Multiple times every shift, emergency physicians evaluate patients with chest pain. It is our disease. It defines our specialty. As a chief complaint, it means considering some of the most serious diseases we see, such as acute coronary syndrome, aortic dissection, and pulmonary embolism, among others. The latter two can be ruled in or out unequivocally and rapidly with CT imaging, though deciding if this is needed can be challenging. Evaluating a patient for ACS, however, can be quite frustrating for everyone involved — physician, hospitalist, cardiologist, and especially the patient — because definitive answers are not easily or rapidly obtainable, and we are left talking with the patient about risks and likelihoods.
We are implored to perform a detailed history and physical exam, and then told that none of the factors has enough reliability to matter. The ECG is diagnostic in cases of STEMI, but far too often is nonspecific. Troponins are good (maybe too good, plaguing us with false positives), but even if undetectable, we are cautioned about lurking ischemia in the form of unstable angina. Time and hindsight are the only definitive ways to arrive at the diagnosis. We give up hope at reaching a diagnosis in the ED for any except the most obvious cases. So we build special areas in our departments or conveyor belts to observation units to watch these patients for a bad outcome, deferring their disposition decisions and plan to others. There must be a better way, and a growing body of research may be pointing the way.
Consider this case. A 40-year-old woman with hypertension experienced 20 minutes of sharp, centrally located chest pain without radiation that occurred while at work that then resolved. She was slightly short of breath and “did not feel well” during the episode, but did not have sweating, nausea, or vomiting. She smokes several cigarettes a day. Her blood pressure is 160/100 mm Hg, but other vital signs are normal and her physical exam is unremarkable. ECG shows nonspecific ST-segment/T-wave changes. Initial troponin is negative.
How would you handle this case?
- Discharge home without further testing or evaluation.
- Discharge home with a stress test within 72 hours.
- Measure troponin level, and discharge home with follow-up if negative.
- Place on observation for serial troponins to rule out myocardial infarction if the stress test is negative.
- Measure troponin two or three times in the ED, and then discharge home with outpatient stress test within 72 hours.
It's not easy to determine if chest pain is caused by cardiac ischemia or myriad other possible diagnoses. There is so much inconsistency between physicians and even for the same physician with different patients who it can feel like you are inventing a new system for every patient.
Identifying Low Risk
The easiest way to shortcut the evaluation for ACS is not to begin the evaluation in no-risk patients. If the chest pain has a clearly identifiable noncardiac cause (e.g., rib fracture or zoster), then don't order the ECG or troponin.
If there is suspicion for ACS and it's on the differential, a myocardial infarction can be diagnosed in the ED if the initial troponin measurement is elevated or if there is a STEMI pattern on ECG (though beware subtle indications and mimics). For others, the impossible task of determining who can be safely discharged arises because there is no acceptable miss rate for myocardial infarction. The diagnosis may require several troponin measurements (up to 12 hours) before becoming elevated. This time delay in diagnosis is why patients undergo prolonged observation in the ED or inpatient unit.
Even after myocardial infarction can be excluded, the guideline committees also prompt us to evaluate for ischemia. This usually involves a provocative test. Ischemic chest pain without myocardial necrosis can occur in unstable and stable angina. Defining the coronary anatomy with coronary angiography or CT coronary angiogram can establish the presence of coronary artery disease. Secondary evidence of coronary atherosclerosis causing ischemic chest pain can also be documented by provocative testing (myocardial perfusion imaging or stress testing). The stress can be exercise or pharmacologic, and the imaging can be an echocardiogram or nuclear. Most centers have moved away from stress ECG testing because of poor performance.
Rapid diagnostic protocols for chest pain have held promise for identifying low-risk patients who may be safely discharged from the ED after a limited initial evaluation, avoiding the need for observation. Several protocols have been developed in Eurasia and Europe using various risk stratification scores, biomarkers (panels, conventional troponin, and high-sensitivity troponin), and ECG. Large validation studies have begun appearing in the literature, which can provide defensible support for their use.
Applicable to Practice
The ADAPT trial is the most applicable to practice in the United States. The study identified low-risk patients with negative conventional troponin at zero and two hours, a nonischemic ECG, and TIMI risk score 0. (Table 1.) The study included 1,975 patients, and 392 (20%) met the low-risk criteria. Only one had a major adverse cardiovascular event (Table 2) over the next 30 days. That is an amazing sensitivity of 99.7 percent.
Patients who are classified as low risk using the ADAPT protocol have such a negligible rate of adverse outcomes that they can be safely discharged from the ED and followed up as an outpatient, forgoing even the mandatory 72-hour stress test. A recent study confirms this good news, which looked at 11,234 patients who had normal vital signs, two normal troponin measurements, and an unconcerning ECG. It found only four patients (0.06%) had an adverse event, two of which were noncardiac and the other two iatrogenic, so you can argue there were actually no adverse events. (JAMA Intern Med 2015 May 17; doi:10.1001/jamainternmed.2015.1674.) A validation study of the ADAPT protocol applied to a United States population found only one missed myocardial infarction in 551 patients (0.18%).
While encouraging, remember that the ADAPT protocol only identified 20 percent of patients who were low risk and eligible for early discharge from the ED. The remaining 80 percent still underwent conventional evaluation and treatment, usually involving observation with serial troponin measurements and provocative testing. Given the very low prevalence of actual disease in patients we evaluate for ACS (15.3%), the specificity would have to drop to achieve a near-perfect sensitivity. In the ADAPT trial, however, the sensitivity was only 23.4 percent. The cost of high sensitivity is a high false-positive rate. We end up evaluating a large number of people without ACS so we don't miss a patient who might suffer a major adverse cardiac event. The positive-predictive value of being classified not low risk is only 19 percent, so one of every five patients you evaluate will suffer an adverse outcome (number needed to test=5).
Of the various accelerated diagnostic studies, the ADAPT protocol is particularly applicable in the United States because it used a conventional troponin I rather than high-sensitivity troponin I or T. When you look at the individual components of ADAPT low-risk score (ECG, TIMI, and troponin), however, you can achieve a sensitivity of 98.3 percent and specificity of 23.5 percent by using just the ECG and TIMI score. This is nearly identical to the full ADAPT score. The addition of troponin does not add significant predictive information about adverse events. To capture the four additional cases needed to increase the sensitivity to 99.7 percent required all 1,975 patients to undergo troponin testing at zero and two hours. That is a number needed to test of more than 500. For perspective, emergency physicians using just their gestalt for ACS without an ECG or troponin have an 86 percent sensitivity and 80 percent specificity.
The goal of accelerated diagnostic pathways is to identify low-risk patients who can be safely discharged from the emergency department and followed as an outpatient. The ADAPT protocol using the TIMI risk score, ECG, and troponin is highly sensitive at predicting adverse events, but suffers from a large false-positive rate.
Share this article on Twitter and Facebook.
Access the links in EMN by reading this on our website or in our free iPad app, both available at www.EM-News.com.
Comments? Write to us at firstname.lastname@example.org.Copyright © 2015 Wolters Kluwer Health, Inc. All rights reserved.