BackgroundThe incidence of cardiovascular disease is on the increase in Africa and with it, an increase in the incidence of out-of-hospital cardiac arrest (OHCA). OHCA carries a high mortality, especially in low-resource settings. Interventions to treat OHCA, such as mass cardiopulmonary resuscitation (CPR) training campaigns are costly. One cost-effective and scalable intervention is telephone-guided bystander CPR (tCPR). Little data exists regarding the quality of tCPR. This study aimed to determine quality of tCPR in untrained members of the public. Participants were also asked to provide their views on the understandability of the tCPR instructions. MethodsThis study followed a prospective, simulation-based observational study design. Adult laypeople who have not had previous CPR training were recruited at public CPR training events and asked to perform CPR on a manikin. Quality was assessed in terms of hand placement, compression rate, compression depth, chest recoil, and chest exposure. tCPR instructions were provided by a trained medical provider, via loudspeaker. Participants were also asked to complete a short questionnaire afterwards, detailing the understandability of the tCPR instructions. Data were analysed descriptively and compared to recommended quality guidance. ResultsFifty participants were enrolled. Hand placement was accurate in 74 % (n = 37) of participants, while compression depth and chest recoil only had compliance in 20 % (n = 10) and 24 % (n = 12) of participants, respectively. The mean compression rate was within guidelines in just under half (48 %, n = 24) of all participants. Only 20 (40 %) participants exposed the manikin's chest. Only 46 % (n = 23) of participants felt that the overall descriptions offered during the tCPR guidance were understandable, while 80 % (n = 40) and 36 % (n = 18) felt that the instructions on hand placement and compression rate were understandable, respectively. Lastly, 94 % (n = 47) of participants agreed that they would be more likely to perform bystander CPR if they were provided with tCPR. ConclusionThe quality of CPR performed by laypersons is generally suboptimal and this may affect patient outcomes. There is an urgent need to develop more understandable tCPR algorithms that may encourage bystanders to start CPR and optimise its quality.