Risk Table for weighted competing risks

[uriahf]

Hi again

I wonder if there is a way (not necessarily through {ggsurvfit} to extract risk table for weighted CIF?
If I understand correctly {tidycmprsk} does not support weighted CIF and {cmprsk} does not support risk tables.

Thanks!

Uriah

[ddsjoberg]

Is this what you're after? If so, be sure you're using the latest survival pkg from CRAN and the ggsurvfit version from here: #208

In the latest survival update, there was a change to how the number censored was returned. CRAN update coming vvv soon.

library(ggsurvfit)
#> Loading required package: ggplot2
packageVersion("survival")
#> [1] '3.6.4'
packageVersion("ggsurvfit")
#> [1] '1.0.0.9002'

survfit2(Surv(ttdeath, death_cr) ~ 1, tidycmprsk::trial, weights = runif(n = nrow(tidycmprsk::trial))) %>%
  tidy_survfit(times = c(0, 12, 24))
#> # A tibble: 6 × 15
#>    time n.risk n.censor cum.event cum.censor n.event estimate std.error conf.low
#>   <dbl>  <dbl>    <dbl>     <dbl>      <dbl>   <dbl>    <dbl>     <dbl>    <dbl>
#> 1     0   96.2      0        0           0      0      0         0         0    
#> 2    12   84.2      0        6.12        0      6.12   0.0636    0.0197    0.117
#> 3    24   44.6     44.6     25.6        44.6   19.5    0.266     0.0359    0.347
#> 4     0   96.2      0        0           0      0      0         0         0    
#> 5    12   84.2      0        5.88        0      5.88   0.0611    0.0207    0.119
#> 6    24   44.6     44.6     26.0        44.6   20.2    0.271     0.0376    0.355
#> # ℹ 6 more variables: conf.high <dbl>, outcome <chr>, estimate_type <chr>,
#> #   estimate_type_label <chr>, monotonicity_type <chr>, conf.level <dbl>

^{Created on 2024-05-08 with reprex v2.1.0}

[uriahf]

That's perfect Daniel!

I thought that ggsurvfit does not support weighted-CIF, I'm not sure why!
Sorry for the trouble 😅