Letter to the editor: is it valid to break down results from long-term trials in bipolar disorder by polarity of relapses?

When analysing and reporting data from long-term drug trials in bipolar disorder, it has become the standard to break down the outcome into the prevention of mania and the prevention of depression. However, as illustrated by a theoretical example, this approach may confer a potential analysis bias. The point is that when mania or depression, whatever appears first, is considered an endpoint, then an endpoint of mania will exclude an endpoint of depression and vice versa. The risk of such bias is reduced when the time course is taken into consideration in the analysis.