pw {drugdevelopR} | R Documentation |
Probabilty that effect in endpoint one is larger than in endpoint two
Description
This function calculated the probability that the treatment effect in endpoint one (or endpoint x) is larger than in endpoint two (or endpoint y), i.e. P(x>y) = P(x-y>0)
Usage
pw(n2, hr1, hr2, id1, id2, fixed, rho)
Arguments
n2 |
total sample size for phase II; must be even number |
hr1 |
assumed true treatment effect on HR scale for endpoint OS |
hr2 |
assumed true treatment effect on HR scale for endpoint PFS |
id1 |
amount of information for |
id2 |
amount of information for |
fixed |
choose if true treatment effects are fixed or random |
rho |
correlation between the two endpoints |
Details
Z=X-Y is normally distributed with expectation mu_x - mu_y and variance sigma_x + sigma_y- 2 rho sdx sdy
Value
The output of the function pw()
is the probability that endpoint one has a better result than endpoint two