Although the cause of chest pain is often noncardiac, coronary artery disease (CAD) affects >18.2 million adults in the United States and remains the leading cause of death for men and women, accounting for >365 000 deaths annually.³ Distinguishing between serious and benign causes of chest pain is imperative. The lifetime prevalence of chest pain in the United States is 20% to 40%,⁴ and women experience this symptom more often than men.⁵ Of all ED patients with chest pain, only 5.1% will have an acute coronary syndrome (ACS), and more than half will ultimately be found to have a noncardiac cause.⁶ Nonetheless, chest pain is the most common symptom of CAD in both men and women.

Chest pain is one of the most common reasons that people seek medical care. The term “chest pain” is used by patients and applied by clinicians to describe the many unpleasant or uncomfortable sensations in the anterior chest that prompt concern for a cardiac problem. Chest pain should be considered acute when it is new onset or involves a change in pattern, intensity, or duration compared with previous episodes in a patient with recurrent symptoms. Chest pain should be considered stable when symptoms are chronic and associated with consistent precipitants such as exertion or emotional stress.

Although the term chest pain is used in clinical practice, patients often report pressure, tightness, squeezing, heaviness, or burning. In this regard, a more appropriate term is “chest discomfort,” because patients may not use the descriptor “pain.” They may also report a location other than the chest, including the shoulder, arm, neck, back, upper abdomen, or jaw. Despite individual variability, the discomfort induced by myocardial ischemia is often characteristic and therefore central to the diagnosis. For this reason, features more likely to be associated with ischemia have been described as typical versus atypical; however, the latter can be confusing because it is frequently used to describe symptoms considered nonischemic as well as noncardiac. Although other nonclassic symptoms of ischemia, such as shortness of breath, nausea, radiating discomfort, or numbness, may be present, chest pain or chest discomfort remains the predominant symptom reported in men and women who are ultimately diagnosed with myocardial ischemia.^3-7 Pain—described as sharp, fleeting, related to inspiration (pleuritic) or position, or shifting locations—suggests a lower likelihood of ischemia.

Most patients who present to the ED with chest pain are women, particularly among those ≥65 years of age.⁸ The ISCHEMIA (International Study of Comparative Health Effectiveness With Medical and Invasive Approaches) trial demonstrated that women with moderate-to-severe ischemia are more symptomatic than men.⁹ Women are less likely to have timely and appropriate care.¹⁰ This could be explained by the fact that women are more likely to experience prodromal symptoms when they seek medical care.¹¹ Women may also present with accompanying symptoms (eg, nausea, fatigue, and shortness of breath) more often than men.^12-14 However, chest pain remains the predominant symptom reported by women among those ultimately diagnosed with ACS, occurring with a frequency equal to men.^3,5-7,15,16

Increased age is a significant risk factor for ACS. However, it is also a risk factor for comorbidities that are associated with alternative diagnoses associated with chest pain. As a result, a more extensive diagnostic workup is required in older patients. Although patients >75 years of age account for 33% of all cases of ACS, alternative diagnoses are still more common than a cardiac cause of chest pain at presentation.^2,3 A substudy of the PROMISE trial has shown that patients >75 years of age, with stable symptoms suggestive of CAD, are more likely to have a positive noninvasive test and more coronary artery calcification than younger people. For these older patients, when compared with anatomic noninvasive testing for obstructive CAD with cardiac CT, a positive stress test result was associated with increased risk of cardiovascular death or MI.⁴

There are marked racial and ethnic disparities when triaging patients who present for the evaluation of chest pain. Despite a greater number of Black patients presenting with angina pectoris relative to other races, this population is less likely to be treated urgently and less likely to have an ECG performed, samples for cardiac biomarkers drawn, cardiac monitoring performed, or pulse oximetry measured.^1-4 Similar treatment disparities are found with Hispanic patients and those who are covered by Medicaid or are uninsured. Derived from a nationally representative sample from the National Hospital Ambulatory Health Care Survey reflecting an estimated 78 million ED visits in the United States over a 10-year period, these findings have been unchanged over time.⁵ Such disparity in the management of chest pain among diverse population subgroups contributes to worse outcomes, including the greater incidence of AMI and fatal coronary events seen in these key population subgroups.^6,7 There are also disparities in the management of patients of South Asian descent who present with ACS, with the diagnosis often missed or delayed, resulting in poor outcomes.^8-11 Consideration of race and ethnicity in the evaluation of patients with suspected ACS and in the outpatient evaluation of symptomatic patients is paramount to improving outcomes. Cultural competency training of providers is recommended to address health disparities in the evaluation and management of diverse patient population subgroups with chest pain.

Although chest pain remains one of the most common reasons that patients seek evaluation, among both sexes, there is a tendency for some patients to minimize perceived risk for cardiac disease, resulting in potentially avoidable delays in care.¹ To alleviate this problem, efforts should be made to educate all people regarding their risk for a cardiac event and educate patients about the need for timely care if a heart attack is suspected. Education is essential regarding the need to call 9-1-1, provide transportation by EMS to the nearest ED, initiate early assessment and management of suspected ACS, including transmittal of prehospital ECGs,² and intervene if complications occur en route to the ED.³ The ACC’s Early Heart Attack Care guide is a resource to help educate the public about early recognition of potential cardiac symptoms and the importance of activating 9-1-1 for transportation.^4,5

The goals in patients presenting to the ED or office with acute chest pain are: 1) identify life-threatening causes; 2) determine clinical stability; and 3) assess need for hospitalization versus safety of outpatient evaluation and management. These concerns entail consideration of the full extent of clinical data. The ACC/AHA STEMI and NSTE-ACS guidelines categorize chest pain cause into 4 types: STEMI, NSTE-ACS, stable angina, and noncardiac.^6,7 The 12-lead ECG, which should be acquired and interpreted within 10 minutes of arrival to a medical facility^1-7,11 (Section 2.3.2, ECG), is pivotal in the evaluation because of its capacity to identify and triage patients with STEMI to urgent coronary reperfusion. Other ST-T abnormalities consistent with possible ischemia also mandate prompt evaluation in a hospital setting. In both cases, transfer should be by EMS; personal automobile for this purpose is associated with increased risk and should be avoided.^3-5 Patients with stable angina or noncardiac chest pain that is not life-threatening should be managed as outpatients.

Patients with chest pain and new ST-elevation, ST depression, or new left bundle branch block on ECG should be treated according to STEMI and NSTE-ACS guidelines.^1,2,6 An initial normal ECG does not exclude ACS. Patients with an initial normal ECG should have a repeat ECG, if symptoms are ongoing, until other diagnostic testing rules out ACS. An ECG may identify other nonischemic causes of chest pain (eg, pericarditis, myocarditis, arrhythmia, electrolyte abnormalities, paced rhythm, hypertrophic cardiomyopathy, pulmonary hypertension, congenital long QT, or normal variant). Figure 4 depicts an algorithm for the role of the ECG to help direct care for individuals presenting with chest pain or chest pain equivalents.

Chest radiographs are rapid, noninvasive tests that can be used to screen for several disorders that may present with chest pain. The yield of chest radiography depends on the pretest probability and will thus be higher when history or physical examination point to a greater likelihood of a given diagnosis. However, chest radiographs often do not lead to a diagnosis that requires intervention,¹ and their use should be guided by clinical suspicion.

Cardiovascular biomarkers can be useful for the diagnostic and prognostic assessment of patients with chest pain. Their most important application in clinical practice is for the rapid identification or exclusion of myocardial injury. The preferred biomarker to detect or exclude myocardial injury is cTn (I or T) because of its high sensitivity and specificity for myocardial tissue.^1-21,33 hs-cTn is preferred and can detect circulating cTn in the blood of most “healthy” individuals, with different sex-specific thresholds.^17,21,34 cTn is organ-specific but not disease-specific. Numerous ischemic, noncoronary cardiac, and noncardiac causes of cardiomyocyte injury can result in elevated cTn concentrations.^17,21,24,25 Therefore, interpretation of cTn results requires integration with all clinical information.^17,21

Although multiple other cardiovascular biomarkers, including some in common clinical use such as natriuretic peptides, have been shown to be associated with the risk of adverse cardiovascular outcomes in patients with chest pain, none have sufficient diagnostic accuracy for myocardial injury to be recommended for that purpose. The use of D-dimer for diagnosis of PE is discussed in Section 4.2.2.

CCTA can visualize and help to diagnose the extent and severity of nonobstructive and obstructive CAD, as well as atherosclerotic plaque composition and high-risk features (eg, positive remodeling, low attenuation plaque).^1-8 Calculation of fractional flow reserve with CT (FFR-CT) provides an estimation of lesion-specific ischemia.⁹ Current radiation dosimetry is low for CCTA, with effective doses for most patients in the 3 to 5 mSv range.¹⁰ CCTA contraindications are reported in Table 5 Although in select situations imaging protocols that evaluate the coronary arteries, aorta, and pulmonary arteries may be useful, the general approach should be to use imaging protocols tailored to the most likely diagnosis, rather than a “triple rule out” approach (Figure 6).

Invasive coronary angiography (ICA) defines the presence and severity of a luminal obstruction of an epicardial coronary artery, including its location, length, and diameter, as well as coronary blood flow.^1,2 For ICA, the primary goal is the characterization and detection of a high-grade obstructive stenosis to define feasibility and necessity of percutaneous or surgical revascularization. The use of physiologic indices (IFR and FFR) provides complementary functional information.¹ Radiation exposure to the patient during an interventional procedure averages 4 to 10 mSv and is dependent on procedural duration and complexity.^3,4

ICA has a spatial resolution of 0.3 mm; as such, it is impossible to visualize arterioles (diameter of 0.1 mm) that regulate myocardial blood flow.⁵ Coronary vascular functional studies can be performed during coronary angiography. Normal angiography does not exclude abnormal coronary vascular function, and it is possible to assess coronary microcirculation and coronary vasomotion. Coronary function testing may assist in management of the underlying condition, in addition to providing prognostic information.^6-8

Symptom-limited exercise ECG involves graded exercise until physical fatigue, limiting chest pain (or discomfort), marked ischemia, or a drop in blood pressure occurs.¹ Candidates for exercise ECG are those: a) without disabling comorbidity (eg, frailty, marked obesity [body mass index >40 kg/m²], peripheral artery disease, chronic obstructive pulmonary disease, or orthopedic limitations) and capable of performing activities of daily living or able to achieve ≥5 metabolic equivalents of exercise (METs)²; and b) without rest ST-T abnormalities (eg, >0.5-mm ST depression, left ventricular hypertrophy, paced rhythm, left bundle branch block, Wolff-Parkinson-White pattern, or digitalis use). Exercise electrocardiographic contraindications are reported in Table 5.

Transthoracic echocardiography (TTE) can visualize and aid in the differential diagnosis among the numerous causes of acute chest pain such as acute aortic dissection, pericardial effusion, stress cardiomyopathy, and hypertrophic cardiomyopathy.^1,2 Although TTE does provide information, for patients with acute chest pain, visualization of left and right ventricular function and regional wall motion abnormalities allows for the assessment of CAD risk and may help to guide clinical decision-making. Performance of TTE at the bedside is ideal for patients with acute chest pain and can be done using point-of-care or handheld devices in institutions where such capabilities are available.

After ACS has been ruled out, stress echocardiography can be used to define ischemia severity and for risk stratification purposes. For TTE and stress echocardiography, ultrasound-enhancing agents are helpful for left ventricular opacification when ≥2 contiguous segments or a coronary territory is poorly visualized.³ Coronary flow velocity reserve in the mid-distal left anterior descending coronary artery has been shown to improve risk stratification and may be helpful in the select patient with known CAD, including nonobstructive CAD.^4-6 Contraindications to stress type (exercise versus pharmacologic) and stress echocardiography are reported in Table 5.

After ACS has been ruled out, rest/stress positron emission tomography (PET) or single-photon emission computed tomography (SPECT) myocardial perfusion imaging (MPI) allows for detection of perfusion abnormalities, measures of left ventricular function, and high-risk findings, such as transient ischemic dilation.^1-8 For PET, calculation of myocardial blood flow reserve (MBFR, the ratio of peak hyperemia to resting myocardial blood flow) adds diagnostic and prognostic information over MPI data.^9-14 Radiation exposure, as reported by an average effective dose, is ∼3 mSv for rest/stress PET with Rb-82 and ∼10 mSv for Tc-99m SPECT; dual-isotope SPECT using thallium is not recommended.^15-17 SPECT/PET contraindications are and contraindications to type of stress test (exercise versus pharmacologic) are reported in Table 5.

Cardiovascular magnetic resonance (CMR) imaging has the capability to accurately assess global and regional left and right ventricular function, detect and localize myocardial ischemia and infarction, and determine myocardial viability. CMR can also detect myocardial edema and microvascular obstruction, which can help differentiate acute versus chronic MI, as well as other causes of acute chest pain, including myocarditis. CMR contraindications are reported in Table 5.

Low-risk patients are those with symptoms suggestive of ACS and whose probability of MACE within 30 days is ≤1%.¹⁷ This estimate is based on clinical information that is readily available during the course of evaluation, typically occurring in the ED. There are several methods to determine that a patient is low risk (Table 8) but, invariably, all involve taking an appropriate history and physical examination, demonstration that the ECG is normal, nonischemic, or unchanged from the previous ECG, and cTn measurement at a single point in time (if presentation is >3 hours from symptom onset and using a high-sensitivity assay) or serially^1-11 (with incorporation of a chest pain risk score into the CDP if using a conventional cTn assay). Importantly, there is no evidence to support routine admission or cardiac testing for chest pain patients who are low risk, although outpatient CAC scanning can provide additional information for longer-term risk stratification.

Patients in the ED without high-risk features and not classified as low risk by a CDP fall into an intermediate-risk group. Intermediate-risk patients do not have evidence of acute myocardial injury by troponin but remain candidates for additional cardiac testing. Some may have chronic or minor troponin elevations. This testing often requires more time than is appropriate for an ED visit. These patients may be placed in an inpatient bed or managed in a dedicated observation unit using a chest pain protocol.

Patients with symptoms suggestive of ACS who are at high risk of short-term MACE include those with new ischemic changes on the ECG, troponin-confirmed acute myocardial injury, new-onset left ventricular systolic dysfunction (ejection fraction <40%), newly diagnosed moderate-severe ischemia on stress imaging, and/or a high risk score on CDP.^4,13,14 ICA is indicated for patients with confirmed ACS based on a robust body of randomized trial evidence and clinical practice guideline indications.^4-7 In the patients with a negative initial evaluation, ICA is also indicated for those categorized as high risk on a validated risk stratification instrument.

For high-risk patients with a documented AMI on the index ED evaluation and who demonstrate on CCTA or ICA normal or nonobstructive CAD, CMR and echocardiography are useful for examining alternative causes for symptoms such as nonischemic cardiomyopathy or myocarditis.^8-11

There are many potential causes of acute chest pain in the months after CABG. Musculoskeletal pain from sternotomy remains the most common. However, other causes such as myocardial ischemia from acute graft stenosis or occlusion,^1,2 pericarditis, PE, sternal wound infection, or nonunion should also be considered. Post-sternotomy pain syndrome is defined as discomfort after thoracic surgery, persisting for at least 2 months, and without apparent cause.¹⁵ The incidence of post-sternotomy pain syndrome has been found to be as low as 7% and as high as 66%,^16-19 with a higher prevalence in women compared with men within the first 3 months of thoracic surgery (51.4% versus 31.3%; P<0.01) but, after 3 months, postoperative sex difference in prevalence was not seen.²⁰ Graft failure within the first year post-CABG using saphenous venous grafts is usually a result of technical issues, intimal hyperplasia, or thrombosis.⁵ Internal mammary artery graft failure within the first-year post-CABG is most commonly attributable to issues with the anastomotic site of the graft.

Reasons for acute chest pain several years after CABG include either graft stenosis or occlusion or progression of disease in a non-bypassed vessel. One year after CABG, ∼10% to 20% of saphenous vein grafts fail, while by 10 years, only about half of saphenous vein grafts are patent.⁵ In contrast, the internal mammary artery has patency rates of 90% to 95% 10 to 15 years after CABG.⁶ Compared with the use of saphenous vein grafts, the use of radial artery grafts for CABG also resulted in a higher rate of patency at 5 years of follow-up.⁷ In addition, knowledge of the native coronary anatomy and type of revascularization (complete or incomplete) is useful for interpretation of functional testing.

In 2015, there were nearly 500 000 people in the United States who received maintenance dialysis to treat end-stage renal disease.¹ Chest pain occurs during hemodialysis in 2% to 5% of patients.^6,7 Causes are numerous and related to the high prevalence of severe cardiovascular disease in this population and the dialysis procedure itself. Causes include AMI or ACS, pericarditis, PE, pleuritis, hemolysis, gastroesophageal reflux, subclavian steal, and musculoskeletal disorders.⁷ Myocardial ischemia is the most frequent serious cause and can be induced by hypotension^6,7 or tachyarrhythmias² occurring during dialysis in patients with CAD. AMI in patients undergoing dialysis is less frequently associated with chest pain than in patients who are not on dialysis, but warning signs may include diaphoresis or dyspnea.³ Unusual but serious causes of chest pain during dialysis are embolism⁶ and vessel perforation by catheter.^4,5 When indicated, cardiac testing for patients on dialysis should be the same as those who are not on dialysis.

The most frequent presenting complaint of cocaine abuse is acute chest pain, resulting from ≥1 of the alkaloid’s many cardiovascular actions.^1,4,5 Cocaine produces a hyperadrenergic state by blocking neuronal reuptake of norepinephrine and dopamine. The accumulation of these catecholamines increases heart rate and blood pressure, sometimes dramatically. These actions and the drug’s simultaneous effect of coronary vasoconstriction and elevated myocardial oxygen demand can produce myocardial ischemia and even infarction in the absence of obstructive CAD. Additional hazardous actions include increased myocardial contractility, cardiac arrhythmias, myocardial toxicity directly or through augmented adrenergic stimulation, increased platelet aggregability, endothelial dysfunction, and hypertensive vascular catastrophes (aortic dissection, cerebrovascular hemorrhage).^4-6

Methamphetamine has also been shown to lead to myocardial ischemia from mechanisms similar to cocaine. Studies have shown that methamphetamine can result in decreased myocardial perfusion. Like cocaine, methamphetamine also may reduce coronary sinus blood flow.⁷ It has been reported that up to 70% of methamphetamine users have an abnormal ECG, with the most common finding being tachycardia.⁸ Additional abnormalities on the ECG have been attributed to presence of hypertension, pulmonary artery hypertension, and cardiomyopathy, all of which have been associated with methamphetamine use.⁹ General principles for risk stratification of patients with chest pain apply to patients with cocaine or methamphetamine use.⁴

Risk communication and shared decision-making using a decision aid such as Chest Pain Choice have been shown to increase patient knowledge, engagement, and satisfaction and decrease the rate of observation unit admission and 30-day cardiac stress testing in both single-center and multicenter randomized trials.^1-3 For low-risk patients, decision aids can facilitate risk communication between the clinician and the patient and increase patients’ understanding of their risk and the importance of outpatient follow-up after discharge from the ED. For intermediate-risk patients, admission to an observation unit or discharge from the ED with further, timely evaluation in an outpatient setting is acceptable. Decision aids such as Chest Pain Choice can effectively facilitate shared decision-making regarding the need for admission, observation, or discharge for further evaluation in an outpatient setting.³

Acute aortic syndrome describes diseases involving disruptions in the aortic wall, including aortic dissection, intramural hematoma, and penetrating aortic ulcer.¹ The annual incidence is 2 to 4 cases/100 000, with higher prevalence with genetic conditions that weaken the aortic wall.² Prominent risk factors include hypertension atherosclerosis and connective tissue disease. Cocaine use may provoke dissection even without other risk factors.

Acute onset of severe chest or back pain heralds acute aortic dissection in 80% to 90% of patients, sometimes characterized as ripping or tearing.³ Progression can produce end-organ hypoperfusion, and proximal extension may cause tamponade, severe acute aortic regurgitation, or rarely, STEMI. Intramural hematomas, which differ from dissection by absence of an identifiable intimal flap, have a lesser understood natural history but are typically evaluated and treated in a similar manner to dissections.

The incidence of PE is estimated at 65 cases per 100 000, but some cases are asymptomatic and others undiagnosed.^5,6 One-third of deaths are sudden, and 60% are undiagnosed before death.⁷ Risk factors for PE are the same for venous thromboembolism and include inherited hypercoagulable states and acquired risk factors (recent surgery, trauma, immobilization, malignancy, smoking, obesity, oral contraception). Recognition of PE can be challenging because symptoms and clinical signs may be nonspecific. Dyspnea followed by chest pain, classically pleuritic, is the most common presenting symptom.¹ Signs of deep venous thrombosis may be present on examination.⁵

Pericarditis and myocarditis share overlapping common causes and likely form a continuum.⁸ In patients with pericarditis, a minimally elevated troponin does not appear to confer a worse prognosis.¹⁴ Most cases of pericarditis in developed nations are viral, although tuberculosis is sometimes a consideration.

Pericarditis classically presents with chest pain that is sharp, pleuritic, and which may be improved by sitting up or leaning forward, although in many instances such findings are not present. A pericardial friction rub may be audible. Widespread ST-elevation with PR depression is the electrocardiographic hallmark, although changes are nonspecific and may be transient.

Clinical manifestations of myocarditis are varied and include chest pain that is often sharp and reflective of epicardial inflammation involving the pericardium. Myocardial dysfunction often causes fatigue and exercise intolerance, and predominance of heart failure distinguishes myocarditis from pericarditis. Troponin is usually elevated.¹⁵

Chest pain may occur in the presence of VHD, particularly stenotic VHD such as aortic valve stenosis and mitral valve stenosis with secondary pulmonary hypertension. Chest pain may also occur after papillary muscle rupture in the setting of MI or in acute degenerative mitral valve disease after spontaneous chordal rupture. Chest pain may also occur in the setting of acute severe aortic insufficiency, which may be related to acute aortic pathology such as an aortic dissection manifesting as severe acute chest pain that may radiate to the back.

The cause of chest pain in patients with aortic valve stenosis may be secondary to coexisting obstructive epicardial CAD¹ or, more commonly, chest pain may occur as a result of coronary microvascular dysfunction² in the presence of very elevated left ventricular pressure caused by a high left ventricular afterload, along with the associated left ventricular hypertrophy. The cause of chest pain in patients with severe mitral valve stenosis is more likely to be secondary to epicardial obstructive CAD¹ although, less likely, chest pain may occur in isolated mitral valve stenosis resulting from low cardiac output and decreased coronary perfusion.¹

Among outpatients who present with chest pain, approximately 10% to 20% have a gastrointestinal cause.¹ Gastrointestinal pain may result from stimulation of chemoreceptors by acid or hyperosmolar substances, of mechanoreceptors by abnormal contraction or distention, or of thermoreceptors.² Some patients have abnormal perceptions of otherwise normal stimuli. Gastroesophageal reflux disease is the most likely cause for recurring unexplained chest pain of esophageal origin.³ Chest pain caused by gastroesophageal reflux disease can mimic myocardial ischemia and may be described as squeezing or burning. The duration can be minutes to hours, often occurs after meals or at night, and can worsen with stress. Depending on the severity, it may or may not resolve spontaneously or with antacids. Esophagitis not related to reflux may be caused by medications, underlying infections such as candidiasis, or radiation injury. Allergic conditions are associated with eosinophilic esophagitis, which is diagnosed by biopsy. Esophageal motility disorders such as achalasia, distal esophageal spasm, and nutcracker esophagus are less common but can present as squeezing retrosternal pain or spasm, often accompanied by dysphagia.

Although the heart-brain relationship is well established,^15-17 its clinical relevance has been enhanced by recognition of stress cardiomyopathy.^18,19 Less dramatic than the latter syndrome but highly prevalent is recurrent chest pain despite angiographically normal coronary arteries and no definable cardiac disease, including an assessment for INOCA.^1-14 Chest pain in these patients has been variously labeled angina, angina-like, “atypical” angina, or noncardiac chest pain based on its deviation from characteristic ischemic cardiac discomfort. Prognosis of patients with noncardiac chest pain is largely devoid of cardiac complications.^4,9,20-23 The close association of this symptom with psychological syndromes such as anxiety, panic attack, depression, somatoform disorder, and cardiophobia suggests that there may be a psychogenic origin in many patients. These factors have also raised consideration of mechanisms for noncardiac chest pain such as central nervous system-visceral interactions, low pain thresholds, hyperbody vigilance, sympathetic activation, as well as anxiety, depression, and panic disorder.^{6,7,9,14,23-30} It has been reported that these patients undergo extensive and repetitive cardiac testing and have low referral to cognitive-behavioral therapists, suggesting a lost opportunity for pharmacologic or cognitive-behavioral therapy.⁶

Acute chest syndrome is a leading cause of death for patients with sickle cell disease.^1,2 Patients with sickle cell disease who are experiencing chest pain require prompt evaluation.³ Although chest pain occurs in most, other manifestations of acute chest syndrome include shortness of breath, fever, arm and leg pain, and the presence of a new density on chest radiography. Older adolescents and adults with sickle cell disease who present with chest pain and shortness of breath should be evaluated for AMI or myocardial ischemia.⁴ AMI occurs in patients with sickle cell disease at a relatively early age, usually without the traditional risk factors for ACS. Death from ACS in patients with sickle cell disease is significantly high in age-, sex-, and race-matched controls.⁵

In the evaluation of symptomatic patients with suspected CAD, use of validated scores to predict the pretest probability of obstructive CAD may be useful to identify low-risk patients for whom testing may be deferred. It is preferable to use contemporary estimates such as those published in the past 10 years, such as the pretest probability proposed by Juarez-Orozco et al.¹ in preference to scores from historical patient series, which may overestimate the frequency of obstructive CAD. Alternatively, low-risk patients may be those <40 years of age or who have symptoms that have a low likelihood of representing ischemia (Section 5.1.2). When available, information on the presence and amount of CAC may be useful for enhancing the pretest probability of obstructive CAD, as shown in Figure 11.² This information can be obtained from performing a CAC scan or, when available, from a visual estimation of CAC based on prior noncardiac chest CT. Among the remaining patients classified as intermediate-high risk, selective testing may improve diagnosis of CAD and for risk stratification purposes.^1-5

Over the past several decades, patient presentation and observed obstructive CAD prevalence has changed, thus affecting patient selection for diagnostic testing. Current observations in the United States include:

For the aforementioned reasons, use of a contemporary pretest probability estimates to define low-risk patients not requiring additional diagnostic testing is a primary goal of the initial evaluation of symptomatic patients with suspected CAD.⁵ Even when contemporary pretest probability estimates are used, they have a low specificity for identifying patients with obstructive CAD. A CAC score of zero can be useful to identify patients with stable chest pain who are low risk, have a low likelihood of obstructive CAD, and a low risk of future cardiovascular events.⁷ Additionally, exercise testing without imaging is also reasonable to perform in low-risk individuals with stable chest pain and no known CAD to exclude myocardial ischemia and assess functional capacity¹⁰ (Figure 12).

The approach to the diagnostic evaluation of patients with no known CAD who are at intermediate to high risk (Figure 12) should be guided by the ability to achieve high-quality imaging as well as local availability and expertise. Intermediate-high risk patients have modest rates of obstructive CAD (∼10%–20%) and risk of clinical events (∼1%–2% per year).^1,5,8,77-80 CCTA is preferable in those <65 years of age and not on optimal preventive therapies, while stress testing may be advantageous in those ≥65 years of age, because they have a higher likelihood of ischemia and obstructive CAD.^34-36,81-83 Although previous guidelines supported direct referral to ICA among patients with stable chest pain, contemporary randomized trials support that candidates for elective coronary angiography may be safely triaged using CCTA^1,84 or noninvasive stress testing.^34,35

Patient characteristics and existing contraindications for a given test modality (Tables 5 and 6) should be considered when choosing a diagnostic test. Imaging of obese patients, especially those with morbid obesity (body mass index >40), can be challenging and requires careful consideration of available equipment. In obese patients, contrast enhancement is useful to improve imaging quality. In certain patients, it may be important to undergo exercise testing so to collect data on the hemodynamic or symptomatic response to exercise. In patients selected for stress imaging who are able to exercise, exercise testing is preferred over pharmacologic stress to improve the diagnostic and prognostic information of the test. Although PET and SPECT are grouped together, PET has improved diagnostic and prognostic performance, especially when quantitative assessment of MBF can be performed.^36-39

Irrespective of the test performed, an overarching goal of the evaluation of symptomatic patients is to identify those who would benefit from GDMT, as defined by the 2014 SIHD guidelines, the 2018 cholesterol-lowering guidelines, and the 2019 prevention guidelines.^13,85-87 For this evaluation, the patient should be engaged in a process of shared decision-making before determining the final choice of the cardiac test modality and in guiding the pathway for treatment decisions.

Randomized trials comparing the effectiveness of CCTA versus stress testing report similar near-term effectiveness (at ∼2–3 years of follow-up).^7,8,10-12,89 In the SCOT-HEART (Scottish Computed Tomography of the Heart) trial, the addition of CCTA to standard of care resulted in a reduction in 5-year CAD death or AMI when compared with standard care alone (predominantly exercise ECG) (HR: 0.59; 95% CI: 0.41-0.84; P=0.004).⁹ From a prespecified analysis from the PROMISE trial, patients with diabetes who underwent CCTA had a lower risk of cardiovascular death or MI when compared with those randomized to stress testing (adjusted HR: 0.38; 95% CI: 0.18-0.79; P=0.01).⁶ Especially for patients with nonobstructive and obstructive CAD, CCTA more often prompts initiation and intensification of preventive and anti-ischemic therapies than other diagnostic strategies.^6,89-96 Several randomized trials compared the effectiveness of CCTA versus direct referral to ICA among symptomatic patients.^1,5 From the CONSERVE trial, a strategy of initial CCTA was associated with lower cost but similar 1-year MACE rates (death, ACS, stroke, urgent/emergency coronary revascularization, or cardiac hospitalization) as direct ICA (4.6% versus 4.6%).⁵

In patients with known CAD, physicians should opt to intensify GDMT first, if there is an opportunity to do so, and defer testing. In patients with a history of obstructive CAD, previous AMI, or previous coronary revascularization, assessing the severity of ischemia may be useful to guide clinical decision-making regarding the use of ICA and intensify preventive and anti-ischemic therapy. Imaging should be considered in those with new onset or persistent stable chest pain (Figure 13). In patients with frequent angina or severe stress-induced ischemia, referral to ICA or CCTA is an option.⁴ Among individuals with known obstructive CAD or ischemic heart disease who have stable symptoms, exercise treadmill testing may be useful for assessing functional capacity, assessing the type and severity of symptoms, and informing the role of coronary revascularization, cardiac rehabilitation, or anti-anginal therapy.^4,38-40

Clinical trials of CMR have included subgroups with obstructive CAD, including 76% and 49% in the MR-IMPACT and MR-IMPACT2 studies, respectively, showing generally comparable diagnostic accuracy to stress SPECT MPI.^23,24 Several large, multicenter registries reveal that stress CMR effectively risk stratifies patients with known CAD.^27-30 In a multicenter registry of 2496 patients with a history of CAD, an abnormal stress CMR had a nearly 2-fold increased mortality hazard.²⁷ From the SPINS Registry (Stress CMR Perfusion Imaging in the United States), patients with known CAD with MPI ischemia and scarring by late gadolinium enhancement had a relative hazard of 1.5 to 2.1 for CV death or nonfatal MI.³⁰ Prognosis worsens for patients by the extent and severity of inducible wall motion abnormalities on stress echocardiography.^66,67 Recent randomized trial evidence supports the role of stress echocardiography to guide clinical decision-making. From the ORBITA (Objective Randomized Blinded Investigation With Optimal Medical Therapy in Stable Angina) trial, there was a greater reduction in the stress echocardiographic wall motion score among patients with single-vessel CAD treated with PCI compared with placebo (P<0.0001).⁶⁸ In a secondary analysis, there was an interaction between the baseline stress echocardiographic wall motion score and the efficacy of PCI for improved angina at 6 weeks of follow-up.⁶⁹ That is, PCI-treated patients with a wall motion score ≥1 were more often angina-free compared with those in the placebo arm.

For patients with known nonobstructive CAD (luminal narrowing 1%–49%), CCTA can be useful for detection of new or worsening obstructive stenosis, atherosclerotic disease progression, and identification of high-risk plaque features, such as low attenuation plaque or positive remodeling^1,2,5-7,25 (Figure 13). Similarly, stress imaging is reasonable to detect myocardial ischemia and can help guide further management and treatment of ischemic burden.^15-24

Irrespective of the test performed, an overarching goal of the evaluation of symptomatic patients with known nonobstructive CAD is to identify those who would benefit from intensification of preventive therapy, as defined by the 2018 cholesterol-lowering guidelines and the 2019 prevention guidelines.^26-29 For this evaluation, the patient should be engaged in a process of shared decision-making before determining the final choice of the cardiac testing modality and in guiding the pathway for treatment decisions.

Recommendations for myocardial blood flow measurements using PET, echocardiography, and CMR are found in Section 5.2.2.

Signs and symptoms of ischemia occur because of focal obstructive CAD, but INOCA is common and may result from alterations in flow within the microvasculature. Thus, many symptomatic patients without obstructive CAD on previous workup may be candidates for assessment of coronary microvascular dysfunction and other causes of INOCA.¹ Patients at highest risk for coronary microvascular dysfunction include women, those with hypertension, diabetes, and other insulin-resistant states.¹⁵ There is substantive evidence that testing focusing on documentation of coronary or microvascular flow abnormalities can aid in the diagnosis of microvascular angina, and abundant evidence supports that the addition of flow alterations improves risk stratification. Invasive coronary reactivity testing allows for the assessment of vasospasm, in addition to nonendothelial–dependent and endothelium-dependent microvascular reactivity.^2,4 From the National Institutes of Health-NHLBI-sponsored WISE (Women’s Ischemia Syndrome Evaluation), impaired coronary flow reserve (ie, <2.32) among women with no obstructive CAD was associated with an elevated hazard for major CAD events with lengthy follow-up of 10 years (P=0.03).² Among women with no obstructive CAD, epicardial vasoconstriction was also significantly associated with higher rates of hospitalization for angina (P=0.0002).² Prognostic evidence is available supporting the novel contribution of PET MBFR techniques; several reports also note a benefit using CMR and echocardiographic techniques. A proposed diagnostic evaluation pathway is outlined in Figure 14.

In the outpatient setting, long-term costs were generally similar between CCTA and stress testing strategies.¹ Higher invasive angiography rates after CCTA are matched by a greater use of downstream stress testing after initial stress testing, resulting in minimal differences in cost at 2 to 3 years of follow-up.^1,2 From the CONSERVE (Coronary Computed Tomographic Angiography for Selective Cardiac Catheterization) trial, 823 patients were randomized to a selective versus direct referral strategy to ICA. Enrollment was limited to patients with nonemergent indications for ICA.³ The selective referral arm included CCTA-guided use of ICA. Cumulative diagnostic costs were $1183 for the selective arm and $2755 for the direct referral arm of the CONSERVE trial (57% lower costs). In the CCTA-guided arm, follow-up stress testing was applied and contributed to reduced referrals to ICA.

A recent tiered testing strategy was evaluated in both the CRESCENT I and II trials.^2,4 From the CRESCENT I trial, CAC was used as the index test, with follow-up CCTA used only in patients with detectable CAC or for those with a high pretest risk.² In this trial, nearly 40% of patients did not undergo CCTA, which reduced diagnostic evaluation costs; no events were reported in this subgroup. By comparison, nearly half of those randomized to the exercise ECG had additional confirmatory diagnostic testing. Overall, 1-year costs were significantly lower in the CAC tiered testing protocol (16% cost savings; P<0.0001).² Moreover, 1-year MACE-free survival was higher in the CAC-guided testing arm (97%) compared with exercise ECG (90%; HR: 0.32; P=0.011).

The economic evidence for the exercise ECG supports that tiered testing may offset its reduced diagnostic accuracy.^1,2 In a decision model, tiered testing of exercise ECG followed by selective stress echocardiography resulted in improved diagnostic accuracy and favorable cost-effective ratios when compared with other testing strategies.² In a Medicare cohort, observed 180-day costs were lowest for the exercise ECG when compared with stress echocardiography, MPI, or CCTA.³ Randomized trial data on cost are available and from the PROMISE trial initial test costs were $174 for exercise ECG, >50% lower than that of other imaging procedures.⁴ At 3-years of follow-up, the mean cost difference was $1731 higher for CCTA (n=4818) when compared with the exercise ECG (n=858); however, the 95% CIs for cost differences was wide ($2 to $3519), and there was no overall difference by randomization.⁴ Overall, results from the PROMISE trial showed that stress testing was associated with similar costs and CAD outcomes over ∼3 years of follow-up.^4,5 In a randomized trial of 824 symptomatic women, cumulative procedural costs were nearly 50% lower for exercise ECG versus MPI SPECT (P<0.0001), with no difference in 2-year event-free survival (P=0.59).⁶

Several cost-effectiveness models have reported an increased incremental cost-effectiveness ratio for stress echocardiography, compared with exercise ECG and other diagnostic procedures.^1-4 In these models, cost-effectiveness was influenced by an improved diagnostic accuracy for stress echocardiography, which led to longer life expectancy.¹ In a recent systematic review, the evidence supports that stress echocardiography or stress MPI are cost-effective for those patients at intermediate pretest risk.⁵ The improved CAD detection for exercise echocardiography resulted in fewer office and ED visits and hospital days, yielding a 20% cost savings when compared with the exercise ECG.⁶ There were 2204 patients that underwent stress echocardiography in the PROMISE trial, and 3-year mean costs were similar to that of CCTA (CCTA – stress echocardiography mean cost difference: –$363; 95% CI: –$1562 to $818).⁷

Among intermediate-risk patients, evidence synthesis supports that stress MPI is a cost-effective test option.¹ From the SPARC (Study of Myocardial Perfusion and Coronary Anatomy Imaging Roles in Coronary Artery Disease) registry, observed 2-year mortality rate was highest for PET MPI (5.5%) compared with CCTA (0.7%) or SPECT MPI (1.6%), with 2-year cost highest for patients undergoing PET.² In the PROMISE trial, nearly two-thirds of patients underwent stress MPI, and the mean cost was similar when compared with CCTA.³ Mean costs were also similar in a randomized trial of 457 patients comparing stress MPI with exercise ECG (P>0.05).⁴ Among higher-likelihood patients in the UK enrolled in the CECaT (Cost Effectiveness of Noninvasive Cardiac Testing) trial, MPI SPECT was the most cost-effective approach.⁵

A synthesis of this cost evidence reveals a pattern whereby CMR perfusion and scar imaging is associated with a favorable incremental cost-effectiveness ratio of <$50 000 per quality-adjusted life years saved.^1,2 From a single report, a CMR strategy informed by the CE-MARC trial was more cost-effective than stress MPI, largely because of the higher diagnostic accuracy for CMR.² However, the most cost-effective strategy was that of initial exercise ECG followed by selective stress CMR and invasive angiography; for this tiered testing approach, additional testing was deemed appropriate in the setting of abnormal or inconclusive findings. In a decision model for intermediate pretest risk patients, a strategy of CMR followed by selective ICA had projected reduced costs by ∼25% when compared with direct referral to ICA.^2,3 From the Stress CMR Perfusion Imaging in the United States registry,⁴ patients with negative findings for ischemia and scar had low downstream costs.⁵