Another common statistical illusion arising from a drug trial

In a previous video (https://youtu.be/6umArFc-fdc) I showed how easy it is to create a statistical illsusion showing high efficacy rtae for a drug even if it has no effect at all. In this video I show a different statistical illusion whereby a hypothetical drug that is definitely not safe for pregnant women may inevitably be shown to be safe because of a simple 'trick' in the way the data is handled. This example is inspired by a recent real study. Because of youtube censorship rules I cannot say what it is but people can find out from my twitter feed (https://twitter.com/profnfenton) and blog (https://probabilityandlaw.blogspot.com/2022/03/why-are-ukhsa-obfuscating-data-on.html)