Survsens: Sensitivity Test for Matched Time to Event Outcomes

In Ngendahimana/sensitivityR5: This package supports implementation of sensitivity analyisis using objects from Match, MatchIt and bmatch

Description

This function performs sensitivity analysis for time to event outcomes in observational studies

Usage

Survsens(x, y = NULL, data = NULL, exp = NULL, outcome = NULL, failtime,
  Gamma, alpha = 0.05, Gammainterval, plot_title = NULL)

Arguments

x	name of the matching object. Should be either a Match or matchit object.
data	dataset.
exp	name of treatment variable.
outcome	name of outcome variable.
failtime	name of event time variable.
Gamma	upper bound for sensitivity parameter.
Gammainterval	value by which to increament sensitivity parameter from zero to Gamma
plot_title	main title of your plot.

References

Rosenbum 2011

Examples

data(toy)
psmodel <- glm(treated ~ covA + covB + covC + covD + covE + covF + Asqr + BC + BD, family=binomial(), data=toy)
toy$ps <- psmodel$fitted
toy$linps <- psmodel$linear.predictors
X <- toy$linps ## matching on the linear propensity score
Tr <- as.logical(toy$treated)
Y <- toy$out3.time
match1 <- Match(Y=Y, Tr=Tr, X=X, M = 1, replace=FALSE, ties=FALSE)
match.it <- matchit(treated ~ covA + covB + covC + covD + covE + covF + Asqr + BC + BD, data = toy, method='nearest', ratio=1)
Survsens(x= match.it,data =toy,exp='treated',outcome = 'out2',failtime = 'out3.time',Gamma=2.4,Gammainterval = 0.01,alpha = 0.05,plot_title = 'Time To Event Outcome Sensitivity Plot')

Ngendahimana/sensitivityR5 documentation built on June 24, 2020, 4:09 a.m.