Ask the Expert – September/October

Joel Schofer, MD FAAEM
LCDR MC USN, Naval Hospital Okinawa

“Ask the Expert” is a Common Sense feature where subject matter experts provide answers to questions provided by AAEM & YPS members. This edition features:

William J. Brady, MD, Professor of Emergency Medicine and Medicine, Vice Chair of Emergency Medicine Department of Emergency Medicine, University of Virginia, Charlottesville, VA

Chris A. Ghaemmaghami, MD, Associate Professor of Emergency Medicine and Medicine, Program Director, Emergency Medicine Residency, Department of Emergency Medicine, University of Virginia, Charlottesville, VA

Original question: Which chest pain patients can be safely discharged from the emergency department?
Amended question: Which chest pain patients initially suspected of acute coronary syndrome after ED evaluation can be safely discharged from the emergency department?

The appropriate evaluation and management of the chest pain patient remains a significant clinical challenge for the emergency physician. The emergency physician will use the history of the event, related medical history, physical examination, 12-lead ECG, and initial serum markers -- if obtained -- to determine the need for further inpatient evaluation.

The evaluation and subsequent management of the chest pain patient is thus deserving of significant review and consideration. Both the number of patients involved, as well as the potential diagnoses encountered in this chest pain patient group, when considered together, account for the significant clinical challenge. Contemporary numbers note the significant magnitude of the chest pain patient in the ED. Of the approximately 100 million ED visits in the United States today, 8% of these patients present with chest pain; five million of these patients are admitted for further evaluation and management of the chest pain syndrome. Ultimately, whether in the ED or the hospital, almost half of these patients are diagnosed with an acute coronary syndrome (ACS) with 1.1 million AMI and 1.2 million USA cases; further, 300,000 of these admitted patients die of cardiovascular ailments. Three million of these chest pain patients are discharged from the ED with a very low rate of missed ACS (approximately 2% for both unstable angina and AMI). Thus, the “numbers” are impressive with respect to the shear magnitude of the presentation.

Not only are the numbers impressive, but also the range of possible diagnoses, ranging from a benign musculoskeletal or gastrointestinal (GI) syndrome to a life threatening pulmonary or cardiovascular event. The differential diagnosis of chest pain includes musculoskeletal/chest wall syndrome, gastro-esophageal reflux, esophageal perforation, acute pancreatitis, other GI entities, pneumonia, pulmonary embolism with infarction, pneumothorax, aortic dissection, and acute coronary syndrome, among many other possible ailments.

In reply to this query, we will limit our response to the chest pain patient initially suspected of ACS who is evaluated in the ED. Thus the question is as follows: Which chest pain patients initially suspected of acute coronary syndrome can be safely discharged from the emergency department after an ED-based evaluation? First of all, it is quite easy to identify those patients clearly deserving admission to the hospital; those patients possessing the following features in the history, examination and diagnostic studies are reasonable candidates for admission to the hospital: (1) known history of coronary artery obstructive lesions with a recurrent, altered, progressive, and/or unrelenting chest pain pattern similar to past ischemic chest discomfort; (2) acute congestive heart failure with or without persistent hypotension; (3) sustained, compromising hypotension temporally unrelated to vasodilator therapy administration; (4) 12-lead electrocardiogram with obvious, clinically concerning abnormality (ST segment deviation, new bundle branch block, or dynamic ST segment/T wave changes during the episode of evaluation); and (5) markedly abnormal serum marker occurring within the appropriate clinical context. This group of chest pain patients would be likely considered “high probability” presentations for ACS based upon the various historical, examination or diagnostic results – and would require admission to the hospital for additional diagnosis and therapy. These patients are diagnosed with STEMI, NSTEMI and unstable angina (USA) and are managed with aggressive anti-anginal, anti-platelet and ant-coagulant therapies; a subgroup is treated with percutaneous coronary intervention in urgent (NSTEMI and USA) and emergent (STEMI) fashion.

At “the other end” of the ACS spectrum of clinical concern, the emergency physician is faced with the “very low probability for acute coronary syndrome” presentation. This group of patients has a markedly lower probability for ACS due to a variety of factors such as atypical host age (i.e., quite young), unusual chest discomfort description, physical examination with obvious abnormality suggesting a non-ACS source of the pain and normal or non-worrisome 12-lead ECG, to name only the most commonly evaluated features of the presentation. These individual factors, however, must not be considered in isolation – rather, these individual issues must be considered as “one small piece” of the larger diagnostic puzzle and evaluated as part of the entire picture. Exceptions to these individual factors abound in the ACS literature. For instance, the “young” 32 year-old male patient can experience ACS; the “sharp” chest pain can be a manifestation of ACS; the chest pain which is reproducible can be encountered in AMI; and the normal or nonspecifically abnormal ECG can be seen early in the course of USA or AMI. If the clinician relies solely on a single feature to “ruleout” ACS, then the diagnosis can, in fact, be missed and the patient managed inappropriately. Yet, if the less concerning features are encountered in a cluster (i.e., simultaneously in a single patient) and are analyzed in this fashion, then the likelihood of ACS is significantly lower. For example, the patient who presents with an unusual chest pain syndrome which is entirely reproducible on examination and demonstrates a normal 12-lead ECG is very unlikely to experience ACS – the combination of these factors, encountered simultaneously in a single patient, is quite powerful in “ruling out” ACS. Such patient can be safely discharged from the ED from an ACS perspective, assuming that an appropriate follow-up mechanism exists.

Once these two extreme ends of the diagnostic spectrum have been removed, the emergency physician is left with the low-to-intermediate probability presentations – this heterogeneous group truly represents the most difficult diagnostic group in the potential ACS ED population and thus the focus of this question. In addressing this rather broad patient group, the EP will use the history, ECG and serum marker analysis as the primary diagnostic tools. In this subgroup of chest pain patients, diagnostic studies (i.e., 12-lead ECG and serum troponin values) are normal or unrevealing; the risk stratification of these patients with nondiagnostic basic study results is thus challenging.

Utilizing various features of the ED evaluation, clinical decision rules have been developed to assist in this challenging process. First, the clinical history is often taught to be a major diagnostic investigation in the evaluation of the ED chest pain patient. Sanchis and colleagues1 explored the value of the clinical history by itself in this evaluation, concluding that the clinical history itself is not a primary determinant of safe discharge. The study was comprised of 1,011 patients presenting to the ED; data regarding the patient’s history reviewed included the clinical presentation (pain characteristics and number of episodes), coronary risk factors and history of ischemic heart disease and noncardiac vascular disease. The various models noted an impressive ability to identify patients at low-risk of adverse outcomes at one year. In fact, 44% of the patients were appropriate discharge candidates with an adverse event rate of only 1.4% at one year; unfortunately, the model performed less well at 30 days with an 8% rate of adverse event, largely resulting from revascularization. The authors concluded that patient selection for discharge is not reliably determined by the clinical history itself, recommending that additional ancillary tools should be considered in this decision.

Considering the clinical history and focusing on the description of the chest sensation, Schillinger et al2 noted that several atypical historical features are associated with a low rate of occurrence of both AMI and adverse outcome. The authors investigated the value of historical features (symptoms, medical history and risk factors) atypical for ACS in the exclusion of acute/subacute coronary events. The study population included 1,288 ED patients with chest pain; the patient histories were separated into typical or atypical for ACS using seven discomfort descriptors. Acute myocardial infarction was found in 13% of patients; a six-month adverse event rate of 19% was observed. Atypical presentations with four or more atypical descriptors was associated with low likelihood of AMI, death and revascularization; in younger patients (aged less than 40 years) with at least four atypical features, AMI was not seen, and the adverse event rate was markedly low. Conversely, typical presentations demonstrated a markedly higher rate of AMI and related poor outcome. The authors concluded that atypical features are of value in “ruling out” ACS, yet the converse is not true – that typical features are not reliable for “ruling in” AMI and related poor outcome. Thus, these two investigations suggest that the history of the event is important but cannot be used by itself to determine the need for further work-up and ultimate ED disposition.

The physical examination has largely been of limited value in this consideration.The exception to this statement includes the identification of complications of ACS, such as hypotension and pulmonary edema – yet these patients are identifiably ill and thus the disposition is reasonably straightforward. Exploring this issue, Schillinger and colleagues3 asked the question – does the presence of pulmonary congestive in individuals with unexplained chest pain identify a subset of ED patients at increased risk of poor outcome? Such patients demonstrated higher rates of ACS diagnosis as well as increased risk for poor outcome. Thus, in this patient group, admission to the hospital for further evaluation and management of the chest pain syndrome and pulmonary congestion is usually warranted.

Other decision rules have incorporated not only the clinical history, but also the remainder of the ED data (examination, ECG and biomarker). In just such an application, the Vancouver Chest Pain Rule4 is focused on the identification of ED chest pain patients with a low risk of acute coronary syndrome – i.e., those patients that can be safely discharged from the ED. In this study, patients greater than 25 years of age were evaluated at presentation with ultimate ACS vs. non-ACS diagnosis assigned at 30 days from the initial visit; data considered in the development of the rule included coronary artery disease risk factors, chest pain characteristics, physical examination and ECG findings and cardiac biomarker results. In the 769 patients studied, approximately 20% had ACS (10% AMI and 11.4% unstable angina); the remaining 80% of the patients were diagnosed with non-ACS conditions. The decision tool demonstrated an impressive sensitivity (98.8%) in the identification of patients who were safely discharged from the ED. In this group, the authors noted that patients who exhibited a normal initial ECG, lacked previous ischemic chest pain and were younger than 40 years demonstrated a very low risk of acute coronary syndrome. In addition, in patients over age 40 years who demonstrated a normal ECG, lacked previous ischemic chest pain, had low-risk pain characteristics, and revealed negative initial/repeat serum markers were also at low risk for ACS. This rule is certainly of value in the younger patient with a negative ED evaluation – it essentially identifies a subgroup of patients who can be safely discharged from the ED after a brief evaluation; it is of less clinical value in older patients or in those individuals with a past history of ischemic heart disease; of course, these subgroups are frequently encountered in the ED.

The use of clinical decision rules in a relatively young patient population can assist with appropriate disposition decisions. Marsan and colleagues5 attempted to develop a clinical decision rule that young adult chest pain patients without known cardiac disease, coronary risk factors and an abnormal ECG were at extremely low risk for ACS and adverse short-term outcome. Initially, the investigators used only the data listed above; the model did not perform very well with 5.4% of patients experiencing ACS and 2.2% of patients having an adverse outcome. With the addition of serum marker results into the model, a much better predictive performance was found – in young adult patients without known cardiac history, with either no classic coronary risk factors or a normal ECG and with initially normal biomarker, the risk of ACS was extremely low (0.14%); no adverse cardiovascular events at one month. Thus, in this group, the history and ECG did not perform well; the addition of the biomarker testing, however, greatly increased the rule’s ability in this young population.

The Goldman criteria and the Thrombolysis in Myocardial Infarction (TIMI) risk score have been used for hospitalized patients in the determination of risk stratification. In this setting, these tools have performed reasonably well, yet neither has identified any subgroup of individuals appropriate for ED discharge. Limkakeng et al6 combined the Goldman criteria with cardiac troponin analysis in an attempt to increase the rule set’s ability to identify those low-risk patients appropriate for discharge. Unfortunately, the combination of the Goldman criteria with serum biomarkers in the ED chest pain patient did not identify a subgroup with less than 1% risk for AMI or poor outcome within 30 days. In a similar fashion, Chase et al7 attempted to use the TIMI risk score to describe ED chest pain patients in a risk stratification sense. The investigators found that the TIMI risk score correlated very nicely with outcome; unfortunately, the scoring system did not separate patients into discrete risk groups, allowing for the identification of individuals appropriate for emergency department release, recommending that the TIMI risk score should not be used in isolation to determine disposition of ED chest pain patients.

Serum marker analysis, primarily using the troponin assay, is an important diagnostic tool in the chest pain patient suspected of ACS. Certainly, positive biomarkers with the typical rise and fall of AMI are suggestive of acute myocardial infarction; interpretation of these values within the clinical context of the patient’s presentation including the 12-lead ECG allows the EP to establish the diagnosis of either NSTEMI or STEMI; in this setting, elevated serum troponin values are associated with adverse cardiovascular outcome. Little information is available regarding the use of serial troponin testing and ED disposition. Ghaemmaghami and colleagues8 have suggested that negative serial troponin determinations, in the setting of a stable patient with a normal to near-normal ECG, is associated with an extremely low adverse event rate in adult chest pain patients who have competed the “rule out MI” ED evaluation. Such information is of extreme value in this patient population – the low to intermediate chest pain population. ED chest pain patients with undetectable circulating levels of cTnI have very low rates of ACS independent of other clinical variables. In a series of patients with undetectable circulating levels of troponin upon ED presentation and at eight hrs after presentation, there were zero deaths or AMI’s and a 1.8% rate of revascularization at 30 days from time of ED visit. Measurements of highly sensitive troponin in a serial manner when combined with assessment of clinical variables (persistence of ischemic pain, hemodynamic instability, ECG changes) is a very powerful method of early diagnosis and risk assessment in the ED chest pain patient. Such information would afford the EP the ability to evaluate the patient, determine that ACS was not present and discharge the individual for timely follow-up for further risk stratification with stress imaging.

After completion of the ED evaluation, the presence of an alternative, non-coronary diagnosis is considered by some to be an appropriate discriminator in the selection of outpatient management candidates. Yet prior studies have not examined the impact of a “noncardiac” EP impression or of the utility of an obvious, alternative, noncardiac diagnosis as safe and appropriate justifications for ED discharge in the chest pain patient. Miller et al9 asked the question “does the initial EP impression of ‘noncardiac chest pain’ reliably predict patients without ACS?” In this review, if the physician’s initial diagnostic impression was noncardiac chest pain after the medical history, physical examination and initial 12-lead ECG, the investigators entered the patient in the study. Of 17,737 patients enrolled, 2,992 had an initial EP impression of noncardiac chest pain; in this group, 2.8% of patients experienced an adverse cardiac event. This adverse event group was characterized as follows: older, more often male, and with more frequent medical histories of diabetes mellitus, coronary artery disease, and congestive heart failure. The authors correctly concluded that, despite an initial impression of noncardiac chest pain, traditional coronary risk factors or established histories of coronary artery disease or CHF should prompt further consideration of ACS.

Beyond the initial impression, Hollander et al10 compared the 30-day negative event rate in ED chest pain patients who were diagnosed with an alternative, noncardiac diagnosis with a group of similar individuals in whom a definitive diagnosis could not be established. The study enrolled 1,995 ED patients with potential ACS – 4% were ultimately diagnosed with AMI during hospitalization while, at thirty days, 4% required revascularization (4%) and 1% died. Thirty percent of patients were diagnosed with an obvious noncardiac diagnosis. The presence of an alternative noncardiac diagnosis was associated with a reduced risk of myocardial infarction, yet this noncardiac group experienced an elevated rate of negative outcome – meaning that chest pain patients, even without a diagnosed ACS etiology, have reasonably high rates of adverse event. This statement does not translate into the recommendation that “all chest pain patients should be admitted to the hospital;” rather, it indicates that appropriate outpatient follow-up is needed for these patient who obviously do not require inpatient management at the time of ED care.

At initial ED presentation, the chest pain patient is evaluated with the goal of ACS identification; with ACS considered unlikely, the EP then must explore the possibility of significant coronary artery disease. Little information is available regarding the prevalence and severity of coronary artery disease in this patient population. This study11 focused on chest pain patients with non-ischemic ECGs and normal serial troponin values who remained clinically stable over the initial 12 hours of care. Of the group who might be considered candidates for early discharge, 33% of these patients had evidence of coronary artery disease at cardiac catheterization. While this high rate of CAD is troubling, it does not necessarily mandate hospital admission for all these patients; rather, it emphasizes the need for careful ED evaluation and prompt medical follow-up after discharge. And, it is important to recall that the ED-based strategy in the chest pain patient changes over the time course of the emergency department stay. Early in the process, the EP is focusing on the detection and management of ACS, particularly STEMI. As time passes in the ED, the focus shifts partially to a combined strategy of ACS detection (NSTEMI and unstable angina) and significant CAD consideration. With completion of the “rule-out MI” protocol, the EP is finally faced with the consideration – what is the likelihood of significant CAD in this patient? This paper very nicely answers this last question.

Once the ED evaluation is complete, the EP must then consider what disposition is most appropriate: admission to the hospital versus discharge with outpatient follow-up (with or without stress imagining). Lai and fellow investigators12 explored this issue of appropriate discharge after ED evaluation in an observation unit for outpatient risk stratification via exercise stress testing. Three hundred forty-four patients were entered in the study with two patients experiencing fatal out-of-hospital cardiac events; twentyseven subsequent chest pain visits to the emergency department occurred with nine hospital admissions and 10 readmissions to the observation unit. The authors suggested that a negative ED evaluation involving serial electrocardiograms and biomarkers can identify patients at very low risk of short-term cardiac events – thus, appropriately selected patients can be safely discharged for subsequent outpatient testing. Other studies have noted the ability of patients to follow-up in a timely fashion for stress imaging, particularly if the appointment is made at the time of ED care.13

Chest pain patients with recent negative inpatient evaluations for coronary artery disease not infrequently present again to the ED with continued or recurrent discomfort. The most appropriate ED evaluation strategy for these patients is often times difficult to identify. Prina and colleagues14 investigated the outcome of hospitalized patients discharged with a diagnosis of chest pain of undetermined origin; in this study, they identified features in the presentation which would warrant further cardiac evaluation should the patient return to the ED after hospital discharge. Those “return” patients with pre-existing diabetes mellitus, established coronary artery disease, or abnormal 12-lead ECG demonstrated higher risks for adverse cardiac event; patients lacking these features experienced an excellent cardiac outcome.

Thus, there is no easy answer to the question “Which chest pain patients initially suspected of acute coronary syndrome can be safely discharged from the emergency department after an ED-based evaluation?” Clearly, the high and very low suspicion presentations can be managed in relatively straightforward fashion. Yet, the low to intermediate group is troublesome. Relatively younger patients with atypical descriptions of the event and unrevealing ED evaluations likely can be safely discharged with appropriate short-term follow-up. Clearly, a stable patient who remains pain free in the ED with negative serial biomarkers and normal to minimally abnormal ECG can be discharged from the ED for further evaluation on an outpatient basis. Older patients and those individuals with a past history of ischemic heart disease likely require more in-depth evaluations.

The ED evaluation can identify patients who are safely discharged with the assumption that appropriate follow-up will occur in timely fashion.It is very important, however, to understand that the ED evaluation of this subgroup of patients is only the first, or initial, step in the process. Once the EP has considered the initial presentation and completed the ED evaluation, he/she can decide if further inpatient care is necessary; if outpatient management is acceptable, then the patient, primary care physician, and/or cardiologist bear the responsibility for completion of the process initiated in the ED.

References:

1. Sanchis J, Bodí V, Núñez J, et al: Limitations of clinical history for evaluation of patients with acute chest pain, non-diagnostic electrocardiogram, and normal troponin. Am J Cardiol 2008;101:613-7.
2. Schillinger M, Sodeck G, Meron G, et al: Acute chest pain--identification of patients at low risk for coronary events. The impact of symptoms, medical history and risk factors. Wien Klin Wochenschr. 2004;116:83-9.
3. Schillinger M, Domanovits H, Paulis M et al: Clinical signs of pulmonary congestion predict outcome in patients with acute chest pain. Wien Klin Wochenschr. 2002;114:917-22.
4. Christenson J, Innes G, McKnight D, et al: A clinical prediction rule for early discharge of patients with chest pain. Ann Emerg Med. 2006;47:1-10.
5. Marsan RJ Jr, Shaver KJ, Sease KL, et al: Evaluation of a clinical decision rule for young adult patients with chest pain. Acad Emerg Med. 2005;12:26-31.
6. Limkakeng A, Gibler WB, Pollack C, et al: Combination of Goldman risk and initial cardiac troponin I for emergency department chest pain patient risk stratification. Acad Emerg Med. 2001;8:696-702.
7. Chase M, Robey JL, Zogby KE, et al: Prospective validation of the Thrombolysis in Myocardial Infarction Risk Score in the emergency department chest pain population. Ann Emerg Med. 2006;48:252-9.
8. Ghaemmaghami C: Personal Communication. July 2008.
9. Miller CD, Lindsell CJ, Khandelwal S, et al: Is the initial diagnostic impression of “noncardiac chest pain” adequate to exclude cardiac disease? Ann Emerg Med. 2004 Dec;44(6):565-74. Erratum in: Ann Emerg Med. 2005;45:87.
10. Hollander JE, Robey JL, Chase MR, et al: Relationship between a clearcut alternative noncardiac diagnosis and 30-day outcome in emergency department patients with chest pain. Acad Emerg Med. 2007;14:210-5.
11. Hillis GS, Oliner C, O’Neil BJ, et al: Coronary artery disease in patients with chest pain who have low-risk clinical characteristics and negative cardiac troponin I. Am J Emerg Med. 2001;19:118-21.
12. Lai C, Noeller TP, Schmidt K, King P, Emerman CL: Short-term risk after initial observation for chest pain. J Emerg Med. 2003;25:357-62.
13. Richards D, Meshkat N, Chu J, Eva K, Worster A: Emergency department patient compliance with follow-up for outpatient exercise stress testing: a randomized controlled trial. Can J Emerg Med Care 2007; 9:435-40.
14. Prina LD, Decker WW, Weaver AL, et al: Outcome of patients with a final diagnosis of chest pain of undetermined origin admitted under the suspicion of acute coronary syndrome: a report from the Rochester Epidemiology Project. Ann Emerg Med. 2004;43:59-67.