Skip to main content

Advertisement

Table 4 Probability of success and failures at interim and final analysis with Bayesian Emax model with informative prior (β1 ~ N(−0.5, 1.2*1.2), β2 ~ N(−2.9, 1.2*1.2) and β3 ~ N(3, 2*2), β1, β2, and β3 are parameter estimates of E0, Emax and ED50 respectively) in the half adaptive design (Scenario 3)

From: Design considerations and analysis planning of a phase 2a proof of concept study in rheumatoid arthritis in the presence of possible non-monotonicity

True Dose Response Bayesian Emax model
Early success Early failure Final success Final failure Total Success Mean subjects
Placebo like flat Curve 0.00 0.13 0.08 0.79 8% 63
Emax Curve 0.87 0.00 0.13 0.00 100% 51
Log Linear Curve 0.71 0.00 0.28 0.01 99% 54
U-Shaped Curve 0.18 0.00 0.41 0.40 59% 61