The importance of randomised vs non-randomised trials

[Extract] We thank Hertzel Gerstein and colleagues for reminding us of the importance of randomised controlled trials (RCTs). However, RCTs require uncertainty about the benefits of an intervention, and once an intervention has already become health policy, ethical issues with doing an RCT arise. For example, most childhood vaccines were introduced without being tested for their effects on overall health. Now, real-world studies suggest that vaccines could have non-specific effects, with important implications for overall health.2 However, most people would consider testing recommended vaccines in RCTs to be unethical. We have shown how triangulation of multiple sources of evidence with different confounder structures can be used to show causality when RCTs are not feasible