An artificial intelligence-enabled electrocardiogram (AI-ECG) was found to effectively identify left ventricular systolic dysfunction (LVSD) in patients who present to the emergency department with dyspnea, according to a study published in Circulation: Arrhythmia and Electrophysiology.

Researchers used a validated AI-ECG algorithm to identify LVSD (defined as left ventricular ejection fraction ≤35%) in 1606 patients (median age 68 years; 47% women; 91% white) with dyspnea who were evaluated in the emergency department at the Mayo Clinic. Patients were included if they had at least 1 standard 12-lead ECG on the date of the visit and an echocardiogram performed within 30 days of presentation.

The AI-ECG algorithm identified LVSD with an area under the receiver operating characteristics curve (AUC) of 0.89 (95% CI, 0.86-0.91) and an accuracy of 85.9%. Sensitivity, specificity, negative predictive value, and positive predictive value were 74%, 87%, 97%, and 40%, respectively.

The AI-ECG was found to identify ejection fraction <50% with an AUC, accuracy, sensitivity, and specificity of 0.85 (95% CI, 0.83-0.88), 86%, 63%, and 91%, respectively. N-terminal pro-B-type natriuretic peptide testing alone at a cut-off of >800 identified LVSD with an AUC of 0.80 (95% CI, 0.76-0.84).


Continue Reading

“The ECG is inexpensive, ubiquitous, painless, quickly obtained, and can be performed with minimal training,” noted the study authors.

Study limitations include the use of International Classification of Diseases diagnosis codes to exclude patients with prior heart failure, which may have led to some patients being missed or inappropriately excluded.

“Our study provides evidence to support real world application of an AI-ECG algorithm in routine clinical practice,” the researchers commented. “The application of an AI-ECG algorithm in the emergency department could improve diagnostic accuracy, facilitate appropriate disposition, and provide an avenue to identify high-risk patients early and link them to appropriate cardiovascular care.”

Reference

Adedinsewo D, Carter RE, Attia Z, et al. An artificial intelligence-enabled ECG algorithm to identify patients with left ventricular systolic dysfunction presenting to the emergency department with dyspnea [published online August 4, 2020]. Circ Arrhythm Electrophysiol. https://doi.org/10.1161/CIRCEP.120.008437