Skip to contents

[Stable]

Functions to calculate odds ratios in estimate_odds_ratio().

Usage

or_glm(data, conf_level)

or_clogit(data, conf_level)

Arguments

data

(data.frame)
data frame containing at least the variables rsp and grp, and optionally strata for or_clogit().

conf_level

(proportion)
confidence level of the interval.

Value

A named list of elements or_ci and n_tot.

Functions

  • or_glm(): Estimates the odds ratio based on stats::glm(). Note that there must be exactly 2 groups in data as specified by the grp variable.

  • or_clogit(): estimates the odds ratio based on survival::clogit(). This is done for the whole data set including all groups, since the results are not the same as when doing pairwise comparisons between the groups.

See also

Examples

# Data with 2 groups.
data <- data.frame(
  rsp = as.logical(c(1, 1, 0, 1, 0, 0, 1, 1)),
  grp = letters[c(1, 1, 1, 2, 2, 2, 1, 2)],
  strata = letters[c(1, 2, 1, 2, 2, 2, 1, 2)],
  stringsAsFactors = TRUE
)

# Odds ratio based on glm.
or_glm(data, conf_level = 0.95)
#> $or_ci
#>        est        lcl        ucl 
#> 0.33333333 0.01669735 6.65441589 
#> 
#> $n_tot
#> n_tot 
#>     8 
#> 

# Data with 3 groups.
data <- data.frame(
  rsp = as.logical(c(1, 1, 0, 1, 0, 0, 1, 1, 0, 0, 1, 1, 0, 1, 0, 0, 1, 1, 0, 0)),
  grp = letters[c(1, 1, 1, 2, 2, 2, 3, 3, 3, 3, 1, 1, 1, 2, 2, 2, 3, 3, 3, 3)],
  strata = LETTERS[c(1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2)],
  stringsAsFactors = TRUE
)

# Odds ratio based on stratified estimation by conditional logistic regression.
or_clogit(data, conf_level = 0.95)
#> $or_ci
#> $or_ci$b
#>        est        lcl        ucl 
#> 0.28814553 0.02981009 2.78522598 
#> 
#> $or_ci$c
#>       est       lcl       ucl 
#> 0.5367919 0.0673365 4.2791881 
#> 
#> 
#> $n_tot
#> n_tot 
#>    20 
#>