An artificial intelligence (AI) algorithm can predict long-term mortality among patients undergoing valve and/or coronary bypass surgery, according to a new study.

Researchers at the Mayo Clinic (Rochester, MN, USA) conducted a study that included 20,627 patients who underwent valve or coronary bypass surgery between 1993 and 2019, with a left ventricular ejection fraction (LVEF) higher than 35%. Patients were screened using an AI-enhanced electrocardiogram (AI-ECG) to independently detect severe ventricular dysfunction on a preoperative electrocardiography, with the primary end being all-cause mortality following cardiac surgery.

The results showed that 83% of patients had a normal AI-ECG screen, and 17% had an abnormal one; patients with an abnormal AI-ECG screen were older and had more comorbidities. The probability of five and ten survival was 86.2% and 68.2% in those with normal AI-ECG screen, compared to 71.4% and 45.1% in abnormal screening. Abnormal AI-ECG screening was independently associated with higher all-cause mortality, and was consistent in patients with an LVEF of 35-55% and those with an LVEF above 55%. The study was published in the December 2021 issue of Mayo Clinic Proceedings.

“The analysis showed that an abnormal AI screen was associated with a 30% increase in long-term mortality after valve or coronary bypass surgery. This correlation was consistent among patients undergoing valve, coronary bypass, or valve and coronary bypass surgery,” said senior author Mohamad Alkhouli, MD. “For clinicians, this may aid in risk stratification of patients referred for surgery and facilitate shared decision-making.”

LVEF is the measurement of how much blood is being pumped out of the left ventricle of the heart with each contraction, and is usually expressed as a percentage. A normal LVEF ranges from 55-70%. An LVEF of less than 40% may confirm a diagnosis of HF. An EF of less than 35% increases the risk of an arrhythmia that can cause sudden cardiac arrest or death, and an ICD may be recommended for these patients.
 

http://m.allmycbs.com/en/index.asp