Regarding the studied initiatives, training by the mean of instructors has been widely used as an alone or combined intervention. Chest compression-only CPR has been the most frequent content of training, since it is a reasonable alternative to standard CPR (compressions plus ventilations) and can be easier to learn [33].

Instructor-led training can be considered the most common way to disseminate BLS skills. It is usually deployed during a BLS course, but instructors can teach BLS even in targeted contexts (public places, schools, families, workplaces) and in a shorter time. Instructor-led training was the only teaching intervention [15,16,18,20,24] or the first tier of a training program to start peer-to-peer training with the adjunct of self-learning videos [19,21,22].

Interestingly, when intervention training is delivered as a “single shot”, not in a program context and with no specific targets [19,20], or as time-limited initiative with a short follow up [23], the interventions do not seem to impact on outcomes. In general, all targeted and multiple sessions programs were able to increase at least bystander CPR.

When initiatives included schools or family members [15,16,24], survival increased as well as bystander CPR, with the exception of the study by Isbye et al. [23]. Furthermore, programs which engaged students as a second-tier trainers [19,23] with a self-learning kit were able to increase the proportion of trained people (with a ratio of 4.9 and 2.5 respectively).

Media campaigns as a sole initiative [25] or as an adjunct to community training programs [17,21,22] contributed to increase bystander CPR rates, but the evidence on survival is conflicting. Interestingly, despite mass media (television, radio, newspapers, and magazines) being considered the most obvious way to spread messages and promote change in behavior [34], the search was able to only find a small number of published studies evaluating their impact. Furthermore, researchers did not find studies including social media which have recently reached a prominent position in influencing opinions, attitudes, and practices. Focus-designed communication strategies have been successfully implemented in many areas of public health to reach targeted improvement and should be considered even in the field of bystander CPR [35,36]. A specific analysis of studies addressing social media as a tool of public campaigns could add useful information.

researchers found that bundled interventions, targeting different components of the “chain of survival”, can improve OHCA survival [26,27,28,29] probably better than isolated ones; however, it is difficult to isolate the effect of each single component of the program on outcomes. This aspect has been highlighted in a recent systematic review and meta-analysis [37], which investigated the effect of community initiatives on survival to discharge or 30-day survival and on bystander CPR rate, finding an improvement in both outcomes (OR, 1.34; 95% CI, 1.14–1.57; I2 = 33% and OR, 1.28; 95% CI, 1.06–1.54; I2 = 82%, respectively). In that review, the authors classified the interventions in community initiatives alone (five studies) and in initiatives that were combined with changes in health care services (ten studies). Community plus health service interventions were associated with a greater bystander CPR rate compared with community alone initiatives, while survival rate did not differ. However, many of the included studies involved non-health professionals with a duty to respond, such as firefighters and policemen, thus confounding the effect of interventions on their ability to engage occasional bystanders, intended as laypeople with no role in the emergency response systems. Interestingly, a restricted analysis on studies targeting laypeople only confirmed a positive association with increased bystander CPR. Nevertheless, in four out of nine of the studies included in this sub analysis, CPR training was associated with a notification system which, by definition, addresses laypeople who adhere to an organized response system on a volunteer basis, such as a “first responder system”. In contrast, the present scoping review excludes this kind of study, with the purpose of investigating and highlighting the effect of community initiatives on the willingness of general population to provide CPR at an earlier stage than the involvement in a first responders program.