Impact of dispatcher-assisted cardiopulmonary resuscitation on performance of termination of resuscitation criteria.

BACKGROUND: Current Advanced Life Support Termination of Resuscitation (TOR) guidelines suggest when to cease cardiopulmonary resuscitation (CPR). With the significant increase of Dispatch-Assisted CPR (DA-CPR) programs, the impact of DA-CPR on the TOR criteria performance is not clear. METHODS: We conducted a secondary analysis of a prospectively collected registry, the Pan-Asian Resuscitation Outcomes Study. We included patients >15 years old with out-of-hospital cardiac arrest between 2014 and 2017 (after implementation of Singapore's DA-CPR program). We excluded patients with non-cardiac etiology, known do-not-resuscitate status, and healthcare provider bystanders. All cases were collected in accordance to Utstein standards. We evaluated the addition of DA-CPR to the diagnostic performance of TOR criteria using logistic regression modeling. The primary outcome was performance for predicting non-survival at 30 days. Sensitivity, specificity, and positive and negative predictive values were calculated. RESULTS: Of the 6009 cases, 319 (5.3%) were still alive at 30 days. Patients had a mean age of 67.9 (standard deviation 15.7) years and were mostly male and Chinese. Almost half of patients had no bystander CPR. The TOR criteria differentiating DA-CPR from unassisted bystander CPR has a specificity of 94% and predictive value of death of 99%, which was not significantly different from undifferentiated CPR criteria. There were differences in adjusted association with survival between unassisted and DA-CPR. CONCLUSION: Advanced life support TOR criteria retain high specificity and predictive value of death in the context of DA-CPR. Further research should explore the differences between unassisted CPR and DA-CPR to understand differential survival outcomes.

Duke Scholars

Author Catherine Ann Staton Emergency Medicine

Author Alexander Tan Limkakeng Jr. Emergency Medicine

Author Alexander Gordee  

Author Maragatha Kuchibhatla Biostatistics & Bioinformatics, Division of Biostatistics