design.tournoux {ph2hetero} | R Documentation |
Calculates Optimal 2-stage Phase II designs given by Caroline Tournoux-Facon (2011).
design.tournoux(alpha=0.05,beta=0.2,p0n,p0p,p1n,p1p,w=1,gamma=0.6)
alpha |
Type I error. |
beta |
Type II error. |
p0n |
Max unacceptable response rate for "negative" subgroup. |
p0p |
Max unacceptable response rate for "positive" subgroup. |
p1n |
Min acceptable response rate for "negative" subgroup. |
p1p |
Min acceptable response rate for "positive" subgroup. |
w |
Ratio between the two subgroups \frac{N_{pos}}{N_{neg}}. |
gamma |
Level of the heterogeneity test. |
Returns a data.frame object which components are number of patients, error-rates and power-levels for each stage. The first line of the data.frame corresponds to the arguments for an One-stage Fleming design and the others are two-stages Fleming designs with :
- Non-stratified heterogeneous Fleming two-stage design when psi=0.
- Stratified adaptive Fleming two-stage with identification of heterogeneity of responses in favour of "negative" subgroup when psi=1.
- Stratified adaptive Fleming two-stage with identification of heterogeneity of responses in favour of "positive" subgroup when psi=2.
Patrick Sfumato and Bastien Cabarrou.
Tournoux-Facon C, De Rycke Y and Tubert-Bitter P (2011).Targeting population entering phase III trials: A new stratified adaptive phase II design. Statistics in Medecine 30(8) 801-11.
Tournoux-Facon C, De Rycke Y and Tubert-Bitter P (2011).How a new stratified adaptive phase II design could improve targeting population. Statistics in Medecine 30(13) 1555-62.
out.tournoux<-design.tournoux(p0n=0.15, p0p=0.15, p1n=0.30, p1p=0.25, w=2, gamma=0.6, alpha=0.05, beta=0.1)