The influence of time on the accuracy of healthcare personnel to diagnose paediatric cardiac arrest by pulse palpation.

Abstract

To determine time and accuracy diagnosing paediatric cardiac arrest (CA) by pulse palpation. Blinded rescuers (82 nurses, 71 doctors) palpated for a brachial pulse in 17 children (1 day-11 years) with non-pulsatile extracorporeal circulation for CA or cardiac failure. Timed rescuer decisions (pulse present/absent) were compared with non-blinded investigator decisions. CA on 55 occasions was diagnosed by 42 (76%) rescuers in mean (+/-SD) time 30+/-19s. Experienced rescuers diagnosed CA in 25+/-14s, inexperienced rescuers in 37+/-24s (p=0.042). CA absent on 98 occasions was confirmed by 77 (79%) rescuers in 13+/-13s. Experienced rescuers confirmed absent CA in 9+/-5s, inexperienced rescuers in 21+/-19s (p=0.0001). Diagnosis of CA compared to confirmation of absence took longer by all rescuers (p<0.0001), experienced rescuers (p<0.0001) and inexperienced rescuers (p=0.018). Twenty-eight of 33 (85%) experienced doctors diagnosed CA or confirmed absence in 13+/-9s, 49 of 61 (80%) experienced nurses in 15+/-12s, 11 of 21 (52%) inexperienced nurses in 18+/-15s and 31 of 38 (82%) inexperienced doctors in 30+/-24s. Overall accuracy was 78% (95%CI 71-84%), sensitivity 0.76 (95%CI 0.64-0.86) and specificity 0.79 (95%CI 0.69-0.86). Experienced doctors were 85% accurate, inexperienced doctors 82%, experienced nurses 80%, inexperienced nurses 52%. Rescuers diagnosing quickly (<10s) had 90% accuracy, in 11-20s 77% accuracy and in 21-30s 62.5% accuracy (p=0.015). Diagnosis of cardiac arrest by pulse palpation alone is unreliable. At least 30s is required but accuracy and speed are related to clinical experience.