New Risk Models Improve Sudden Cardiac Arrest Prediction

Identifying who is at risk remains difficult, and most out-of-hospital events are fatal, underscoring the need for timely risk stratification. Reliance on low left ventricular ejection fraction alone overlooks many at-risk patients. Researchers have now advanced prediction by mapping warning-symptom patterns and tracking risk trajectories linked to recurrent cardiovascular events.

Investigators at Cedars-Sinai Health Sciences University (Los Angeles, CA, USA) reported two complementary advances. In one study published in Circulation: Arrhythmia and Electrophysiology on July 8, 2026, machine learning analyzed combinations of warning symptoms with clinical history to flag imminent risk. The analysis drew from two long-standing community studies in Oregon and Ventura County, California, comparing 364 people who called 911 with symptoms and survived sudden cardiac arrest to 313 callers with similar symptoms who did not arrest. Shortness of breath paired with coronary artery disease or heart failure distinguished those who went on to arrest. Chest pain combined with coronary artery disease predicted imminent arrest in women, while chest pain with heart failure did so in men.

Warning symptoms most often occurred at least 15 minutes before arrest, allowing time to activate emergency response. Prior work by the same group found that 81% of people delayed calling 911, reducing chances of successful resuscitation by ambulance paramedics. The investigators indicated that these findings could inform risk-prediction algorithms for urgent care and emergency medicine to help minimize such delays.

A second study in the Journal of the American Heart Association on June 23, 2026 used the Observational Study of Cardiac Arrest Risk (O.S.C.A.R.) cohort, which has tracked about 400,000 Los Angeles County residents since 2017. Researchers followed more than 6,700 patients hospitalized for heart failure and more than 2,900 hospitalized for acute coronary syndrome. Recurrent events signaled rising risk: a second coronary artery blockage was associated with more than a threefold increase in sudden cardiac arrest, while a second heart failure hospitalization nearly doubled risk, which increased further with each additional heart failure admission. Results aligned with patterns observed in the Framingham Heart Study, although the heart failure association there did not reach statistical significance.

“This may mean running additional tests. It may mean educating the patient about what cardiac arrest is and the importance of having a family member who knows to call 911 and start CPR immediately if their loved one collapses,” said Kyndaron Reinier, Ph.D., M.P.H., associate director of epidemiology in the Center for Cardiac Arrest Prevention in the Smidt Heart Institute at Cedars-Sinai.

“Near-term and long-term predictions are parallel approaches that can help us move past roadblocks we face in preventing death from this lethal heart event,” said Sumeet Chugh, M.D., vice dean and chief AI health research officer at Cedars-Sinai.

Related Links
Cedars-Sinai Health Sciences University