Compare parametric survival model distributions

Fit the same parametric survival regression model using multiple distributions and compare model-fit statistics such as AIC and BIC.

Usage

surv_model_compare(
  data,
  time,
  event,
  exposures,
  adjust_for = NULL,
  distributions = c("weibull", "exponential", "lognormal", "loglogistic"),
  digits = 2,
  format = c("flextable", "gt", "tibble"),
  theme = c("minimal")
)

Arguments

data

A data.frame containing survival time, event status, and predictor variables.

time

Survival follow-up time. Quoted and bare names are accepted.

event

Event indicator. Quoted and bare names are accepted. Numeric 0/1, numeric 1/2, logical, character, and factor variables are accepted. For two-level character or factor variables, the second level is treated as the event.

exposures

Character vector of main exposure variable names. Quoted names are recommended in scripts, and bare names are also accepted.

adjust_for

Optional character vector of adjustment variables. These variables are included in every candidate model.

distributions

Parametric survival distributions to compare. Defaults to c("weibull", "exponential", "lognormal", "loglogistic"). Quoted and bare values are accepted. Common spellings such as "log-normal" and "log-logistic" are also accepted.

digits

Number of digits for displayed model statistics.

format

Output format. One of "flextable" (default), "gt", or "tibble".

theme

Table styling preset.

Value

A list of class c("gtregression","surv_model_compare", ...) with elements:

table

A flextable, gt_tbl, or NULL when format = "tibble".

table_body

Tibble with model-fit statistics.

table_display

Display data frame used to render the table.

models

Named list of fitted survreg models.

time,event,exposures,adjust_for,distributions,format,source

Metadata fields.

Details

The same model formula is fitted for each candidate distribution using survival::survreg(). Lower AIC or BIC values indicate better relative model fit among the compared distributions. These statistics should be used with clinical judgment and visual checks; they do not prove that a distribution is scientifically correct.

Examples

lung_data <- data_lungcancer
lung_data$trt <- factor(lung_data$trt, levels = c(1, 2),
                        labels = c("Standard", "Test"))
lung_data$prior <- factor(lung_data$prior, levels = c(0, 10),
                          labels = c("No", "Yes"))

surv_model_compare(
  data = lung_data,
  time = time,
  event = status,
  exposures = c("trt", "celltype", "prior"),
  adjust_for = c("age", "karno")
)

surv_model_compare(
  data = lung_data,
  time = "time",
  event = "status",
  exposures = c(trt, prior),
  distributions = c(weibull, "log-logistic"),
  format = tibble
)
#> # A tibble: 2 × 9
#>   Distribution   AIC   BIC logLik Scale     N Events Best.AIC Best.BIC
#>   <chr>        <dbl> <dbl>  <dbl> <dbl> <dbl>  <dbl> <lgl>    <lgl>   
#> 1 weibull      1503. 1515.  -748. 1.16    137    128 TRUE     TRUE    
#> 2 loglogistic  1508. 1519.  -750. 0.784   137    128 FALSE    FALSE