From: The use of percentage change from baseline as an outcome in a controlled trial is statistically inefficient: a simulation study
Correlation
ρ = 0.2
ρ = 0.35
ρ = 0.5
ρ = 0.65
ρ = 0.8
POST
70.5%
FRACTION
45.1%
56.4%
67.0%
82.7%
97.1%
CHANGE
50.7%
59.2%
84.8%
97.7%
ANCOVA
72.3%
76.1%
82.3%
90.8%
98.6%