Cardiac arrest & resuscitation

Early Identification of Recurrent Out-of-hospital Cardiac Arrest Using Clinical ECG Features

ResuscitationPublished retrospective cohort study

Authors' conclusion

Established ECG features achieved good prediction for rearrest with identification improving closer to time of rearrest. Future work leveraging novel features or more advanced artificial intelligence models may further improve prediction.

Study context

Population
adult out-of-hospital cardiac arrest (OHCA) patients who achieved return of spontaneous circulation (ROSC) between 2018–2024analyzed, of the 1,489 OHCA patients achieving ROSC
Participants
1,298 participantsEvents: 528 (40.7%)
Predictors
ECG features (QRS width, standard deviation of NN intervals, root mean square of successive NN interval differences, and percentage of NN intervals that differ by more than 50 milliseconds) within individual 15-second segments throughout a 5-minute window prior to rearrest as well as their temporal trends over the 5-minute analytic window
Predicted outcome
Recurrent cardiac arrest (rearrest)

Model performance

As reported in the abstract

AUROC
Identification highest1 minute prior to rearrest
0.74
Models using temporal trends
0.67
  • Identification declining beyond 3 minutes

Further findings

Bystander CPR

all p<0.001

Survival

all p<0.001

Favorable neurologic outcomes

all p<0.001

Age · Rearrest patients were older
all p<0.001

Recurrent cardiac arrest (rearrest), models using temporal trends

Strongest predictors were heart rate variability measures and QRS width

ECG features in those with rearrest

Lower heart rate, wider QRS, and higher HR variability in those with rearrest

Original publication

Early Identification of Recurrent Out-of-hospital Cardiac Arrest Using Clinical ECG Features. Resuscitation. Published 29 October 2026. doi:10.1016/j.resuscitation.2026.111370

Composed by AI from the published abstract and checked against it. How Briefs are made →

Get them every week

New studies on this topic in My Week and in your Monday e-mail.

We e-mail you a sign-in link and a code. No password, no spam.