pretty_km_output: Fancy Table Output of KM (survfit) Fit
In z2thet/MoffittFunctions: Moffitt Functions

Description Usage Arguments Details Value Examples

This function takes a Kaplan-Meier model fit object (from survival::survfit) and calculate survival estimates at a specified time, and Median Survival Estimates. This can be performed on an overall KM fit or a fit including a categorical variable (strata).

pretty_km_output(
  fit,
  time_est = NULL,
  group_name = NULL,
  title_name = NULL,
  surv_est_prefix = "Time",
  surv_est_digits = 2,
  median_est_digits = 1,
  output_type = NULL
)

`fit`	survfit object (with or without single strata variable)
`time_est`	numerical vector of time estimates. If NULL (default) no time estimates are calculated
`group_name`	strata variable name. If NULL and strata exists then using variable
`title_name`	title to use
`surv_est_prefix`	prefix to use in survival estimate names. Default is Time (i.e. Time:5, Time:10,...)
`surv_est_digits`	number of digits to round p values for survival estimates for specified times
`median_est_digits`	number of digits to round p values for Median Survival Estimates
`output_type`	output type, either NULL (default), "latex", or "html" (making special charaters latex friendly)

Currently works with multiple strata in the fit (i.e. survfit(Surv(time, event) ~ x1 + x2)), although level and Group column names may be off.

A tibble with: Name (if provided), Group (if strata variable in fit), Level (if strata variable in fit), Median Estimate, Time:X (Survival estimates for each time provided, if any). In no strata variable tibble is one row, otherwise nrows = number of strata levels.

# Basic linear model example
set.seed(542542522)
ybin <- sample(0:1, 100, replace = TRUE)
ybin2 <- sample(0:1, 100, replace = TRUE)
y <- rexp(100,.1)
x1 <- factor(sample(LETTERS[1:2],100,replace = TRUE))
x2 <- factor(sample(letters[1:4],100,replace = TRUE))
my_fit <- survival::survfit(survival::Surv(y, ybin) ~ 1)
my_fit2 <- survival::survfit(survival::Surv(y, ybin) ~ x1)
my_fit3 <- survival::survfit(survival::Surv(y, ybin) ~ x2)
my_fit_y2 <- survival::survfit(survival::Surv(y, ybin2) ~ 1)

pretty_km_output(fit = my_fit3, time_est = c(5,10), title_name = 'Overall Fit')

library(dplyr) 
km_info <- bind_rows(
  pretty_km_output(fit = my_fit, time_est = c(5,10), 
        group_name = 'Overall', title_name = 'Overall Survival---ybin'),
  pretty_km_output(fit = my_fit2, time_est = c(5,10), 
        group_name = NULL, title_name = 'Overall Survival---ybin'),
  pretty_km_output(fit = my_fit3, time_est = c(5,10), 
        group_name = 'x2', title_name = 'Overall Survival---ybin'),
  pretty_km_output(fit = my_fit_y2, time_est = c(5,10), 
        group_name = 'Overall', title_name = 'Overall Survival---ybin2'),
) %>% select(Group, Level, everything())

kableExtra::kable(km_info, 'html', caption = 'Survival Percentage Estimates at 5 and 10 Years') %>% 
  kableExtra::collapse_rows(1:2, row_group_label_position = 'stack', headers_to_remove = 1:2)
  
  # Real World Examples
  data(Bladder_Cancer)
surv_obj <-survival::Surv(Bladder_Cancer$Survival_Months, Bladder_Cancer$Vital_Status == 'Dead')
    downstage_fit <- survival::survfit(surv_obj ~ PT0N0, data = Bladder_Cancer)
  
  pretty_km_output(fit = downstage_fit, time_est = c(24, 60), 
       surv_est_prefix = 'Month', surv_est_digits = 3)