Pred_Prob: Posterior predictive probability, given 1 observed proportion

In IDDI-BE/PhII_Bayes: Bayesian tools for PhII clinical trial design

Description

Posterior predictive probability for posterior probability outcome at the end of the trial, given currently observed number of patients and successes Definition: Sum(P(X2=i|x1)I[P(Delta>Dcut|x1,X2=i)>PostProb]) for all X2=i,...,n2, with P(Delta>Dcut|x1,X2=i) ~beta(beta_par[1]+x1+i,beta_par[2]+N-x1-i) with prior distribution beta_par with X2~beta-binomial(n2,beta_par[1]+x1,beta_par[2]+n1-x1)

Usage

Pred_Prob(N, n1, x1, beta_par, Dcut, PostProb)

Arguments

N	total number of patients at the end of the study
n1	number of patients currently observed
x1	number of successes currently observed
beta_par	two shape parameters c(alpha,beta) for prior beta distribution
Dcut	Proportion corresponding with some hypothesis
PostProb	Threshold for outcome at the end of the study, in terms of Bayesian posterior probability P(theta>Dcut|x1+x2)>PostProb

Value

Predictive Probability for outcome at the end of the study

References

Lee JJ, Liu DD.A predictive probability design for phase II cancer clinical trialsClinical Trials 2008; 5: 93–106

Examples

## Not run: 
Pred_Prob(N=68,n1=38,x1=27,beta_par=c(1  ,1)  ,Dcut=0.664,PostProb=0.95)
Pred_Prob(N=40,n1=23,x1=16,beta_par=c(0.6,0.4),Dcut=0.6  ,PostProb=0.9 ) # Example Lee 2008, 
p97, Table 1 (PP in text)

## End(Not run)

IDDI-BE/PhII_Bayes documentation built on May 19, 2021, 3:04 p.m.