Someone asked me a very good question regarding interim analysis. If the call for an interim analysis is specified as occurring one-third (and two-third) of the way through a study, what does that mean. In a study with 60 patients lasting a full year, does that mean until 20 (40) patients have arrived, or does it mean until 4 (8) months? Also, if you are counting discrete events, such as adverse events, does it mean one-third of the expected number of adverse events?

One-third does indeed mean one-third of the patients rather than one-third of the time. This can lead to some complexity when accrual is a lot faster or a lot slower than originally planned, and simple rules such as Pocock and O'Brien-Fleming may need to be adapted if the interim analysis is consistent with a time interval, but seriously inconsistent with the sample size. If you are chained to a calendar, you might consider using a more complex approach such as an alpha-spending function.

When counting number of adverse events, it is the number of patients, not the number of patients experiencing an adverse events that defines the stopping rule. This is more by tradition than for any technical reason. The one exception to this rule is the use of survival curves. Sometimes interim analyses for survival curves are defined using number of patients (or better still) the number of patients hours. Sometimes it can use the number of events, instead.

You can find an earlier version of this page on my original website.