Summary

Aims

Materials and methods

Consecutive adult, out-of-hospital cardiac arrests of all causes were studied. One hundred and seventy-six episodes (March 2002–October 2003) without feedback were compared to 108 episodes (October 2003–September 2004) where automatic feedback on CPR was given. Automated verbal and visual feedback was based on measured quality with a prototype defibrillator. Quality of CPR was the main outcome measure and survival was reported as specified in the protocol.

Results

Average compression depth increased from (mean ± S.D.) 34 ± 9 to 38 ± 6 mm (mean difference (95%CI) 4 (2, 6), P < 0.001), and median percentage of compressions with adequate depth (38–51 mm) increased from 24%to 53%(P < 0.001, Mann–Whitney U-test) with feedback. Mean compression rate decreased from 121 ± 18 to 109 ± 12 min⁻¹ (difference −12 (−16, −9), P = 0.001). There were no changes in the mean number of ventilations per minute; 11 ± 5 min⁻¹ versus 11 ± 4 min⁻¹ (difference 0 (−1, 1), P = 0.8) or the fraction of time without chest compressions; 0.48 ± 0.18 versus 0.45 ± 0.17 (difference −0.03 (−0.08, 0.01), P = 0.08). With intention to treat analysis 7/241 control patients were discharged alive (2.9%) versus 5/117 with feedback (4.3%) (OR 1.5 (95%CI; 0.8, 3), P = 0.2). In a logistic regression analysis of all cases, witnessed arrest (OR 4.2 (95%CI; 1.6, 11), P = 0.004) and average compression depth (per mm increase) (OR 1.05 (95%CI; 1.01, 1.09), P = 0.02) were associated with rate of hospital admission.

Conclusions

Automatic feedback improved CPR quality in this prospective non-randomised study of out-of-hospital cardiac arrest. Increased compression depth was associated with increased short-term survival.

Trial registration

ClinicalTrials.gov (NCT00138996), http://www.clinicaltrials.gov/.