EPsProg_bias: Expected probability of a successful program for bias...

In Sterniii3/drugdevelopR: Utility-Based Optimal Phase II/III Drug Development Planning

Expected probability of a successful program for bias adjustment programs with time-to-event outcomes

Description

To discount for overoptimistic results in phase II when calculating the optimal sample size in phase III, it is necessary to use the following functions, which each describe a specific case:

• EPsProg_L(): calculates the expected probability of a successful for an additive adjustment factor (i.e. adjust the lower bound of the one-sided confidence interval), however the go-decision is not affected by the bias adjustment

• EPsProg_L2(): calculates the expected probability of a successful for an additive adjustment factor (i.e. adjust the lower bound of the one-sided confidence interval) when the go-decision is also affected by the bias adjustment

• EPsProg_R(): calculates the expected probability of a successful for a multiplicative adjustment factor (i.e. use estimate with a retention factor), however the go-decision is not affected by the bias adjustment

• EPsProg_R2(): calculates the expected probability of a successful for a multiplicative adjustment factor (i.e. use estimate with a retention factor) when the go-decision is also affected by the bias adjustment

Usage

EPsProg_L(
  HRgo,
  d2,
  Adj,
  alpha,
  beta,
  step1,
  step2,
  w,
  hr1,
  hr2,
  id1,
  id2,
  fixed
)

EPsProg_L2(
  HRgo,
  d2,
  Adj,
  alpha,
  beta,
  step1,
  step2,
  w,
  hr1,
  hr2,
  id1,
  id2,
  fixed
)

EPsProg_R(
  HRgo,
  d2,
  Adj,
  alpha,
  beta,
  step1,
  step2,
  w,
  hr1,
  hr2,
  id1,
  id2,
  fixed
)

EPsProg_R2(
  HRgo,
  d2,
  Adj,
  alpha,
  beta,
  step1,
  step2,
  w,
  hr1,
  hr2,
  id1,
  id2,
  fixed
)

Arguments

HRgo	threshold value for the go/no-go decision rule
d2	total events for phase II; must be even number
Adj	adjustment parameter
alpha	significance level
beta	1-beta power for calculation of sample size for phase III
step1	lower boundary for effect size
step2	upper boundary for effect size
w	weight for mixture prior distribution
hr1	first assumed true treatment effect on HR scale for prior distribution
hr2	second assumed true treatment effect on HR scale for prior distribution
id1	amount of information for hr1 in terms of number of events
id2	amount of information for hr2 in terms of number of events
fixed	choose if true treatment effects are fixed or random, if TRUE hr1 is used as fixed effect

Value

The output of the functions EPsProg_L(), EPsProg_L2(), EPsProg_R() and EPsProg_R2() is the expected probability of a successful program.

Examples

res <- EPsProg_L(HRgo = 0.8, d2 = 50, Adj = 0.4, 
                           alpha = 0.025, beta = 0.1, 
                           step1 = 1, step2 = 0.95, 
                           w = 0.3, hr1 = 0.69, hr2 = 0.81,
                           id1 = 280, id2 = 420, fixed = FALSE)
          res <- EPsProg_L2(HRgo = 0.8, d2 = 50, Adj = 0.4, 
                           alpha = 0.025, beta = 0.1, 
                           step1 = 1, step2 = 0.95, 
                           w = 0.3, hr1 = 0.69, hr2 = 0.81,
                           id1 = 280, id2 = 420, fixed = FALSE)
          res <- EPsProg_R(HRgo = 0.8, d2 = 50, Adj = 0.9, 
                           alpha = 0.025, beta = 0.1, 
                           step1 = 1, step2 = 0.95, 
                           w = 0.3, hr1 = 0.69, hr2 = 0.81,
                           id1 = 280, id2 = 420, fixed = FALSE)
          res <- EPsProg_R2(HRgo = 0.8, d2 = 50, Adj = 0.9, 
                           alpha = 0.025, beta = 0.1, 
                           step1 = 1, step2 = 0.95, 
                           w = 0.3, hr1 = 0.69, hr2 = 0.81,
                           id1 = 280, id2 = 420, fixed = FALSE)

Sterniii3/drugdevelopR documentation built on Jan. 26, 2024, 6:17 a.m.