tvcure: Fit a proportional hazards cure model

In gwilliford/tvcure: Estimate Proportional Hazards Cure Models with Time-Varying Data

Description

Fit a proportional hazards cure model using the expectation-maximization algorithm detailed in Peng and Dear (2000) and Sy and Taylor (2000). Time-varying covariates may be incorporated using a "counting" type survival object in the formula.

Usage

tvcure(
  survform,
  cureform,
  data,
  subset = NULL,
  na.action = na.omit,
  offset = NULL,
  link = "logit",
  brglm = T,
  var = T,
  nboot = 100,
  parallel = T,
  emmax = 1000,
  eps = 1e-07
)

Arguments

survform	A formula for the hazard function. Must have a Surv object on the right-hand side of type "right" or "counting".
cureform	A formula for the cure function. Must begin with a tilde followed by variables to include in the equation
data	A data frame containing the data to be used in estimation.
na.action	Specifies how missing data should be handled
offset	Specify an offset variable
link	Link function for the cure equation. Must be either "logit" or "probit".
brglm	Logical value indicating whether bias-reduced logistic regression should be used to estimate the cure equation using brglm2
var	Logical value indicating whether standard errors should be estimated.
nboot	The number of bootstrap samples to draw for estimating standard errors.
parallel	Logical value indicating whether bootstrap replications should be run using parallel processing. This option requires the user to set up a snow object and register it using the doSNOW package.
emmax	Specifies the maximum number of iterations for the EM algorithm.
eps	Convergence criterion

Details

The coefficients in the cure equation are parameterized in terms of the probability of being susceptible to an event. Positive coefficients indicate that a variable is associated with higher susceptibility to experiencing the event of interest (i.e., a lower probability of being cured).

References

Yingwei Peng and Keith B.G. Dear. 2000. “A Nonparametric Mixture Model for Cure Rate Estimation.” Biometrics, 56(1), 237-243.

Sy, Judy P. and Jeremy M.G. Taylor. 2000. “Estimation in a Cox proportional hazards cure model.” Biometrics, 56(1), 227-236.

Code for this package was based in part on smcure

gwilliford/tvcure documentation built on Dec. 20, 2021, 1:51 p.m.