You might have seen this “Trial Sequential Analysis” in the odd systematic review or meta-analysis, and seen some strange graphs with ‘cumulative z-score boundaries’ in them. And then either quickly closed the PDF or flicked away to catch up with an episode of Peppa Pig.
But fear not. TSA is not as strange an idea as it’s often displayed.
You know that in trials, when they do the interim analysis, they need to account for the fact that many looks at the data might increase the risk of a false positive result?
It’s just that sort of thing, but applied to meta-analysis.
See – simple.
But if that was too stressful …