Global prevalence of basic life support training: A systematic review and meta-analysis.

BACKGROUND AND AIMS: Out-of-hospital cardiac arrest exerts a large disease burden, which may be mitigated by bystander cardiopulmonary resuscitation and automated external defibrillation. We aimed to estimate the global prevalence and distribution of bystander training among laypersons, which are poorly understood, and to identify their determinants.

METHODS: We searched electronic databases for cross-sectional studies reporting the prevalence of bystander training from representative population samples. Pooled prevalence was calculated using random-effects models. Key outcome was cardiopulmonary resuscitation training (training within two-years and those who were ever trained). We explored determinants of interest using subgroup analysis and meta-regression.

RESULTS: 29 studies were included, representing 53,397 laypersons. Among national studies, the prevalence of cardiopulmonary resuscitation training within two-years and among those who were ever trained, and automated external defibrillator training was 10.02% (95% CI 6.60 -14.05), 42.04% (95% CI 30.98-53.28) and 21.08% (95% CI 10.16-34.66) respectively. Subgroup analyses by continent revealed pooled prevalence estimates of 31.58% (95%CI 18.70-46.09), 58.78% (95%CI 42.41-74.21), 18.93 (95%CI 0.00-62.94), 64.97% (95%CI 64.00-65.93), and 50.56% (95%CI 47.57-53.54) in Asia, Europe, Middle East, North America, and Oceania respectively, with significant subgroup differences (p < 0.01). A country's income and cardiopulmonary resuscitation training (ever trained) (p = 0.033) were positively correlated. Similarly, this prevalence was higher among the highly educated (p<0.00001).

CONCLUSIONS: Large regional variation exists in data availability and bystander training prevalence. Socioeconomic status correlated with prevalence of bystander training, and regional disparities were apparent between continents. Bystander training should be promoted, particularly in Asia, Middle East, and low-income regions. Data availability should be encouraged from under-represented regions.