Analyse Dataset with accelarated failure time models

Usage

analyse_aft(level = 0.95, dist = "weibull", alternative = "two.sided")

Arguments

level: confidence level for CI computation
dist: passed to survival::survreg
alternative: alternative hypothesis for the tests "two.sided" or "one.sieded"

Value

an analyse function that returns a list with the elements

p p value of the score test (two.sided) or the Wald test (one.sided)
alternative the alternative used
coef coefficient for trt
lower lower 95% confidence intervall boundary for the coefficient
upperlower 95% confidence intervall boundary for the coefficient
CI_level the CI level used
N_pat number of patients
N_evt number of events

Details

alternative can be "two.sided" for a two sided test of equality of the summary statistic or "one.sided" for a one sided test testing H0: treatment has equal or shorter survival than control vs. H1 treatment has longer survival than control.

Examples

condition <- merge(
  assumptions_delayed_effect(),
  design_fixed_followup(),
  by = NULL
) |>
  head(1)
dat <- generate_delayed_effect(condition)
analyse_aft()(condition, dat)
#> $p
#> [1] 0.01282547
#> 
#> $alternative
#> [1] "two.sided"
#> 
#> $coef
#>       trt 
#> 0.2890074 
#> 
#> $lower
#> [1] 0.0613899
#> 
#> $upper
#> [1] 0.5166248
#> 
#> $CI_level
#> [1] 0.95
#> 
#> $N_pat
#> [1] 300
#> 
#> $N_evt
#> [1] 300
#> 
analyse_aft(dist="lognormal")(condition, dat)
#> $p
#> [1] 0.01436426
#> 
#> $alternative
#> [1] "two.sided"
#> 
#> $coef
#>      trt 
#> 0.350311 
#> 
#> $lower
#> [1] 0.069841
#> 
#> $upper
#> [1] 0.630781
#> 
#> $CI_level
#> [1] 0.95
#> 
#> $N_pat
#> [1] 300
#> 
#> $N_evt
#> [1] 300
#>