BCG vaccination has beneficial off-target ('non-specific') effects on non-mycobacterial infections. On this premise, trials set out to investigate whether BCG provides off-target protection against COVID-19. A literature search identified eleven randomised 'BCG COVID-19' trials, with conflicting results. Factors including participant age, sex and comorbidities, BCG vaccination strain and dose, outcome measure and duration of follow up likely explain the inconsistencies. The findings of these trials and the differences in their study design are discussed using the PICOT framework to highlight these factors. Understanding how to control these factors to best exploit BCG's off-target effects will be important in designing future trials and intervention strategies.