Chest Pain Rule-Out? For Two ED Scores, It’s a Draw

BALTIMORE, MD—Two risk scores used to predict events in patients presenting to the emergency department with low-risk, acute chest pain are seemingly comparable, though one proved to be more effective at identifying obstructive disease, according to a new retrospective analysis.

Varun Bhasin, MD (Stony Brook University Hospital, NY), who presented his findings comparing the History, Electrocardiogram, Age, Risk factors, and Troponin (HEART) score with the clinical CAD consortium (CADC) model last week at the annual Society of Cardiovascular Computed Tomography (SCCT) meeting, told TCTMD he would encourage physicians and hospitals to use either tool. “There isn't one that I would necessarily recommend over the other,” he said. “It's just important to look at the data and each [patient] always has to be considered on a case-by-case basis, considering what can be used. At times you can absolutely calculate both scores.”

The HEART score was developed about a decade ago in the Netherlands, and using five factors—history, ECG, age, risk factors, and troponin—has been shown to reduce hospital stays and lower costs. The CADC includes age, sex, chest pain type, diabetes, hypertension, dyslipidemia, and smoking and is endorsed by the European Society of Cardiology as potentially leading to decreased use of additional invasive or noninvasive imaging.

Tiberio Frisoli, MD (Henry Ford Hospital, Detroit, MI), who was not involved in the study, told TCTMD that the results underscore the importance of institutions employing at least one of these scoring systems to help stratify risk. “The two scores quite frankly overlap a lot and are much more similar than they are different,” he said. “I would say that it's probably wise that emergency room doctors and providers in general become familiar and facile with one of these tools because chest pain is such a common and expensive emergency room chief complaint.”

At his health system specifically, Frisoli said he has seen growing use of the HEART score, “but I don't think it has anywhere near universal adoption across the country.” Other tools like high-sensitivity troponin testing are gaining traction, he noted, but “having this clinical model is only a helpful adjunctive tool.”

HEART Versus CADC Scores

For the study, Bhasin and colleagues included 1,981 patients without known CAD from their institution who had negative initial troponin and ECG results and calculated both their HEART and CADC model scores. Chest pain was classified as typical, atypical, or non-anginal according to physician report.

Both scores were deemed equally good at predicting which patients had a MACE event within 30 days, but the CADC model was better at identifying obstructive CAD.

Area Under the Curve for Predicting 30-Day MACE and Obstructive CAD

Additionally, among both the 48.3% of patients for whom the CADC model indicated a 5% or lesser probability of obstructive CAD and the 48.9% who had a HEART score of 2 or less, the observed 30-day MACE rate was 0.6%. “Per the literature, a less than 1% missed MACE rate is acceptable,” Bhasin noted during his presentation, yet interestingly all missed MACE events were in females.

“Looking forward to where we go next from this study, one thing would be a gender-specific analysis,” he said. The reason he didn’t include one in this presentation was “simply because we didn't have enough data to conduct that analysis—not enough males who were missed to compare with the females who were missed. And it was only about six female patients that were missed at those thresholds. So, if there was more data, that's something that I'd definitely want to do or something that should be studied in the future.”

Pick Your Preference, But Choose One

For now, Bhasin recommends clinicians go with their preference for picking a score to use. “Physicians should look at the literature and look at studies like this that compare the performance to find out what the best risk score is,” he said. “They're both quite effective. The clinical CAD consortium had that higher discriminative power for the prediction of obstructive coronary disease compared to the HEART score, but clinically, that statistical difference may not be something that compromises your ability to use the HEART score.”

Frisoli said he prefers the HEART score. “I think it's a little easier to use because there are fewer variables in it, but it seems to me at first glance that the CADC score does not seem to be overly difficult.” With the modified HEART score—a version that eliminates the troponins—"all you're really doing is giving a point from 1 to 3 for history, ECG, age, and risk factors, so that can literally be calculated in under 60 seconds. I'm going to assume that the CADC model can probably be done that quickly as well in somebody that knows how to use it.”

Whether the addition of gender to the CADC model is the reason for its “slightly better” performance in predicting obstructive disease “would be interesting to see,” Frisoli said. Regardless, “I think it's yet another example of how these easy-to-use, noninvasive, inexpensive tools can be quite helpful in busy emergency rooms in risk stratifying which chest pain patients are truly low-risk and can be managed conservatively with discharge and outpatient follow-up versus which ones deserve more careful, short-term attention with a stress test or some other investigation such as a cardiology consultation.”

He said he would ultimately like to see “a very, very large prospective outcomes trial” in order to show the true effectiveness of these risk scores, but “unfortunately, it may be very difficult to find funding for a trial that would require so many patients.”

Nevertheless, “the more retrospective registry data we have, that can only be helpful,” Frisoli said. “The data supporting the use of these scores frankly is already quite good.”

Bhasin noted, in response to a question following his presentation, that coronary artery calcium scoring was not included in this patient series.