Aim Data on the quality of lay person CPR during emergencies are sparse. We present compression quality data derived from use of a novel CPR feedback device during actual cases prior to ambulance arrival.

Method The credit-card sized CPRcardTM device provided visual indication of compression depth and rate in real-time, and stored the data. Median rate, depth; proportion within targets (100–120/minute; depth:4–6 cm); and flow-time were used to determined compression quality. Bystanders’ emergency performances were compared to their training performances.

Results Median depth during emergencies vs trainings was 39 mm (95% CI: 30 to 49 mm, p=0.028) vs 55 mm (95% CI: 50 to 57 mm, p=0.028); and median rates were 114 cpm (95% CI: 109 to 120 cpm, p=0.104) vs 109 cpm (95% CI: 105 to 112 cpm, p=0.104). Of total emergency vs training delivered compressions, 6% (95% CI: 0% to 49%, p=0.008) vs 63% (95% CI: 56 to 90%, p=0.008) were within target depth; 54% (95% CI: 32% to 79%, p=0.028) vs 94% (95% CI: 81 to 97%, p=0.028) were within target rate. Of the lay bystanders’ during emergencies vs trainings, 4 (50%, p=0.398) vs 5 (71%, p=0.398) met both compression and depth targets. Emergency vs training compression flow-time was 95% (95% CI: 85% to 99%, p=0.099) vs 100% (95% CI: 96 to 100%, p=0.099), respectively. Lay bystanders overall reported positive experience using the card but some expressed reluctance to compress deeply for fear of harming the victims.

Conclusion Training compressions were better quality. The results show the quality of chest compressions delivered by lay bystanders in actual cases, and highlights depth as an area of concern that could improve with training enhancement.

Conflict of interest None

Funding Ministry of Health grant