Figure 1From: The use of percentage change from baseline as an outcome in a controlled trial is statistically inefficient: a simulation studyMathematical description of the four methodsBack to article page