Does it work?

A very brief guide on how to spin results

For reasons that we will not get into now, the two old geezers got involved in the oldest chestnut in the basket: how to present the “right” results—not the wrong results, not the false results—just the right ones, ignoring the ones that are not really very right.

We thought we would kick off with the example of influenza vaccines, based on the Cochrane review in healthy adults, which has been ongoing since 1998. It has been updated six times and finally “stabilised” in 2018. Translated into English, that means no more updates are currently envisaged for the reasons given here.

So, how do inactivated bog-standard influenza vaccines fare in preventing influenza when compared to standard care or real placebo (i.e. saline)? Healthy adults are the ones most capable of mounting an antibody response, so this is the best-performing group. These are real randomised trials with a combined denominator of 71,221 for the outcome of influenza, confirmed by laboratory tests. So, it's real influenza, …