Compute censoring weighted pseudo observations

Assuming that the censoring depends on covariates, the pseudo observations are calculated with the inverse probability of censoring weighted approach, where the censoring probabilities are estimated using Cox's proportional hazards model.

pseudo_coxph(
  formula,
  time,
  cause = 1,
  data,
  type = c("cuminc", "survival", "rmean"),
  formula.censoring = NULL,
  ipcw.method = NULL
)

Arguments

formula

A formula specifying the outcome model. The left hand side must be a Surv object specifying a right censored survival or competing risks outcome. The status indicator, normally 0=alive, 1=dead. Other choices are TRUE/FALSE (TRUE = death) or 1/2 (2=death). For competing risks, the event variable will be a factor, whose first level is treated as censoring. The right hand side is the usual linear combination of covariates.

time

Numeric constant specifying the time at which the cumulative incidence or survival probability effect estimates are desired.

cause

Numeric or character constant specifying the cause indicator of interest.

data

Data frame in which all variables of formula can be interpreted.

type

One of "survival", "cuminc", or "rmean"

formula.censoring

A right-sided formula specifying which variables to use in the model for the censoring distribution.

ipcw.method

Which method to use for calculation of inverse probability of censoring weighted pseudo observations. "binder" the default, uses the number of observations as the denominator, while the "hajek" method uses the sum of the weights as the denominator.

Value

A vector of pseudo observations

See also

coxph

Examples

POi <- pseudo_coxph(Surv(time, status) ~ 1, 1500, cause = 1,
  data = colon, type = "survival", formula.censoring = ~ sex + age,
  ipcw.method = "hajek")

mean(POi)
#> [1] 0.5980773