Table of Contents Print a PDF References Figures & Tables
< Previous Next >


Antman et al., Management of Patients With STEMI: Executive Summary
J Am Coll Cardiol 2004;44:671-719

ACC/AHA Guidelines for the Management of Patients With ST-Elevation Myocardial Infarction—Executive Summary

A Report of the American College of Cardiology/American Heart Association Task Force on Practice Guidelines (Writing Committee to Revise the 1999 Guidelines for the Management of Patients With Acute Myocardial Infarction)

Developed in Collaboration With the Canadian Cardiovascular Society


6. Initial Recognition And Management In The Emergency Department

A variety of treatment options (Figure 3) (24-40) are available that can reduce mortality and morbidity in patients with STEMI, but the effectiveness of these therapies diminishes rapidly within the first several hours after symptoms onset (162,185). The traditional ED evaluation of patients with chest pain relies heavily on the patient’s history, physical examination, and the ECG. This approach not infrequently fails to identify patients who are actually suffering from STEMI, which results in an inappropriate discharge home from the ED (186). Such missed MI patients are at relatively high risk of death or complications for the next 4 to 6 weeks after ED discharge (187-192).

In a large study on this subject, Pope et al. (188) found that 889 of 10 689 patients who presented to 10 US hospital EDs with chest pain or other symptoms that suggested acute cardiac ischemia had STEMI; 19 patients (2.1%, 95% CI 1.1% to 3.1%) were discharged from the ED. Patients with STEMI were more likely not to be hospitalized if they were nonwhite (odds ratio [OR] for discharge 4.5; 95% CI 1.8 to 11.8) or had a normal or nondiagnostic ECG (OR 7.7; 95% CI 2.9 to 20.2). The risk-adjusted mortality ratio for MI patients who were not hospitalized compared with those who were hospitalized was 1.9 (95% CI 0.7 to 5.2).

6.1. Optimal Strategies for Emergency Department Triage

Class I

Hospitals should establish multidisciplinary teams (including primary care physicians, emergency medicine physicians, cardiologists, nurses, and laboratorians) to develop guideline-based, institution-specific written protocols for triaging and managing patients who are seen in the prehospital setting or present to the ED with symptoms suggestive of STEMI. (Level of Evidence: B)

The advent of highly effective, time-dependent treatment for STEMI coupled with the need to reduce healthcare costs adds further incentive for clinicians to get the right answer quickly and to reduce unnecessary admissions and length of hospital stay. Investigators have tried various diagnostic tools such as clinical decision algorithms, cardiac biomarkers, echocardiography, and myocardial perfusion imaging in an attempt to avoid missing patients with MI or unstable angina. The most successful strategies to emerge thus far are designed to identify MI patients and, when clinically appropriate, screen for unstable angina and underlying coronary artery disease. Most strategies use a combination of cardiac biomarkers, short-term observation, diagnostic imaging, and provocative stress testing. An increasing number of high-quality centers now use structured protocols, checklists, or critical pathways to screen patients with suspected MI or unstable angina (193-205). It does not appear to matter whether the institution designates itself a chest pain center. Rather, it is the multifaceted, structured approach to the problem that appears to provide clinical, cost-effective benefit (206,207). One randomized trial has confirmed the safety, efficacy, and cost-effectiveness of the structured decisionmaking approach compared with standard, unstructured care (208).

6.2. Initial Patient Evaluation

Class I
1. The delay from patient contact with the healthcare system (arrival at the ED or contact with paramedics) to initiation of fibrinolytic therapy should be less than 30 minutes. Alternatively, if PCI is chosen, the delay from patient contact with the healthcare system (typically, arrival at the ED, or contact with paramedics) to balloon inflation should be less than 90 minutes. (Level of Evidence: B)

2. The choice of initial STEMI treatment should be made by the emergency medicine physician on duty based on a predetermined, institution-specific, written protocol that is a collaborative effort of cardiologists (both those involved in coronary care unit management and interventionalists), emergency physicians, primary care physicians, nurses, and other appropriate personnel. For cases in which the initial diagnosis and treatment plan are unclear to the emergency
physician or are not covered directly by the agreedupon protocol, immediate cardiology consultation is advisable. (Level of Evidence: C)

Regardless of the approach used, all patients presenting to the ED with chest discomfort or other symptoms suggestive of STEMI or unstable angina should be considered high-priority triage cases and should be evaluated and treated on the basis of a predetermined, institution-specific chest pain protocol. The protocol should include several diagnostic possibilities (Figure 9) (4). The patient should be placed on a cardiac monitor immediately, with emergency resuscitation equipment, including a defibrillator, nearby. An ECG should be performed and shown to an experienced emergency medicine physician within 10 minutes of ED arrival. If STEMI is present, the decision as to whether the patient will be treated with fibrinolytic therapy or primary PCI should be made within the next 10 minutes (Figure 7) (180). The goal for patients with STEMI should be to achieve a door-to-needle time of within 30 minutes and a door-to-balloon time of within 90 minutes (Figure 6) (155). If the initial ECG is not diagnostic, the patient remains symptomatic, and there is a high clinical suspicion for STEMI, serial ECGs at 5- to 10-minute intervals or continuous ST-segment monitoring should be performed.

Ideally, such decisions should be made by the emergency medicine physician on duty in the ED based on a predetermined, institution-specific, written protocol that has been developed with input from cardiologists (both those involved in coronary care unit management and interventionalists), emergency medicine physicians, primary care physicians, nurses, and other appropriate personnel. For non-interventional hospitals, this will usually require formal, written transfer agreements and protocols that will permit expeditious transfer of patients who require urgent mechanical
revascularization to the nearest appropriate interventional facility (Figure 6) (155). The protocol should also include the level of training and certification of personnel required to accompany the patient during transfer, the minimum equipment requirements, and the type(s) of transport vehicles (e.g., standard ground ambulance, mobile intensive care unit, helicopter, or fixed-wing aircraft) that can be used on the basis of patient condition. For cases in which the initial diagnosis and treatment plan are unclear to the emergency medicine physician or are not covered directly by the agreed-upon protocol, immediate cardiology consultation is advisable.

6.2.1. History

Class I
The targeted history of STEMI patients taken in the ED should ascertain whether the patient has had prior episodes of myocardial ischemia, such as stable or unstable angina, MI, coronary bypass surgery, or PCI. Evaluation of the patient’s complaints should focus on chest discomfort, associated symptoms, sex and age-related differences in presentation, hypertension, diabetes mellitus, possibility of aortic dissection, risk of bleeding, and clinical cerebrovascular disease (amaurosis fugax, face/limb weakness or clumsiness, face/limb numbness or sensory loss, ataxia, or vertigo). (Level of Evidence: C)

The history taken in the ED must be concise and detailed enough to establish the probability of STEMI but should be obtained expeditiously so as not to delay implementation of reperfusion therapy.

Chest Discomfort. The severity of discomfort varies and is typically graded on a scale of 1 to 10, with 10 being the mostsevere pain. It is important to keep in mind that many patients will not admit having chest “pain” but will acknowledge the presence of chest “discomfort,” because of their definition of pain. The chest discomfort is often described as a crushing, vice-like constriction, a feeling equivalent to an “elephant sitting on the chest,” or heartburn. Usually, the discomfort is substernal but may originate in or radiate to areas such as the neck, jaw, interscapular area, upper extremities, and epigastrium. The duration of the discomfort, which typically lasts longer than 30 minutes, may wax and wane and may be remitting. It may be described as “indigestion in the chest” and occasionally may be relieved with belching. The possibility of precipitation of STEMI by use of illicit drugs such as cocaine should be considered.

The targeted history of patients with STEMI taken in the ED should ascertain whether the patient has had prior episodes of myocardial ischemia such as stable or unstable angina, MI, coronary bypass surgery, or PCI. Evaluation of the patient’s complaints should focus on chest discomfort, associated symptoms, sex- and age-related differences in presentation, hypertension, diabetes mellitus, possibility of aortic dissection, risk of bleeding, and clinical cerebrovascular disease (amaurosis fugax, face/limb weakness or clumsiness, face/limb numbness or sensory loss, ataxia, or vertigo).

Associated Symptoms. Other symptoms to be aware of when taking a patient’s history include nausea and vomiting. Diaphoresis associated with a pale complexion may also appear, as well as weakness or profound fatigue. Dizziness, lightheadedness, syncope, and paresthesia may occur because of pain and hyperventilation.

Hypertension
. Hypertension should be assessed, because chronic, severe, poorly controlled hypertension or severe uncontrolled hypertension on presentation is a relative contraindication to fibrinolytic therapy (see Section 6.3.1.6.3.2).

Sex- and Age-Related Differences in Presentation. It has been noted in studies that women present with STEMI at an older age and later after the onset of symptoms than men (53,210). There must be an elevated index of suspicion during the evaluation of women for STEMI. Although some variation exists, when large databases of MI patients are examined, symptom profiles for STEMI by sex generally appear more similar than different between men and women (211-215). Elderly patients with STEMI are significantly less likely than younger patients to complain of chest discomfort. However, elderly patients with STEMI are more likely to complain of shortness of breath, as well as other atypical symptoms such as syncope or unexplained nausea. (181).

Diabetes Mellitus. Diabetics may have impaired angina (pain) recognition, especially in the presence of autonomic neuropathy. A diabetic may misinterpret dyspnea, nausea, vomiting, fatigue, and diaphoresis as disturbance of diabetic control. Up to 50% of diabetic individuals with type 2 diabetes for longer than 10 years will have autonomic nervous system dysfunction manifested by impaired heart rate variability.
Diabetics with STEMI should be evaluated for renal dysfunction (216).

Possibility of Aortic Dissection. Severe tearing pain radiating directly to the back associated with dyspnea or syncope and without ECG changes indicative of STEMI should raise the suspicion for aortic dissection, and appropriate studies should be undertaken. Clinicians should have a heightened index of suspicion for aortic dissection in elderly hypertensive patients. However, it must be kept in mind that the dissection may extend to the pericardial sac and produce cardiac tamponade or disrupt the origin of a coronary artery.

Risk of Bleeding. Patients should be questioned about previous bleeding problems, e.g., during surgery or dental procedures, history of ulcer disease, cerebral vascular accidents, unexplained anemia, or melena. The use of antiplatelet, antithrombin, and fibrinolytic agents as part of the treatment for STEMI will exacerbate any underlying bleeding risks.

Clinical Cerebrovascular Disease. The patient with STEMI frequently has medical conditions that are risk factors for both MI and stroke. Evidence for prior episodes suggestive of clinical cerebrovascular disease should be sought. For example, the patient should be asked whether he/she has ever had symptoms of transient retinal or cerebral ischemia such as amaurosis fugax, face/limb weakness or clumsiness, face/limb numbness or sensory loss, ataxia, or vertigo. Transient ischemic attacks (TIAs) typically last less than 30 minutes, whereas symptoms that last more than 60 to 90 minutes are more likely to indicate the presence of a stroke (217). In addition, the patient should be asked whether he/she has ever had an ischemic stroke, intracerebral hemorrhage [ICH], or subarachnoid hemorrhage. A brief summary of the details for diagnosis of the different stroke subtypes is available (218). Finally, a history of cognitive decline/dementia may indicate the presence of cerebral amyloid angiopathy and increased risk of ICH, and information regarding head and facial trauma should be obtained.

6.2.2. Physical Examination

Class I

1. A physical examination should be performed to aid in the diagnosis and assessment of the extent, location, and presence of complications of STEMI. (Level of Evidence: C)

2. A brief, focused, and limited neurological examination to look for evidence of prior stroke or cognitive deficits should be performed on STEMI patients before administration of fibrinolytic therapy. (Level of Evidence: C)

A brief physical examination may promote rapid triage (Table 5), whereas a more detailed physical examination aids in the differential diagnosis and is useful for assessing the extent, location, and presence of complications of STEMI (Tables 6 and 7) (219).


Evidence of prior stroke or dementia may be suggested by the finding on examination of focal neurological or cognitive deficits (Table 6). A brief but focused examination can identify focal neurological or cognitive deficits.

6.2.2.1. Differential Diagnosis

The differential diagnosis of STEMI includes conditions that can be exacerbated by fibrinolysis and anticoagulation (Table 8). The pain of aortic dissection is typically described as searing, ripping, or tearing and frequently radiates to the back or lower extremities. The pain is worse at onset and lasts for hours. Major pulses may be absent, and a murmur of aortic regurgitation may be present. A transesophageal echocardiogram, computed tomography (CT) scan, or magnetic resonance imaging scan is useful for establishing the diagnosis of aortic dissection. Active peptic ulcer disease can be present with chest or epigastric pain, sometimes radiating posteriorly, and may be associated with syncope, hematemesis, or melena. Free subdiaphragmatic air may be seen on upright chest X-ray in perforations. Acute pericarditis may show PR-segment depression and ST-segment elevation on the ECG but without reciprocal ST-segment depression (220). Pain from pericarditis is usually pleuritic and can radiate to the shoulder and trapezius ridge and is often relieved by sitting up and leaning forward, characteristics not found in STEMI. A rub is often present. Pulmonary embolus, with or without infarction, presents with dyspnea and knifelike pleuritic pain, sometimes with hemoptysis. Pulmonary embolism can present with chest pain similar to that of STEMI. Costochondral pain is described as sharp or sticking, with associated localized tenderness. Pneumothorax may present with acute dyspnea, pleuritic pain, and differential decrease in breath sounds with hyperresonance over 1 lung field. Acute cholecystitis may mimic STEMI, and rightupper- quadrant abdominal tenderness should be sought on physical examination.

6.2.3. Electrocardiogram

Class I
1. A 12-lead ECG should be performed and shown to an experienced emergency physician within 10 minutes of ED arrival on all patients with chest discomfort (or
anginal equivalent) or other symptoms suggestive of STEMI. (Level of Evidence: C)

2. If the initial ECG is not diagnostic of STEMI but the patient remains symptomatic, and there is a high clinical suspicion for STEMI, serial ECGs at 5- to 10- minute intervals or continuous 12-lead ST-segment monitoring should be performed to detect the potential development of ST elevation. (Level of Evidence: C)

3. In patients with inferior STEMI, right-sided ECG leads should be obtained to screen for ST elevation suggestive of RV infarction. (See Section 7.6.6 and the ACC/AHA/ASE 2003 Guideline Update for the Clinical Application of Echocardiography). (Level of Evidence: B)


The 12-lead ECG in the ED is at the center of the therapeutic decision pathway because of the strong evidence that ST-segment elevation identifies patients who benefit from reperfusion therapy (221). Mortality increases with the number of ECG leads showing ST elevation. Important predictors of mortality on the initial 12-lead ECG include left bundle branch block (LBBB) and anterior location of infarction (Figure 10) (222,223). Diagnostic criteria of greater than 0.1 mV in leads V1 through V4 may have reduced specificity for STEMI in patients with early repolarization. Some evidence exists to support the use of greater than or equal to 0.2 mV anteroseptal elevation as a preferable threshold for diagnosing STEMI, because a higher proportion of patients are correctly classified as having STEMI than with a threshold of greater than 0.1 mV in these leads (221).

In the absence of ST elevation, there is no evidence of benefit of fibrinolytic therapy for patients with normal ECG or nonspecific changes, and there is some suggestion of harm (including increased bleeding risk) for patients with ST-segment depression only (221,224). Notwithstanding this, fibrinolytic therapy may be appropriate when there is marked STsegment depression confined to leads V1 through V4 and accompanied by tall R waves in the right precordial leads and upright T waves indicative of a true posterior injury current and circumflex coronary occlusion. In circumstances where there is a suggestive clinical history and suggestive evidence of true posterior infarction, confirmatory data from posterior leads (i.e., V7 and V8) as well as 2-dimensional echocardiography may be especially helpful; this latter investigation has a high negative predictive value (225,226). Primary PCI is another reperfusion strategy that may be effective in patients with true posterior MI (see Section 6.3.1.6.4.2).

Initial errors in ECG interpretation can result in up to 12% of patients being categorized inappropriately, demonstrating a potential benefit of accurate computer-interpreted electrocardiography and fax transmission to an expert (227). It is less likely that STEMI is present if the upward-directed STsegment changes are concave rather than convex (228). Because lethal ventricular arrhythmias may develop abruptly in patients with STEMI, all patients should be monitored electrocardiographically on arrival in the ED. It is important to examine serial tracings approximately 5 to 10 minutes apart, or if symptoms recur, during evaluation in the ED for development of ST elevation if the initial ECG is nondiagnostic. ST elevation may also be detected by intermittent visual inspection of the oscilloscope or auditory alarms in systems with continuous ST-segment monitoring capability.

Although the Fibrinolytic Therapy Trialists’ (FTT) Collaborative Group overview indicates that patients with new or presumably new LBBB are at high risk when presenting with presumed MI, this ECG presentation is a frequent cause of delay or lack of reperfusion therapy because of the concern of the validity of the ECG criteria for MI diagnosis and the risk of therapy. This is also a situation in which direct PCI may be preferable to fibrinolytic therapy (156). It has been suggested that patients with new or presumably new LBBB coupled with a typical ischemic history be approached with a plan to rule in MI using 1 of 3 ECG criteria that provide independent diagnostic value. These consist of ST elevation greater than or equal to 0.1 mV in leads with a positive QRS, ST depression greater than or equal to 0.1 mV in V1 to V3, and ST elevation greater than or equal to 0.5 mV in leads with a negative QRS (229,230).

6.2.4. Laboratory Examinations

Class I
Laboratory examinations should be performed as part of the management of STEMI patients but should not delay the implementation of reperfusion therapy. For specific laboratory examinations, see Table 9. (Level of Evidence: C)

In addition to serum cardiac biomarkers for cardiac damage, several routine evaluations have important implications for management of patients with STEMI (Table 9). Although these studies should be undertaken when the patient is first examined, therapeutic decisions should not be delayed until results are obtained because of the crucial role of time to therapy in STEMI.

6.2.5. Biomarkers of Cardiac Damage

Class I
1. Cardiac-specific troponins should be used as the optimum biomarkers for the evaluation of patients with STEMI who have coexistent skeletal muscle injury. (Level of Evidence: C)

2. For patients with ST elevation on the 12-lead ECG and symptoms of STEMI, reperfusion therapy should be initiated as soon as possible and is not contingent on a biomarker assay. (Level of Evidence: C)

Class IIa
Serial biomarker measurements can be useful to provide supportive noninvasive evidence of reperfusion of the infarct artery after fibrinolytic therapy in patients not undergoing angiography within the first 24 hours after fibrinolytic therapy. (Level of Evidence: B)

Class III
Serial biomarker measurements should not be relied upon to diagnose reinfarction within the first 18 hours after the onset of STEMI. (Level of Evidence: C)

The nomenclature of acute coronary syndromes is illustrated in Figure 2 (8-10). The central position of the 12-lead ECG and initial triage of patients are emphasized. Serum cardiac biomarkers (creatine kinase [CK], CK-MB, cardiac-specific troponins, myoglobin) are useful for confirming the diagnosis of MI and estimating infarct size. Serum cardiac biomarkers also provide valuable prognostic informatin. For patients with ST-segment elevation, the diagnosis of STEMI is secure; initiation of reperfusion therapy should not be delayed while awaiting the results of a cardiac biomarker assay (231,232) (Table 10). Quantitative analysis of cardiac biomarker measurements provides prognostic information as well as a noninvasive assessment of the likelihood that the patient has undergone successful reperfusion when fibrinolytic therapy is administered (Figure 11) (233,234).

Because there are differences in the clinical need for biomarke in STEMI versus NSTEMI patients and differences in the characteristics of the various cardiac biomarkers, preferential use of a particular biomarker should be base on the clinical syndrome. CK-MB is found in the skeletal muscle and blood of healthy subjects; therefore, the cutoff value for an elevated CK-MB is typically set a few units above the upper end of the reference (normal) range. In contrast, because cardiac troponin I (cTnI) and cardiac troponin T (cTnT) are not normally detected in the blood of healthy people, the definition of an abnormally increased level is a value that exceeds that of 99% of a reference control group. Given the nearly absolute myocardial tissue specificity and high sensitivity for even microscopic zones of myocardial necrosis, the ACC and the European Society of Cardiology subsequently declared cardiac troponins to be the preferred biomarker for diagnosing MI (233). A single cutoff point was recommended such that an MI would be diagnosed if, as a result of myocardial ischemia, cTnI or cTnT were detected at least once within 24 hours of the index clinical event at a level exceeding the 99th percentile of the mean value measured in a normal control population (233). The superior sensitivity makes troponin the preferred marker for patients with UA/NSTEMI. In contrast, patients with STEMI are recognized on the basis of the 12-lead ECG, and in general, subsequent confirmation of MI can be ascertained by measurement of any of the available cardiac biomarkers. Occasionally, a very small infarct will be missed by CK-MB; therefore, troponin should be measured for patients suspected to have STEMI who have negative serial CK-MBs.

It should be recognized that in patients with STEMI, cTnT and cTnI may first begin to rise above the reference limit by 3 to 6 hours from the onset of ischemic symptoms. Therefore, a significant number of patients will present o the emergency room with negative biomarkers. Myoglobin, a low-molecular-weight heme protein found in cardiac and skeletal muscle, is not cardiac specific but is released more rapidly from infarcted myocardium than CK-MB and may be detected as early as 2 hours after STEMI.

In some patients, cardiac-specific troponins may not be detectable for up to 6 hours after onset of chest pain. Thus, when CK-MB, cTnI, or cTnT levels are elevated in less than 6 hours after the onset of discomfort in patients with STEMI, clinicians should suspect that an antecedent episode of unstable angina was in fact MI and the patient is exhibiting a stuttering course of occlusion and release of the infarct artery. Data from the Global Utilization of Streptokinase and TPA for Occluded Arteries (GUSTO) III Study suggest that patients with STEMI who have elevated cTnT levels and who are less than 6 hours from the onset of discomfort have an increased mortality risk (235).

CK-MB is the preferred, widely available cardiac biomarker for most patients with STEMI, for whom the need to diagnose reinfarction and noninvasively assess reperfusion is
greater than the need to make the diagnosis. By mapping the time course of the rise and fall of a biomarker (typically CKMB), clinicians can detect an interruption of the progressive fall of the biomarker level to a point below the upper reference limit (Figure 11) (233,234). Re-elevation of the biomarker level is evidence of myocardial reinfarction (Figure 12). A more rapidly rising and falling biomarker such as CKMB or myoglobin is superior for diagnosing reinfarction. As a consequence of continuous release from a degenerating contractile apparatus in necrotic myocytes, elevations of cTnI may persist for 7 to 10 days after MI, and elevations of cTnT may persist for up to 10 to 14 days. The more protracted time course of kinetic release of cTnI and cTnT limits the ability of clinicians to make the diagnosis of reinfarction within several days after the index STEMI event. An algorithm illustrating the decision making process that incorporates biomarker measurements, ECG findings, clinical symptoms, and, if available, autopsy data for making the diagnosis of reinfarction is shown in Figure 12.

In addition to monitoring the patient for resolution of schemic-type chest discomfort and regression of the magnitude of ST-segment elevation on the ECG, clinicians can obtain serial measurements of serum cardiac markers to buttress the noninvasive diagnosis of reperfusion of the infarctrelated artery after fibrinolytic therapy (Figure 11) (233,234,236). An early peak of CK-MB (12 to 18 hours) suggests reperfusion. Because of its rapid-release kinetics, myoglobin is also an attractive marker for the early diagnosis of reperfusion.

CK-MB isoforms are another serum cardiac biomarker less frequently used for evaluating patients with STEMI. CK-MB exists in only 1 form in myocardial tissue but in different isoforms (or subforms) in the plasma. An absolute level of CKMB2 greater than 1 U/L or a ratio of CK-MB2 to CK-MB1 of 1.5 has improved sensitivity and specificity for diagnosis of MI within the first 6 hours compared with conventional assays for CK-MB (237).

6.2.5.1. Bedside Testing for Serum Cardiac Biomarkers

Class I
1. Although handheld bedside (point-of-care) assays may be used for a qualitative assessment of the presence of an elevated level of a serum cardiac biomarker, subsequent measurements of cardiac biomarker levels should be done with a quantitative test. (Level of Evidence: B)

2. For patients with ST elevation on the 12-lead ECG and symptoms of STEMI, reperfusion therapy should be initiated as soon as possible and is not contingent on a bedside biomarker assay. (Level of Evidence: C)


Handheld rapid bedside assays are clinically available for measuring cTnI, cTnT, myoglobin, and CK-MB, but in general, bedside assays are less sensitive and less precise than quantitative assays. Small desktop rapid analyzers are also available for the same purpose. A rapid, high-voltage electrophoretic system is available for measuring CK-MB isoforms. Monitoring the timing of the appearance of a positive bedside assay result may provide clinicians with a tool for a semiquantitative estimate of a serum cardiac biomarker level at the patient’s bedside (238). A positive bedside test should be confirmed by a conventional quantitative test. However, reperfusion therapy should not be delayed while one awaits the results of a quantitative assay.

6.2.6. Imaging

Class I
1. Patients with STEMI should have a portable chest X-ray, but this should not delay implementation of reperfusion therapy (unless a potential contraindication is suspected, such as aortic dissection). (Level of Evidence: C)

2. Imaging studies such as a high-quality portable cest X-ray, transthoracic and/or transesophageal echocardiography, and a contrast chest CT scan or magnetic resonance imaging scan should be used for differentiating STEMI from aortic dissection in patients for whom this distinction is initially unclear. (Level of Evidence: B)

Class IIa
Portable echocardiography is reasonable to clarify the diagnosis of STEMI and allow risk stratification of patients with chest pain who present to the ED, especially if the diagnosis of STEMI is confounded by LBBB or pacing or if there is suspicion of posterior STEMI with anterior ST depressions. (See Section 7.6.7, Mechanical Causes of Heart Failure/Low-Output Syndrome.) (Level of Evidence: B)

Class III
Single-photon emission CT (SPECT) radionuclide imaging should not be performed to diagnose STEMI
in patients for whom the diagnosis of STEMI is evident on the ECG. (Level of Evidence: B)

Various forms of imaging are often used to evaluate patients with symptoms that are suggestive of MI or acute coronary syndrome. Cardiac imaging can be of value in further determining the cause of chest discomfort in patients suspected of having an acute MI or unstable angina but whose initial ECG is normal or nondiagnostic. The 2 most studied techniques thus far have been echocardiography and radionuclide imaging.

Bedside echocardiography is useful for diagnosis and risk stratification of chest pain patients in the ED (226). A highquality portable chest X-ray, transthoracic and/or transesophageal echocardiography, and a contrast chest CT scan can be useful for differentiating acute MI from aortic dissection in patients for whom this distinction is clinically unclear. SPECT radionuclide imaging at rest is not routinely indicated to establish the diagnosis of MI in patients with STEMI, although it can provide valuable, accurate diagnostic and prognostic information in patients who present to the ED with symptoms suggestive of acute cardiac ischemia and a normal or nondiagnostic ECG (239). During the recuperative phase of hospitalization for STEMI, SPECT imaging can be used to study myocardial perfusion and to look for segmental abnormalities of LV wall motion.

6.2.7. Global Risk Assessment Tools

Global risk assessment provides an opportunity to integrate various patient characteristics into a single score that can convey an overall estimate of a patient’s prognosis over a given period of time. Beyond being informative about prognosis, the general value of these risk assessment tools is that they can influence clinical strategies. In general, the risk of the intervention should be commensurate with the underlying risk of the patient without the intervention and the expected benefit of the intervention. That is, a high-risk intervention should usually not be used for a very low-risk patient. The expected increase in risk associated with the intervention would very likely outweigh the expected benefit.

Several risk assessment tools have been proposed for patients with STEMI (240-243). One such tool uses clinical and ECG characteristics to predict risk of mortality for a patient if and if not treated with fibrinolytic therapy, as well as the risk of intracranial hemorrhage and major bleeding. This decision aid suggests that some patients with small infarctions may not have a substantial benefit from fibrinolytic therapy, particularly those who may have a risk factor for bleeding. These estimates are based on trials and registries. The use of this aid in clinical practice did not increase the use of fibrinolytic therapy overall (244). Whether the widespread application of these tools can improve decision making is not clear. Nevertheless, they provide estimates of risk that may be useful in the tailoring of therapy for individual patients. In general, however, patients who present with STEMI require evaluation for rapid reperfusion therapy and treatment with aspirin, beta-blockers, and ACE inhibitors. Nevertheless, any patient with a risk from the intervention that exceeds their STEMI risk reduction will, on average, do better without that treatment. This group will generally include patients with a higher risk from the intervention or a lower absolute risk reduction (generally because of a low absolute STEMI risk). This issue may be particularly important for younger patients, who tend to have a lower absolute risk of mortality (245), and for the elderly, who tend to have a higher risk from interventions, particularly with respect to fibrinolytic therapy (246). Precise estimates of risks and benefits are useful because the low STEMI risk in younger patients is often accompanied by a lower risk of interventions. In contrast, in the elderly, the higher intervention risk is accompanied by a higher STEMI risk (and thus a larger absolute reduction in risk with the intervention) (247).

The use of any risk assessment tool should not contribute to any delay in providing the time-sensitive assessment and treatment strategies that patients with STEMI require. Further research is necessary to determine how these tools may best contribute to optimizing patient outcomes.

6.3. Management

6.3.1. Routine Measures

6.3.1.1. Oxygen


Class I
Supplemental oxygen should be administered to patients with arterial oxygen desaturation (SaO2 less than 90%). (Level of Evidence: B)

Class IIa
It is reasonable to administer supplemental oxygen to all patients with uncomplicated STEMI during the first 6 hours. (Level of Evidence: C)


It has become universal practice to administer oxygen, usually by nasal prongs, to virtually all patients suspected of having acute ischemic-type chest discomfort, although it is not known whether this therapy limits myocardial damage or reduces morbidity or mortality. If oxygen saturation monitoring is used, therapy with supplemental oxygen is indicated if the saturation is less than 90%. Experimental results indicate that breathing oxygen may limit ischemic myocardial injury (248), and there is evidence that oxygen administration reduces ST-segment elevation (249). The rationale for use of oxygen is based on the observation that even with uncomplicated MI, some patients are modestly hypoxemic initially, presumably because of ventilation-perfusion mismatch and excessive lung water (250).

In patients with severe congestive heart failure, pulmonary edema, or a mechanical complication of STEMI, significant hypoxemia may not be corrected with supplemental oxygen alone. Continuous positive-pressure breathing or endotracheal intubation and mechanical ventilation may be required in such cases (251).

For patients without complications, excess administration of oxygen can lead to systemic vasoconstriction, and high flow rates can be harmful to patients with chronic obstructive airway disease. In the absence of compelling evidence for established benefit in uncomplicated cases, and in view of its expense, there appears to be little justification for continuing its routine use beyond 6 hours.

6.3.1.2. Nitroglycerin

Class I
1. Patients with ongoing ischemic discomfort should receive sublingual nitroglycerin (0.4 mg) every 5 minutes for a total of 3 doses, after which an assessment should be made about the need for intravenous nitroglycerin. (Level of Evidence: C)

2. Intravenous nitroglycerin is indicated for relief of ongoing ischemic discomfort, control of hypertension, or management of pulmonary congestion. (Level of Evidence: C)

Class III
1. Nitrates should not be administered to patients with systolic blood pressure less than 90 mm Hg or greater than or equal to 30 mm Hg below baseline, severe bradycardia (less than 50 beats per minute [bpm]), tachycardia (more than 100 bpm), or suspected RV infarction. (Level of Evidence: C)

2. Nitrates should not be administered to patients who have received a phosphodiesterase inhibitor for erectile dysfunction within the last 24 hours (48 hours for tadalafil). (Level of Evidence: B)


The physiological effects of nitrates include reducing preload and afterload through peripheral arterial and venous dilation, relaxation of epicardial coronary arteries to improve coronary flow, and dilation of collateral vessels, potentially creating a more favorable subendocardial to epicardial flow ratio (252-254). Vasodilation of the coronary arteries, especially at or adjacent to sites of recent plaque disruption, may be particularly beneficial for the patient with acute infarction. Nitrate-induced vasodilatation may also have particular utility in those rare patients with coronary spasm presenting as STEMI.

Clinical trial results have suggested only a modest benefit from nitroglycerin used acutely in STEMI and continued subsequently. A pooled analysis of more than 80 000 patients treated with nitrate-like preparations intravenously or orally in 22 trials revealed a mortality rate of 7.7% in the control group, which was reduced to 7.4% in the nitrate group. These data are consistent with a possible small treatment effect of nitrates on mortality such that 3 to 4 fewer deaths would occur for every 1000 patients treated (152).

Nitroglycerin may be administered to relieve ischemic pain and is clearly indicated as a vasodilator in patients with STEMI associated with LV failure. Nitrates in all forms should be avoided in patients with initial systolic blood pressures less than 90 mm Hg or greater than or equal to 30 mm Hg below baseline, marked bradycardia or tachycardia (256), or known or suspected RV infarction. Patients with RV infarction are especially dependent on adequate RV preload to maintain cardiac output and may experience profound hypotension during administration of nitrates (257). Phosphodiesterase inhibitors potentiate the hypotensive effects of nitrates because of their mechanism of action in releasing nitric oxide and increasing cyclic guanosine monophosphate (258). Therefore, it is useful clinical practice to ascertain whether such agents have been used, and nitrates should not be administered to patients who have received a phosphodiesterase inhibitor for erectile dysfunction in the prior 24 hours (48 hours for tadalafil).

Nitroglycerin is commonly given sublingually at doses of 0.4 mg when patients present with STEMI. Arterial pressure may decline precipitously because of limited control of the initial dose and rate of absorption. An intravenous infusion of nitroglycerin allows clinicians to titrate the therapy in response to the patient’s blood pressure. A useful intravenous nitroglycerin regimen employs an initial infusion rate of 5 to 10 mcg per minute with increases of 5 to 20 mcg per minute until symptoms are relieved or mean arterial blood pressure is reduced by 10% of its baseline level in normotensive patients and by up to 30% for hypertensive patients, but in no case below a systolic pressure of 90 mm Hg or a drop greater than 30 mm Hg below baseline. In view of their marginal treatment benefits, nitrates should not be used if hypotension limits the administration of beta-blockers, which have more powerful salutary effects.

6.3.1.3. Analgesia

Class I
Morphine sulfate (2 to 4 mg IV with increments of 2 to 8 mg IV repeated at 5- to 15-minute intervals) is the analgesic of choice for management of pain associated with STEMI. (Level of Evidence: C)


Pain relief is an important element in the early management of the patient with STEMI. There is a tendency to underdose patients with STEMI because of the desire to assess the response to anti-ischemic or reperfusion therapy. This should be avoided, because patients with STEMI have a hyperadrenergic state particularly early after the onset of coronary occlusion. Conversely, it should not be assumed that resolution of discomfort after administration of analgesics indicates reperfusion has occurred (see Section 6.3.1.6.3.7 for further discussion). Pain, which is commonly severe in the acute phase of the event, contributes to increased sympathetic activity.

Pain management should be directed toward acute relief of symptoms of ongoing myocardial ischemia and necrosis and toward general relief of anxiety and apprehension, the latter of which can heighten pain perception. Surges of catecholamines have been implicated as having a role in plaque fissuring and thrombus propagation and in reducing the threshold for ventricular fibrillation (259). Because the pain of STEMI is related to ongoing ischemia, interventions that affect the oxygen supply-demand relationship (i.e., by either
increasing supply or decreasing demand) may lessen the pain of STEMI (260).

Control of cardiac pain is typically accomplished with a combination of nitrates, opiate analgesic agents, oxygen, and beta-adrenergic blockers. Treatment with these agents extends from the ED to the critical care unit. An important consideration when using intravenous nitrates is not to lower blood pressure to a level that would preclude adequate dosage of morphine sulfate for pain control. Morphine sulfate remains the analgesic agent of choice for management of pain associated with STEMI, except in documented cases of morphine sensitivity. The dose required for adequate pain relief varies in relation to age and body size, as well as blood pressure and heart rate. Anxiety reduction secondary to morphine administration reduces the patient’s restlessness and the activity of the autonomic nervous system, with a consequent reduction of the heart’s metabolic demands. Morphine administration for patients with pulmonary edema is clearly beneficial and may promote peripheral arterial and venous dilation, reducing the work of breathing and slowing the heart rate secondary to combined withdrawal of sympathetic tone and augmentation of vagal tone (259,260).

Side effects of morphine administration such as hypotension can be minimized by keeping the patient supine and elevating the lower extremities if systolic pressure goes below 100 mmHg systolic, assuming pulmonary edema is not present. The concomitant use of atropine in 0.5- to 1.5-mg doses intravenously may be helpful in reducing the excessive vagomimetic effects of morphine if significant bradycardia or hypotension occurs. Although respiratory depression is relatively uncommon, patients’ respirations should be monitored, particularly as their cardiovascular status improves. The narcotic reversing agent naloxone, 0.1 to 0.2 mg intravenously, can be given initially if indicated and repeated after 15 minutes if necessary. Nausea and vomiting as potential side effects of large doses of morphine may be treated with a phenothiazine (260).

See “Hospital Management” (Section 7.2.4) for additional discussion of analgesia.

6.3.1.4. Aspirin

Class I
Aspirin should be chewed by patients who have not taken aspirin before presentation with STEMI. The initial dose should be: 162 mg (Level of Evidence: A) to 325 mg (Level of Evidence: C). Although some trials have used enteric-coated aspirin for initial dosing, more rapid buccal absorption occurs with non–enteric-coated aspirin formulations.

At a dose of 162 mg or more, aspirin produces a rapid clinical antithrombotic effect caused by immediate and near-total inhibition of thromboxane A2 production. The Second International Study of Infarct Survival (ISIS-2) has shown conclusively the efficacy of aspirin alone for treatment of evolving acute MI, with an absolute risk difference in 35-day mortality of 2.4% (relative risk reduction [RRR] 23%) (261). When aspirin was combined with streptokinase, the absolute risk difference in mortality was 5.2% (RRR 42%). A metaanalysis demonstrated that aspirin reduced coronary reocclusion and recurrent ischemic events after fibrinolytic therapy with either streptokinase or alteplase (262). Accordingly, aspirin now forms part of the early management of all patients with suspected STEMI and should be given promptly, certainly within the first 24 hours, at a dose between 162 and 325 mg and continued indefinitely at a daily dose of 75 to 162 mg (263). Although some trials have used entericcoated aspirin for initial dosing, more rapid buccal absorption occurs with non–enteric-coated formulations (264).

Unlike fibrinolytic agents, there is little evidence for a time-dependent effect of aspirin on early mortality. However, data do support the contention that a chewable aspirin is absorbed more quickly than one swallowed in the early hours after infarction, particularly after opiate therapy. The use of aspirin is contraindicated in those with a hypersensitivity to salicylate. Aspirin suppositories (300 mg) can be used safely and are the recommended route of administration for patients with severe nausea and vomiting or known upper-gastrointestinal disorders. In patients with true aspirin allergy (hives, nasal polyps, bronchospasm, or anaphylaxis), clopidogrel or ticlopidine may be substituted.

6.3.1.5. Beta-Blockers

Class I
Oral beta-blocker therapy should be administered promptly to those patients without a contraindication, irrespective of concomitant fibrinolytic therapy or performance of primary PCI. (Level of Evidence: A)

Class IIa
It is reasonable to administer IV beta-blockers promptly to STEMI patients without contra-indications, especially if a tachyarrhythmia or hypertension is present. (Level of Evidence: B)


During the first few hours after the onset of STEMI, betablocking agents may diminish myocardial oxygen demand by reducing heart rate, systemic arterial pressure, and myocardial contractility. In addition, prolongation of diastole caused by a reduction in heart rate may augment perfusion to ischemic myocardium, particularly the subendocardium. As a result, immediate beta-blocker therapy appears to reduce 1) the magnitude of infarction and incidence of associated complications in subjects not receiving concomitant fibrinolytic therapy, 2) the rate of reinfarction in patients receiving fibrinolytic therapy, and 3) the frequency of life-threatening ventricular tachyarrhythmias.

In patients not receiving fibrinolytic therapy, intravenously administered beta-blocking agents exert a modestly favorable influence on infarct size (265). Large early trials suggested a mortality benefit as well. In ISIS-1 (266), more than 16 000 patients with suspected acute MI were enrolled within 12 hours of onset of symptoms; immediate atenolol, 5 to 10 mg IV, followed by oral atenolol, 100 mg daily, reduced 7-day mortality from 4.3% to 3.7% (p less than 0.02; 6 lives saved per 1000 treated). The mortality difference between those receiving and not receiving atenolol was evident by the end of day 1 and was sustained subsequently. In the Metoprolol In Acute Myocardial Infarction (MIAMI) trial (267), more than 5700 subjects with evolving MI were randomly assigned to receive placebo or metoprolol, up to 15 mg IV in 3 divided doses followed by 50 mg orally every 6 hours for 48 hours and then 100 mg twice per day thereafter. Fifteen-day mortality was reduced with metoprolol from 4.9% to 4.3%. As in ISIS-1, the mortality difference between those given placebo and those receiving metoprolol was evident by the end of day 1, after which it was sustained.

In subjects receiving concomitant fibrinolytic therapy, intravenously administered beta-blocking drugs diminish the incidence of subsequent nonfatal reinfarction and recurrent ischemia. In addition, they may reduce mortality if given particularly early (i.e., within 2 hours) after onset of symptoms. In the Thrombolysis In Myocardial Infarction Phase II (TIMI-II) trial (268), in which all patients received IV alteplase, those randomly assigned to receive metoprolol, 15 mg IV, followed by oral metoprolol, 50 mg twice per day for 1 day and then 100 mg twice per day thereafter, had a diminished incidence of subsequent nonfatal reinfarction and recurrent ischemia compared with those begun on oral metoprolol 6 days after the acute event. Among those treated especially early (i.e., within 2 hours of symptom onset), the composite end point, death or reinfarction, occurred less often in those given immediate IV metoprolol than in those who did not receive it.

The benefits of routine early IV use of beta-blockers in the fibrinolytic era have been challenged by 2 later randomized trials of IV beta-blockade (269,270) and by a post hoc analysis of the use of atenolol in the GUSTO-I trial (271). A subsequent systematic review of early beta-blocker therapy in STEMI found no significant reduction in mortality (67). Therefore, data on the early use of intravenous beta-blockade in STEMI are inconclusive, and patterns of use vary.

Beta-blockers should not be administered to patients with STEMI precipitated by cocaine use because of the risk of exacerbating coronary spasm (272). If IV beta-blockade induces an untoward effect, such as atrioventricular (AV) block, excessive bradycardia, or hypotension, the condition is quickly reversed by infusion of a beta-adrenergic agonist (i.e., isoproterenol 1 to 5 mcg/min). The presence of moderate LV failure early in the course of STEMI should preclude the use of early IV beta-blockade until the heart failure has been compensated but is a strong indication for the oral use of beta-blockade before discharge from the hospital.

The following are relative contraindications to beta-blocker therapy: heart rate less than 60 bpm, systolic arterial pressure less than 100 mm Hg, moderate or severe LV failure, signs of peripheral hypoperfusion, shock, PR interval greater than 0.24 second, second- or third-degree AV block, active asthma, or reactive airway disease.

Randomized trials of beta-blocker therapy in patients with STEMI undergoing PCI without fibrinolytic therapy have not been performed. However, it seems reasonable pending further information to extrapolate data from those receiving another form of revascularization, fibrinolytic therapy, to the PCI population. The more contemporary CAPRICORN (Carvedilol Post-infarct Survival Controlled Evaluation) trial (273), which includes patients undergoing either form of revascularization, confirms the benefits of beta-blocker therapy in patients with transient or sustained postinfarction LV dysfunction.

6.3.1.6. Reperfusion

6.3.1.6.1. General Concepts

Class I
All STEMI patients should undergo rapid evaluation for reperfusion therapy and have a reperfusion strategy implemented promptly after contact with the medical system. (Level of Evidence: A)

Although rapid spontaneous reperfusion of the infarct artery may occur, in the majority of patients there is persistent occlusion of the infarct artery in the first 6 to 12 hours while the affected myocardial zone is undergoing necrosis. Prompt and complete restoration of flow in the infarct artery can be achieved by pharmacological means (fibrinolysis), PCI (balloon angioplasty with or without deployment of an intracoronary stent under the support of pharmacological measures to prevent thrombosis), or surgical measures (Figure 3) (24-40). Despite the extensive improvement in intraoperative preservation with cardioplegia and hypothermia and in numerous surgical techniques, it is not logistically possible to provide surgical reperfusion in a timely fashion, and therefore patients with STEMI who are candidates for reperfusion routinely receive either fibrinolysis or a catheter-based treatment.

Evidence exists that expeditious restoration of flow in the obstructed infarct artery after the onset of symptoms in patients with STEMI is a key determinant of short- and longterm outcomes regardless of whether reperfusion is accomplished by fibrinolysis or PCI (24,274,275). As discussed previously (Section 4.1), efforts should be made to shorten the time from recognition of symptoms by the patient to contact with the medical system. All healthcare providers caring for patients with STEMI from the point of entry into the medical system must recognize the need for rapid triage and implementation of care in a fashion analogous to the handling of trauma patients. When considering recommendations for timely reperfusion of patients with STEMI, the Writing Committee reviewed data from clinical trials, focusing particular attention on enrollment criteria for selection of patients for randomization, actual times reported in the trial report rather than simply the allowable window specified in the trial protocol, treatment effect of the reperfusion strategy on individual components of a composite primary end point (e.g., mortality, recurrent nonfatal infarction), ancillary therapies (e.g., antithrombin and antiplatelet agents), and the interface between fibrinolysis and referral for angiography and revascularization. When available, data from registries were also reviewed to assess the generalizability of observations from clinical trials of reperfusion to routine practice. Despite the wealth of reports on reperfusion for STEMI, it is not possible to produce a simple algorithm given the heterogeneity of patient profiles and availability of resources in various clinical settings at various times of day. This section introduces the recommendations for an aggressive attempt to minimize the time from entry into the medical system to implementation of a reperfusion strategy using the concept of medical system goals. More detailed discussion of these goals and the issues to be considered in selecting the type of reperfusion therapy may be found in Section 6.3.1.6.2, followed by a discussion of available resources in Section 6.3.1.6.2.1.

The medical system goal is to facilitate rapid recognition and treatment of patients with STEMI such that door-to-needle (or medical contact–to-needle) time for initiation of fibrinolytic therapy can be achieved within 30 minutes or that door-to-balloon (or medical contact–to-balloon) time for PCI can be kept under 90 minutes. These goals may not be relevant for patients with an appropriate reason for delay, such as uncertainty about the diagnosis (particularly for the use of fibrinolytic therapy), need for evaluation and treatment of other life-threatening conditions (e.g., respiratory failure), or delays associated with the patient’s informed choice to have more time to consider the decision. In the absence of such types of circumstances, the emphasis is on having a system in place such that when a patient with STEMI presents for medical care, reperfusion therapy can be provided as soon as possible within these time periods. Because there is not considered to be a threshold effect for the benefit of shorter timesto reperfusion, these goals should not be understood as “ideal” times but rather the longest times that should be considered acceptable. Systems that are able to achieve even more rapid times for patients should be encouraged. Also, this goal should not be perceived as an average performance standard but as a goal that an early treatment system in every hospital should seek for every appropriate patient.

A critically important goal of reperfusion is to restore flow in the infarct artery as quickly and as completely as possible, but the ultimate goal of reperfusion in STEMI is to improve myocardial perfusion in the infarct zone. Despite adequate restoration of flow in the epicardial infarct artery, perfusion of the infarct zone may still be compromised by a combination of microvascular damage and reperfusion injury (276-278). Microvascular damage occurs as a consequence of downstream embolization of platelet microemboli and thrombi followed by the release of substances from activated platelets that promote occlusion or spasm in the microvasculature. Reperfusion injury results in cellular edema, free radical formation, calcium overload, and acceleration of the apoptotic process. Cytokine activation in the infarct zone leads to neutrophil accumulation and inflammatory mediators that contribute to tissue injury.

Thus, construction of an ideal reperfusion regimen in patients with STEMI not only should focus on the primary means of restoring flow in the epicardial infarct artery (pharmacological or catheter-based) but should also include adjunctive and ancillary treatments that minimize the amount of microvascular damage and protect the jeopardized myocardial infarct zone that contains cells in various stages of ischemia, necrosis, and apoptosis (279,280). The Writing Committee endorses further research to identify the optimum strategies for achieving these goals.

6.3.1.6.2. Selection Of Reperfusion Strategy

The literature provides very strong evidence that among patients with suspected STEMI and without contraindications, the prompt use of reperfusion therapy is associated with improved survival (156). Despite such strong evidence, studies continue to indicate that reperfusion therapy is underutilized and often not administered soon after presentation (281-283). Indecision about the choice of reperfusion therapy should not deter physicians from using these strategies or delay them in administering therapy.

There is controversy about which form of reperfusion therapy is superior in various clinical settings. Part of the uncertainty derives from the continual introduction of new agents, devices, and strategies, which quickly make previous studies less relevant to contemporary practice. With pharmacological reperfusion therapies, there are new agents, dosing regimens, adjunctive treatments, and combined strategies with procedures that are in a continual process of refinement and evaluation. Similarly, with catheter-based approaches, there are new devices, adjunctive therapies, technologies, and combined strategies with medications that are being introduced and evaluated. As a result, the evidence base regarding the best approach to reperfusion therapy is quite dynamic.

Several issues should be considered in selecting the type of reperfusion therapy, as discussed below.

Time From Onset of Symptoms.
Time from onset of symptoms to fibrinolytic therapy is an important predictor of MI size and patient outcome (284). The efficacy of fibrinolytic agents in lysing thrombus diminishes with the passage of time (279). Fibrinolytic therapy administered within the first 2 hours (especially the first hour) can occasionally abort MI and dramatically reduces mortality (Figure 13) (156,159). The National Heart Attack Alert Working Group (179) recommends that EDs strive to achieve a 30-minute door-toneedle time to minimize treatment delays. Prehospital fibrinolysis reduces treatment delays by up to 1 hour and reduces mortality by 17% (285).

The amount of myocardium at risk, presence of collateral blood flow, and duration of coronary occlusion are major determinants of myocardial infarct size (286-289). In animal models (18), occlusions persisting greater than 30 minutes produce myonecrosis. Reperfusion at 90 minutes salvages approximately half of the myocardium at risk. Myocardial salvage is minimal after 4 to 6 hours of ischemia unless ischemic preconditioning and/or collateral flow have modified the wave front of necrosis.

A time-dependent decrease in efficacy of fibrinolytic therapy may also contribute to the higher mortality rate in patients with longer symptom duration (279). In contrast, the ability to produce a patent infarct artery is much less dependent on symptom duration in patients undergoing PCI. Several reports claim no influence of time delay on mortality rates when PCI is performed after 2 to 3 hours of symptom duration (290,291). One study suggests that time to PCI is only important for patients presenting with shock (292). Another showed that time was associated with outcome in higher-risk but not lower-risk patients (293). Conversely, others have reported increasing mortality rates with increasing door-to-balloon times (294,295). Importantly, after adjustment for baseline characteristics, time from symptom onset to balloon inflation is significantly correlated with 1- year mortality in patients undergoing primary PCI for STEMI (relative risk [RR] equals 1.08 for each 30 minute delay from symptom onset to balloon inflation, p equals 0.04) (275,275a). Interestingly, although the CAPTIM (173) and PRAGUE-2 (177) studies reached different conclusions about the overall superiority of PCI over fibrinolysis, important observations were made in the subset of patients presenting very early after the onset of symptoms. In the subset of patients presenting within 3 hours of the onset of symptoms in PRAGUE-2, mortality was equivalent in those treated with streptokinase and those transferred with PCI (177). Patients treated within 2 hours of symptom onset in CAPTIM had improved outcomes with prehospital tissue plasminogen activator (tPA) versus transfer for PCI (176). (See Section 6.3.1.6.2.1.)

It is also possible that time-to-treatment analyses have been confounded by other variables (293, 296). First, higherrisk patients report later to the hospital and may respond better to PCI than to fibrinolytic agents. Second, shorter doorto- balloon times may be a surrogate for better quality of care and adherence to treatment guidelines. The Task Force on the Management of Acute Myocardial Infarction of the European Society of Cardiology (297) and this Committee both recommend a target medical contact– or door-to-balloon time of less than 90 minutes.

Risk of STEMI. Several models have been developed that assist clinicians in estimating the risk of mortality in patients with STEMI (240-242,298,299). Although these models vary somewhat in the factors loaded into the risk prediction tool and also vary with respect to statistical measures of their discriminative power (e.g., C statistic), all the models provide clinicians with a means to assess the continuum of risk from STEMI. None of the models have been tested prospectively by randomizing patients to a reperfusion strategy based on estimated mortality at presentation. Retrospective analyses do suggest that the absolute difference in mortality at 30 days between PCI and fibrinolysis increases in favor of PCI as the estimated risk of mortality with fibrinolysis increases (300). Conversely, as the estimated mortality benefit with fibrinolysis decreases, the absolute mortality benefit of PCI decreases, with equipoise appearing (i.e., similar 30-day mortality rates) when the estimated mortality with fibrinolysis is approximately 2% to 3% (300).

When the estimated mortality with fibrinolysis is extremely high, as is the case in patients with cardiogenic shock, compelling evidence exists that favors a PCI strategy. The SHOCK trial (SHould we emergently revascularize Occluded Coronaries for cardiogenic shocK?) demonstrated that patients with cardiogenic shock have a better 1-year survival if they have undergone early coronary revascularization (184). At 1 year, patients in the early revascularization group had a mortality rate of 53% compared with 66% for the group that had initial medical stabilization followed by no or late revascularization (184,301). Observational data from NRMI suggest superiority of PCI over fibrinolysis for patients with Killip class greater than or equal to II (302).

Risk of Bleeding. Choice of reperfusion therapy is also affected by the patient’s risk of bleeding. When both types of reperfusion are available, the higher the patient’s risk of bleeding with fibrinolytic therapy, the more strongly the decision should favor PCI. If PCI is unavailable, then the
benefit of pharmacological reperfusion therapy should be balanced against the risk. A decision analysis suggested that fibrinolytic therapy should be favored against no reperfusion treatment until the risk of a life-threatening bleed exceeds 4% in older patients who have a risk profile similar to those in the classic randomized trials of fibrinolytic therapy (247).
Risk scores for bleeding after fibrinolytic therapy allow for the calculation of this risk (246). Because they are derived from less restricted populations, the scores that are most generalizable are those derived from observational studies (246).

Time Required for Transport to Skilled PCI Laboratory. The availability of interventional cardiology facilities is a key determinant of whether PCI can be provided. For facilities that can offer PCI, the literature suggests that this approach is superior to pharmacological reperfusion (303). The trials comparing pharmacological and PCI strategies, however, were conducted before the advent of more recent pharmacological and PCI strategies. When a composite end point of death, nonfatal recurrent MI, or stroke is analyzed, much of the superiority of a PCI strategy is driven by a reduction in the rate of nonfatal recurrent MI (Figure 14) (40). The rate of nonfatal recurrent MI can be influenced both by the adjunctive therapy used (Figure 3) (24-40) and by the proportion of patients who are referred for PCI when the initial attempt at fibrinolysis fails or myocardial ischemia recurs after initially successful pharmacological reperfusion (Figure 14) (155).

The experience and location of the PCI laboratory also plays a role in the choice of therapy. The trials were performed in centers with highly experienced teams, and their results may not be generalizable to all PCI laboratories throughout the country. Not all laboratories can provide prompt, high-quality primary PCI. Even centers with interventional cardiology facilities may not be able to provide the staffing required for 24-hour coverage of the catheterization laboratory. Despite staffing availability, the volume of cases in the laboratory may be insufficient for the team to acquire and maintain skills required for rapid PCI reperfusion strategies. A study from NRMI investigated the effect of volume on the outcomes of patients treated with PCI versus pharmacological reperfusion strategies (303). They studied 446 acute-care hospitals, with 112 classified as low-volume (fewer than or equal to 16 procedures), 223 as intermediatevolume (17 to 48 procedures), and 111 as high-volume (49 or more procedures) based on their annual primary angioplasty volume. They reported that patients hospitalized at intermediate- and high-volume centers had lower mortality with PCI reperfusion, whereas in the low-volume centers, there was no significant difference between the 2 reperfusion strategies. In another article from the NRMI investigators, the volume of primary PCI procedures, but not pharmacological treatment, was inversely associated with the mortality rate for patients with STEMI (304).

A decision must be made when a STEMI patient presents to a center without interventional cardiology facilities. Fibrinolytic therapy can generally be provided sooner than primary PCI (Figure 7) (180). As the time delay for performing PCI increases, the mortality benefit associated with expeditiously performed primary PCI over fibrinolysis decreases (305). Compared with a fibrin-specific lytic agent, a PCI strategy may not reduce mortality when a delay greater than 60 minutes is anticipated versus immediate administration of a lytic (Figure 15) (305).

The balance of risk/benefit between the transfer of patients for PCI and more immediate treatment with fibrinolytic therapy remains uncertain. The DANAMI-2 trial (DANish trial in Acute Myocardial Infarction), conducted in Denmark, found that patients treated at facilities without interventional cardiology capabilities had better composite outcomes with transfer for PCI within 2 hours of presentation than with pharmacological reperfusion treatment at the local hospital (306). Whether these results could be replicated elsewhere is not known. An alternative to transfer is for hospitals without on-site cardiac surgery to develop the capability to provide primary mechanical reperfusion therapy. A study by Aversano and colleagues with 11 hospitals in Massachusetts and Maryland suggested that this approach may improve outcomes (307). It can be expected, however, that only a limited number of hospitals could develop such a program, and it has yet to be determined whether a certain volume of cases would be necessary to maintain the effectiveness of the service. The economic implications of expansion of the number of PCI-capable centers that are able to maintain an inventory of the necessary catheters and other devices and provide 24-hour coverage, 7 days per week, deserve further evaluation from the perspectives of individual institutions and the global healthcare delivery system. See additional discussion in Section 6.3.1.6.2.1.

Given the current literature, it is not possible to say definitively that a particular reperfusion approach is superior for all patients, in all clinical settings, at all times of day (173,176,177) (Danchin N; oral presentation, American Heart Association 2003 Annual Scientific Sessions, Orlando, FL, November 2003). The main point is that some type of reperfusion therapy should be selected for all appropriate patients with suspected STEMI. The appropriate and timely use of some reperfusion therapy is likely more important than the choice of therapy, given the current literature and the expanding array of options. Clinical circumstances in which fibrinolytic therapy is generally preferred or an invasive strategy is generally preferred are shown in Table 11.

6.3.1.6.2.1. Available Resources

Class I
STEMI patients presenting to a facility without the capability for expert, prompt intervention with primary PCI within 90 minutes of first medical contact should undergo fibrinolysis unless contraindicated. (Level of Evidence: A)


The preferred reperfusion therapy for STEMI must take into account the location of the patient, the response time and expertise of the paramedical/ambulance personnel, their relationship to the regional healthcare facility(s), and the availability, capability, and expertise of the medical personnel at the facility. The most recent NRMI data continue to support advantages of primary PCI versus fibrinolysis in high-PCI–volume hospitals but not among those institutions with low volume (fewer than or equal to 16 procedures per year) (303). Approximately 20% of US hospitals have cardiac catheterization laboratories; and less than that number have the capacity for primary PCI. The C-PORT study (Atlantic Cardiovascular Patient Outcomes Research Team), which randomized 451 fibrinolysis-eligible patients with STEMI treated in 11 community hospitals with diagnostic catheterization but not onsite PCI facilities, is of interest. Patients were randomized within 12 hours of symptom onset to an accelerated alteplase regimen (median door-to-lytic time was 46 minutes) versus primary PCI (median door-to-balloon time 101.5 minutes). At 6 months, the incidence of death, re-MI, and stroke was 12.4% for PCI and 19.9% for fibrinolytic therapy (p equals 0.03). Because only 18% of the intended sample size was actually enrolled, this study is significantly underpowered, and its conclusion can only be hypothesis-generating rather than definitive (307). Importantly, most C-PORT patients were randomized between 0800 and 1600 hours, an experience consistent with NRMI data. The NRMI report also demonstrated a substantially longer door-to-balloon time when patients with STEMI undergo direct PCI outside of daylight hours (308). The Zwolle group evaluated 1702 consecutive patients and found that the 47% of patients who presented outside “routine duty hours” (i.e., 1800 to 0800 hours) had a higher rate of both PCI failure and 30-day mortality than those within the 0800 to 1800 period (6.9% and 4.2% versus 3.8% and 1.9%, respectively; p less than 0.01) (309). The reasons for these differences are unclear but could relate to both patient and process-of-care factors, including variations in both cognitive function and manual dexterity in sleep-deprived healthcare providers (308,310).

Two studies germane to STEMI care and resource utilization are noteworthy. The first, CAPTIM, a comparison of angioplasty and prehospital fibrinolysis (accelerated alteplase) in STEMI, fell short of its planned 1200-patient enrollment, and hence was underpowered (173). Eight hundred forty patients were randomized to prehospital fibrinolysis versus primary PCI. The primary end point was the composite of all-cause mortality, nonfatal recurrent MI, and nonfatal disabling stroke at 30 days, which occurred in 8.2% of patients assigned fibrinolytic therapy and 6.2% of patients assigned to PCI (p not significant [NS]). The components for death, reinfarction, and disabling stroke were 3.8%, 3.7%, and 1.0% for fibrinolytic therapy and 4.8%, 1.7%, and 0% for PCI. Unlike C-PORT, this trial liberally used rescue angioplasty (28%), which probably accounts for the relatively low reinfarction rate in the lytic-treated group. A subsequent analysis from CAPTIM of the 55% of patients treated within 2 hours of symptom onset revealed a mortality trend in favor of prehospital fibrinolysis versus primary PCI (2.2% versus 5.7%, p equals 0.058), whereas those patients treated beyond 2 hours had a 5.9% versus 3.7% (p equals 0.47) 30-day mortality rate, respectively (176). Interestingly, there was a significant reduction in the frequency of cardiogenic shock for patients treated within 2 hours with prehospital fibrinolysis (1.3% versus 5.3%, p equals 0.032), whereas the frequency of this event after 2 hours was similar (i.e., 3.9% versus 4.4%, respectively) (176).

The DANAMI-2 study, which compared primary PCI versus accelerated alteplase, enrolled 1572 patients versus the 2000 patients planned (306). Patients were eligible if they had a sum of greater than 0.4 mV of ST elevation in 2 contiguous leads on their presenting ECG within 12 hours of symptom onset; however, patient enrollment consisted of 37% of those screened, and patients deemed to be at high risk during ambulance transport were excluded (306). Twentynine hospitals, of which 5 conducted primary PCI and were located a mean of 35 miles from referring hospitals (maximum 95 miles), participated. The median door-to-needle time for patients randomized to fibrinolysis was approximately 50 minutes for patients presenting either to a community (referral) hospital or an invasive center. For patients who presented to a community hospital (where 1129 patients were enrolled), the time from initial presentation to balloon inflation at an invasive center was 108 minutes; the door-toballoon time was 93 minutes for patients presenting to an invasive center and randomized to PCI. The primary composite end point of death, reinfarction, and stroke through 30 days occurred in 14.2% of the fibrinolysis-treated patients and 8.5% of the PCI-treated patients (p less than 0.001). The individual end-point components of death, reinfarction, and stroke occurred in 7.8%, 6.3%, and 2.0% of the fibrinolysistreated patients and 6.6%, 1.6%, and 1.1% of the PCI-treated patients, respectively. In addition to the exclusion of highrisk patients for transport, some caveats in DANAMI-2 are noteworthy: 1) The antithrombotic dosing regimen was in excess of ACC/AHA guidelines. 2) The protocol specified that repeat fibrinolysis was to be used for failed reperfusion, reinfarction, and recurrent ST-elevation ischemia; this strategy was used in 26 patients within 12 hours after randomization, and only 1.9% underwent rescue PCI. 3) Patients with prior stroke were included, and an imbalance in this baseline characteristic was present, (i.e., 4.0% for fibrinolysis versus 2.7% for PCI ([1-sided p equals 0.06]). 4) The difference in reinfarction rates between the 2 groups was likely exaggerated by the exclusion of those patients associated with invasive procedures (311).

Hence, on the basis of the data, patients with STEMI presenting to a facility without the capability for expert, prompt intervention with primary PCI within 90 minutes of first medical contact should undergo fibrinolysis unless contraindicated. (See Sections 6.3.1.6.4.2, Primary PCI, and 6.3.1.6.4.2.4, Interhospital Transfer for Primary PCI.)

6.3.1.6.3. Pharmacological Reprefusion. Rationale For Fibrinolytic Therapy. Although the clinical features of coronary obstruction were described nearly a century ago (312,313), thrombotic obstruction of the infarct artery as a cause of STEMI was not proven until 1980 (20). The benefits of fibrinolytic therapy are maximal when there is prompt, adequate restoration of flow in the epicardial infarct artery and perfusion of the myocardium in the infarct zone. Controlled clinical trials have demonstrated the potential forfunctional, clinical, and mortality benefits only if fibrinolytic therapy is given within 12 hours. (See additional discussion on the use of antithrombins and antiplatelet agents as ancillary therapy in Sections 6.3.1.6.8.1 and