StratifiedMedicine (version 0.2.2)

param_cox: Parameter Estimation: Cox Regression

Description

For each identified subgroup, fit separate cox regression models. Point-estimates and variability metrics in the overall population are obtained by aggregating subgroup specific results (adaptive weighting or sample size weighting).

Usage

param_cox(Y, A, X, mu_hat, Subgrps, alpha_ovrl, alpha_s,
  combine = "adaptive", ...)

Arguments

Y

The outcome variable. Must be numeric or survival (ex; Surv(time,cens) )

A

Treatment variable. (a=1,...A)

X

Covariate space.

mu_hat

Patient-level estimates (See PLE_models)

Subgrps

Identified subgroups (can be the overall population)

alpha_ovrl

Two-sided alpha level for overall population

alpha_s

Two-sided alpha level at subgroup

combine

For overall population, method of combining subgroup-specific results. Default is "adaptive", "SS" corresponds to sample size weighting.

...

Any additional parameters, not currently passed through.

Value

Data-set with parameter estimates (log hazard ratio) and corresponding variability metrics, for overall and subgroups. Subgrps=0 corresponds to the overall population by default.

• param.dat - Parameter estimates and variability metrics (est=logHR, SE=SE(logHR), LCL/UCL = lower/upper confidence limit on logHR scale, pval = p-value).

References

Andersen, P. and Gill, R. (1982). Cox<U+2019>s regression model for counting processes, a large sample study. Annals of Statistics 10, 1100-1120.

See Also

param_combine

Examples

# NOT RUN {
library(StratifiedMedicine)
# Survival Data #
require(TH.data); require(coin)

# MOB-Weibull Subgroup Model ##
res_weibull = submod_train(Y, A, X, Xtest=X, family="survival",
                            submod="submod_weibull")
plot(res_weibull$mod)

## Parameter-Estimation ##
params = param_cox(Y, A, X, Subgrps = res_weibull$Subgrps.train, alpha_ovrl=0.05,
                   alpha_s=0.05)
params
# }
# NOT RUN {
# }