Which chest pain patients can go home?

Research on risk algorithms and tests helping to identify unnecessary hospitalizations.

A 50-year-old man with diabetes arrives in the ED, visibly sweating and reporting shooting pains from his left arm to his chest. Although tests come back negative for myocardial infarction (MI), the ED physician decides to admit the patient for further evaluation based on his history and cardiac risk factors.

Similar scenarios are replayed millions of times a year at hospitals across the country, even though several studies have found that many patients admitted for chest pain are not at short-term risk for serious cardiac events. Physicians are often hesitant to rely solely on test results when deciding whether or not to admit and often feel that keeping patients with worrisome symptoms in the hospital is the safest option.

Photo by Thinkstock
Photo by Thinkstock

However, a recent study that reviewed thousands of admissions for chest pain at 3 large teaching hospitals over 5 years suggests that many hospitalizations could be safely avoided if physicians adhered to a simple 3-pronged algorithm for risk stratification. Investigators found that adverse cardiac events during hospitalization were extremely rare among patients admitted after a negative ED evaluation based on 2 negative cardiac troponin tests, normal vital signs, and non-ischemic electrocardiographic (ECG) findings.

“If patients met those 3 criteria, their risk of having a bad outcome over the next 24 to 48 hours was 0.06%,” said the study's lead author, Michael Weinstock, MD, adjunct professor of emergency medicine at The Ohio State University, and chair of the ED at Mt. Carmel St. Ann's in Columbus, Ohio. “Patients often get admitted because they have a concerning story, but the real question is how high is their risk in the short term, and we didn't really know the answer to that before.”

Developing evidence-based practices

The study led by Dr. Weinstock, published in the July JAMA Internal Medicine, highlights the lack of a universally accepted algorithm on which to base decisions about admission for chest pain. While most hospitals use similar methods for initial evaluation, the criteria for admission usually hinge on clinical judgment, and practices vary widely across hospitals.

“The decision of whether or not to admit is complicated because the range of possible diagnoses causing the patient's discomfort is so broad,” said James Januzzi Jr., MD, a cardiologist and professor of medicine at Harvard Medical School in Boston. “While many of the diagnoses are low risk and can be managed safely in the outpatient setting, others are life-threatening and have serious potential consequences for the patient and physician if missed.”

The standard workup in the ED for chest pain typically includes assessment of vital signs, an ECG to test for coronary ischemia, and multiple assays for cardiac troponin levels, considered the gold standard biomarker for diagnosing MI.

After admission, however, there is lack of consensus about which tests are considered most clinically useful and cost-effective in assessing the need for cardiac catheterization.

For example, more hospitals are using computed tomography (CT) angiography, seeing it as a potentially more accurate way of diagnosing coronary artery disease (CAD). However, a recent study found that CT angiography was no more accurate than conventional imaging in selecting patients for invasive management.

In the study, dubbed PROSPECT, patients with acute chest pain admitted to telemetry wards of an inner-city hospital were randomly assigned to undergo CT angiography or radionuclide stress myocardial perfusion imaging (MPI). The percentages of patients receiving cardiac catheterization and no subsequent revascularization after 1 year (that is, catheterization that was likely unnecessary) were similar in the 2 groups, suggesting that both are reasonable diagnostic options.

However, CT angiography conferred other advantages over MPI, said study co-author Paul Huang, MD, an assistant professor of medicine and hospitalist at Montefiore Medical Center/Albert Einstein College of Medicine in New York City. For example, CT angiography accurately diagnosed other significant conditions—including pulmonary embolism, pneumonia, and aortic aneurysm—in about 5% of patients in the study, which was published in the Aug. 4 Annals of Internal Medicine.

“CT angiography has the ability to rule out some other causes of chest pain that could be important and could be missed with stress nuclear studies,” he said. “It also resulted in significantly less radiation and a better patient experience than nuclear stress testing.” He noted that because CT angiography requires contrast dye, it may not be the right option for patients with chronic kidney disease.

Developing standard practices for patients after admission is critical in order to prevent unnecessary testing and minimize length of stay, said Allen Liles Jr., MD, hospital medicine program director at the University of North Carolina Medical Center in Chapel Hill, N.C.

Dr. Liles spearheaded development of a Hospitalist Chest Pain Risk Score tool to assess patients' risk for acute coronary syndrome (ACS) following admission from the ED. The tool uses a point system to classify patients based on known risk factors for CAD and ACS: personal and family history of CAD, atypical or typical angina chest pain, ECG results, and troponin value.

In a retrospective study of 259 patients referred to the UNC hospitalist service for evaluation, the investigators found that the risk score would have identified nearly one-third of patients as low risk and potentially eligible for immediate discharge. After 90 days, none of the patients in the low-risk group had experienced a cardiac event versus 5.8% in the high-risk group. The study findings were published in the Journal of Academic Hospital Medicine in January.

“It's often hard to assess what chest pain means, but we've shown that if you can answer these clinical questions, coupled with a couple of inexpensive lab tests, it's safe to keep that person out of the hospital,” said Dr. Liles. “It used to be standard for these patients to spend a night in the hospital, but we're realizing with more diagnostic capability that a fairly substantial group of patients can be kept out of the hospital and managed in the outpatient setting.”

High-sensitivity tests show promise

Measurement of cardiac troponin levels is a critical component in the initial assessment of chest pain patients. Patients with increasingly elevated concentrations of troponin—defined as exceeding the 99th percentile of values—over several hours are considered to be at high risk for MI and referred for immediate intervention.

In the past, physicians often had to wait 6 hours or more after the initial troponin test for conclusive results, but new high-sensitivity tests have cut that time in half, allowing for earlier intervention. Recently, researchers in Europe validated a 1-hour algorithm using highly sensitive assays for troponin T and I, which are expected to be approved soon for use in the United States.

They studied 870 patients presenting to EDs with suspected acute MI. Physicians were able to accurately rule out or rule in acute MI in 75% of patients after 1 hour based on the results of high-sensitivity troponin T tests, according to results published in Archives of Internal Medicine in 2012. The rule-out criterion was a baseline level of less than 12 ng/L and an absolute change within the first hour of less than 3 ng/L; the rule-in group had a baseline value of 52 ng/L or greater or an absolute change within the first hour of 5 ng/L.

The 25% of patients who could not be ruled in or out for MI based on the criteria in the 1-hour algorithm study were placed in an observational zone, and 19% of those were ultimately diagnosed with MI. These results were validated in a second study with 1,320 patients, published in the Canadian Medical Association Journal in April.

“We also looked at 30-day mortality in the rule-out group and found that it was 0%, proving the safety of this strategy,” said study co-author Maria Rubini, MD, a resident in internal medicine and cardiology at the University Hospital of Basel in Switzerland. “We think it is very safe if you combine this algorithm with detailed clinical assessment, including chest pain characteristics, the 12-lead ECG, and a physical exam.”

In the rule-in zone, the algorithm provided high positive predictive value to help clinicians to rapidly identify patients that would benefit from early coronary angiography. Although the test was not so specific that all ruled-in patients were having a heart attack, it identified other serious conditions requiring coronary angiography, such as Takotsubo cardiomyopathy. “These patients are at high risk of death and/or major arrhythmias and usually require admission to a monitored unit,” Dr. Rubini said.

While high-sensitivity testing has helped with more rapid detection of MI, it should not be used in isolation, and physicians should always keep the differential diagnosis in mind, said Dr. Januzzi. For example, other conditions, such as myocarditis, can lead to a rise in troponin.

“The downside of new high-sensitivity tests is that we now detect heart injury in a broad range of other medical diagnoses that we did not previously suspect were present, such as pneumonia,” he said. “The bottom line is that there is no one magic blood test that will give you a correct diagnosis. You have to link an abnormal troponin test with a patient history that would support a diagnosis of MI.”

However, he acknowledged that the new tests are a valuable new tool in helping physicians detect and diagnose MI earlier.

“It's likely that physicians will admit more people at first because the compulsion with elevated troponin is to be worried,” he said. “But as we begin to understand how to interpret the results correctly, it's likely that outcomes of patients will improve.”

Weighing the benefits and risks

Physicians' compulsion to admit chest pain patients based on unfounded worry was the target of the study in JAMA Internal Medicine.

In addition to finding that low-risk patients were unlikely to have cardiac events, the authors noted that 1 in 164 hospitalized patients have a preventable adverse event that contributes to their death and serious harm from hospitalization is 10- to 20-fold more common. Hospital-acquired infections, pneumonia, falls, and other iatrogenic events are among the risks facing patients after admission. In addition, patients are subjected to costly and potentially unnecessary tests.

“Inadvertent harms may not be part of a physician's risk-benefit analysis because they often don't see what happens to the patient after they're admitted,” said Dr. Weinstock. “But it should be part of their thought process when assessing patients for admission.”

Efficient outpatient follow-up and management are often better options for lower-risk patients, he said. Patients can be discharged with a referral or appointment to see a cardiologist within 48 hours if more tests are needed.

“If I feel that a patient is anxious to leave and they are competent to get follow-up care, I will send them to an outpatient cardiologist,” said Dr. Huang at Montefiore.

It's also important to communicate the risks of hospitalization and discuss treatment options with patients, said Dr. Liles.

“Shared decision making is very powerful,” he said. “If a patient is very concerned about their symptoms, a brief admission for observation might be warranted to provide reassurance. But a lot of patients understand the dangers of iatrogenic harms and unnecessary testing and would prefer to go home if possible.”

Patient preferences should be woven into the decision-making process, agreed Dr. Weinstock.

“Remember to ask what the patient wants to do,” he said. “You should have an honest discussion about what is in the best long-term interests of the patient.”