Heart Rate And QRS Duration Predict Immediate Outcome In Pulseless Electrical Activity (PEA) During In-hospital Cardiac Arrest (IHCA)

Introduction: PEA is commonly observed in in-hospital cardiac arrest. ROSC is currently the only indication of treatment response. Studies suggest that QRS duration (QRSd) and heart rate (HR) develop differently in patients who obtain ROSC or not. The aim of this study was to assess prospectively how QRSd and HR affect the immediate outcome of patients with PEA. Method: We investigated 327 episodes of IHCA in 298 patients, collected at two US and one Norwegian hospital. We assessed the ECG in 559 segments of PEA, measuring QRSd and HR in pauses of compressions, and noted the clinical state that followed PEA. We investigated the development of HR, QRSd, and transitions to ROSC or noROSC in a joint linear mixed/ time-to-event model, using software R version 4.0.3 with the package ‘JMbayes2’. Results: A HR increase by 50bpm increased the intensity (“hazard”) of gaining ROSC by 42% (p<0.01), and a 50ms decrease in QRSd increased the intensity of gaining ROSC by 29% (p<0.01). A decreasing HR had no significant impact; however, if QRSd increased by 50ms this increased the probability of transitioning to other lethal states (ventricular tachycardia or fibrillation, asystole, or death) by 24% (p<0.01). Still, several patients experienced increasing QRSd before obtaining ROSC. The figure shows an example of how QRSd (lower left) and HR (upper left) developed in one PEA segment overlaid the linear model (blue line). The estimated probability of obtaining ROSC (green line with 95% CI) or transitioning to other arrest states (red line) over the next 4 minutes are displayed to the right. This patient obtained ROSC at approx. 7min. Conclusion: HR and QRSd conveys information of the immediate outcome (ROSC/noROSC) in PEA. This may guide the team in their efforts, and possibly allow for individual tailoring of treatment, e.g., by delaying a (potentially harmful) epinephrine administration. Due to low specificity, the model cannot support termination of resuscitation.