View source: R/ATSS_Design_Stage1.R
ATSS_Design_Stage1 | R Documentation |
ATSS_Design_Stage1( ) provides an Adaptive Threshold and Sample Size Simon Design (ATSS Simon) method for Simon's two stage design in oncology trials when the realized sample sizes in the first stage is different from the planned sample sizes in the first stage. When under-enrollment or over-enrollment occurs at the first stage, we identify the design parameters (r1*, r*, n*) based on the actual sample size n1* ar the first stage to satisfy the type I error rate and power. In addition, the ATSS Simon design also satisfies the other criteria as in the originally planned design, such as minimizing the average sample size under the null hypothesis H0.
ATSS_Design_Stage1(p0, p1, n1_star, alpha, beta)
p0 |
Unacceptable efficacy rate |
p1 |
Desirable efficacy rate |
n1_star |
The actual number of patients in stage 1 |
alpha |
Original Type-I error rate |
beta |
Original Type-II error rate |
a data frame includes the Adaptive Threshold and Sample Simon Design interim analysis' adjusted first stage threshold r1*, second stage threshold r*, actual number of patients in the first stage n1*, new design planned two stages' patients n*, attained Type-I error rate and Power, Average sample size under null hypothesis EN(p0) and Probability of early termination under null hypothesis PET(p0).
Yunhe Liu, & Haitao Pan. (2024). Clinical Trial Design Methods for Managing Under- and Over-Enrollment in Simon's Two-Stage Design, Submitted.
# Adaptive Threshold and Sample Size Simon Design interim analysis case 1
ATSS_Design_Stage1(0.05, 0.20, 20, 0.10, 0.10)
# r1* r* n1* n* Type I Power EN(p0) PET(p0)
# ATSS_Design_Stage1 1 3 20 35 0.08 0.901 23.962 0.736
# Adaptive Threshold and Sample Size Simon Design interim analysis case 2
ATSS_Design_Stage1(0.10, 0.30, 18, 0.10, 0.10)
# r1* r* n1* n* Type I Power EN(p0) PET(p0)
# ATSS_Design_Stage1 2 4 18 26 0.099 0.904 20.13 0.734
Add the following code to your website.
For more information on customizing the embed code, read Embedding Snippets.