METHODS: Using a validated questionnaire, 6248 participants were asked to rate their willingness to perform bystander chest compression with mouth-to-mouth ventilation and chest compression-only CPR. Their past familial experiences of receiving cardiopulmonary resuscitation (CPR) and medical help in various cardiac arrest and nonfatal cardiac events were also recorded.

RESULTS: Kruskal-Wallis test with post hoc Dunn's pairwise comparisons showed that the following were significantly more willing to perform CPR with mouth-to-mouth ventilation: familial experience of "nonfatal cardiac events" (mean rank = 447) vs "out-of-hospital cardiac arrest with no CPR" (mean rank = 177), U = 35442.5, z = -2.055, p = 0.04; "in-hospital cardiac arrest and successful CPR" (mean rank = 2955.79) vs "none of these experiences" (mean rank = 2468.38), U = 111903, z = -2.60, p = 0.01; and "in-hospital cardiac arrest with successful CPR" (mean rank = 133.45) vs "out-of-hospital arrest with no CPR" (mean rank = 112.36), U = 4135.5, z = -2.06, p = 0.04. For compression-only CPR, Kruskal-Wallis test with multiple runs of Mann-Whitney U tests showed that "nonfatal cardiac events" group was statistically higher than the group with "none of these experiences" (mean rank = 3061.43 vs 2859.91), U = 1194658, z = -2.588, p = 0.01. The groups of "in-hospital cardiac arrest with successful CPR" and "in-hospital cardiac arrest with transient return of spontaneous circulation" were the most willing groups to perform compression-only CPR.