nphPower (version 1.1.0)

cureHR: Control hazard and hazard ratio generation function

Description

Generate control hazard and hazard ratio function used for sample size calculation for cure model

Usage

cureHR(pi0, pi1 = NULL, k0, lmd0, theta, HRType, tchg = NULL)

Value

a list of components including

ctrl_hr

a hazard function for the control group

hr

a hazard ratio function

Arguments

pi0

cure rate for the control group

pi1

cure rate for the treatment group, Default: NULL

k0

shape parameter of the Weibull distribution for the control group

lmd0

rate parameter of the Weibull distribution for the control group

theta

hazard ratio function

HRType

hazard ratio function type. susceptible indicates the hazard ratio function applies to the susceptible only; overall indicates the hazard ratio function applies to the overall population; delayed indicates a cure model with delayed treatment effects. See details.

tchg

delayed timepoint for HRType = delayed, Default: NULL

Details

DETAILS The control group has a survival function of \(S_o0=\pi_0+(1-\pi_0)S_0\),where \(\pi_0\) is the cure rate and \(S_0\) is the survival function for the susceptible population. For HRType = susceptible, the user also needs to provide the cure rate for the experimental group. The provided hazard ratio applies to the susceptible population only. The returned hazard ratio function is the overall one. For HRType=delayed, the returned hazard ratio is derived based on the paper of Wei and Wu (2020) .

References

Wei, J. and Wu, J., 2020. Cancer immunotherapy trial design with cure rate and delayed treatment effect. Statistics in medicine, 39(6), pp.698-708.

See Also

integrate

Examples

p0 <- 0.2; p1 <- 0.3; param <- c(1, log(2)/12);
theta_eg <-function(t){t^0*0.7}
fit <- cureHR(p0, p1, param[1], param[2],theta_eg, HRType="susceptible")
# with delayed effects
theta_eg2 <- function(t){(t<=9)+(t>9)*0.7}
fit2 <- cureHR(p0, p1, param[1], param[2],theta_eg2, HRType="delayed", tchg=9)