RACE-CARS: county-wide CPR and EMS program did not raise cardiac-arrest survival in North Carolina

The RACE-CARS trial (Randomized Cluster Evaluation of Cardiac ARrest Systems), presented on 8 November 2026 at the American Heart Association’s Scientific Sessions in Chicago and published simultaneously in the New England Journal of Medicine, tested whether a community-wide program could raise survival after out-of-hospital cardiac arrest.

Design

From July 2021 to June 2026, 60 North Carolina counties (about 8.5 million residents, nearly 80% of the state) were randomized: 31 got the program, 29 continued usual practice. The program combined free community CPR training in schools, workplaces and local government; training 911 dispatchers to recognize arrest and coach callers through CPR; and stronger first-responder defibrillation, including 405 trial-provided defibrillators. The analysis covered 22,459 adults with cardiac arrest (median age 64; 37% women; 33% Black; 29% rural), using the CARES registry. The trial was funded by the NIH/NHLBI.

Results

• Survival to hospital discharge with favorable brain function: about 9% with the program versus 9.4% without — no significant difference.
• Bystander CPR: 44.8% versus 42.3%, not significantly different overall; in an exploratory analysis of the final year, odds of bystander CPR were 33% higher in program counties.
• Defibrillation before EMS arrival: 13.5% versus 15.3%.
• Training events: 5,724, reaching 201,586 people — only 4.8% of residents, varying widely by county.

Principal investigator Christopher Granger of Duke said the late rise in bystander CPR suggests sustained outreach may be needed to show a survival benefit.

Limitations

Counties implemented the program unevenly; 911 training, CPR training and first-responder engagement were partly self-reported; 911 staffing shortages, the COVID-19 pandemic and Hurricane Helene disrupted work. The final-year CPR increase is exploratory. The lead author’s advice remains: learn CPR, call 911 immediately and follow the dispatcher. This is not personal medical advice.