get_surv_prob: Get Survival Probability at Specified Times

In survutils: Utility Functions for Survival Analysis

Description

get_surv_prob retrieves the survival probability at a specific time from a survival curve object from the survival::survfit function. The survival curve object can only have one group.

Usage

get_surv_prob(fit, times)

Arguments

fit	survival::survfit object.
times	Vector of times to lookup survival probabilities.

Value

Vector of survival probabilities based on the input times.

Examples

library("survival")

# Get Survival Probabilities Based on Whole Cohort
fit <- survfit(Surv(time, status) ~ 1, data = colon)
times <- c(100, 200, 300)
get_surv_prob(fit, times)

# Get Survival Probabilities for Each rx Group
library("purrr")
library("dplyr")
library("tidyr")

surv.prob.res <- 
  colon %>%
  split(.$rx) %>%
  map(~ survfit(Surv(time, status) ~ 1, data = .)) %>%
  map(get_surv_prob, times)

surv.prob.res.df <- as_data_frame(surv.prob.res)
colnames(surv.prob.res.df) <- names(surv.prob.res)
surv.prob.res.df <-
  surv.prob.res.df %>%
  mutate(surv_prob_time = times)

gather(surv.prob.res.df, "group", "surv_prob", Obs:`Lev+5FU`)

Example output

[1] 0.9741226 0.9239601 0.8689171

Attaching package: 'dplyr'

The following objects are masked from 'package:stats':

    filter, lag

The following objects are masked from 'package:base':

    intersect, setdiff, setequal, union

Warning message:
`as_data_frame()` is deprecated, use `as_tibble()` (but mind the new semantics).
This warning is displayed once per session. 
# A tibble: 9 x 3
  surv_prob_time group   surv_prob
           <dbl> <chr>       <dbl>
1            100 Obs         0.973
2            200 Obs         0.917
3            300 Obs         0.848
4            100 Lev         0.971
5            200 Lev         0.911
6            300 Lev         0.854
7            100 Lev+5FU     0.979
8            200 Lev+5FU     0.944
9            300 Lev+5FU     0.906

survutils documentation built on May 2, 2019, 6:38 a.m.