Logistic regression fitting and formatting helpers

Usage

leo_logistic(
  df,
  y_out,
  x_exp,
  x_cov = NULL,
  case_value = 1,
  x_exp_type = "auto",
  verbose = TRUE
)

leo_logistic_format(x, style = "wide")

Arguments

df: Data frame containing the binary outcome, exposure, and covariates.
y_out: Character scalar giving the binary outcome column name.
x_exp: Character scalar giving the exposure column name.
x_cov: NULL, a character vector of covariate column names, or a list of covariate column-name vectors. All models are fitted on the same complete-case cohort defined by y_out, x_exp, and all covariates that appear in x_cov. If x_cov is a named list, those model names are preserved in the output columns.
case_value: Value in y_out that indicates case status.
x_exp_type: Exposure type handling for x_exp. Use "auto" to infer from the input type, "continuous" to force a numeric logistic term, or "categorical" to force factor coding.
verbose: Logical; print progress messages.
x: Result returned by leo_logistic().
style: One of "wide", "tidy", or "gtsummary".

Value

A leo_logistic object containing the default wide table in $result, the underlying tidy rows in $result_tidy, model metadata, and fitted glm objects.

Details

leo_logistic() fits one or more logistic regression models for a single exposure and binary outcome. leo_logistic_format() converts the returned object into a wide summary table, tidy result table, or gtsummary output.

Examples

set.seed(123)
n <- 300
logi_df <- data.frame(
  outcome = rbinom(n, 1, 0.35),
  prs = rnorm(n),
  sex = factor(sample(c("Male", "Female"), n, TRUE)),
  bmi_group = factor(sample(c("Low", "Mid", "High"), n, TRUE), levels = c("Low", "Mid", "High"))
)

model_cont <- list(
  "Crude" = NULL,
  "Model A" = c("sex"),
  "Model B" = c("sex", "bmi_group")
)

res_prs <- leo_logistic(
  df = logi_df, y_out = "outcome", x_exp = "prs",
  x_cov = model_cont, verbose = FALSE
)
res_prs$result
#>   Exposure Outcome Case N Control N      Class Crude OR Crude 95% CI
#> 1      prs outcome    103       197 Continuous    0.969 0.760, 1.235
#>   Crude P value Model A OR Model A 95% CI Model A P value Model B OR
#> 1         0.802       0.97   0.760, 1.235           0.802      0.963
#>   Model B 95% CI Model B P value
#> 1   0.754, 1.228           0.761
leo_logistic_format(res_prs, style = "tidy")
#>     Model Exposure Outcome Level Case N Control N    OR       95% CI P value
#> 1   Crude      prs outcome  <NA>    103       197 0.969 0.760, 1.235   0.802
#> 2 Model A      prs outcome  <NA>    103       197 0.970 0.760, 1.235   0.802
#> 3 Model B      prs outcome  <NA>    103       197 0.963 0.754, 1.228   0.761
#>        Class                              Formula
#> 1 Continuous                   outcome ~ exposure
#> 2 Continuous             outcome ~ exposure + sex
#> 3 Continuous outcome ~ exposure + sex + bmi_group

model_cat <- list(
  "Crude" = NULL,
  "Model A" = c("sex")
)

res_bmi <- leo_logistic(
  df = logi_df, y_out = "outcome", x_exp = "bmi_group",
  x_cov = model_cat, x_exp_type = "categorical", verbose = FALSE
)
res_bmi$result
#>    Exposure Outcome Case N Control N                  Class Crude OR
#> 1 Low (Ref) outcome     31        68 Categorical (3 levels)    1.000
#> 2       Mid outcome     34        66 Categorical (3 levels)    1.130
#> 3      High outcome     38        63 Categorical (3 levels)    1.323
#>   Crude 95% CI Crude P value Model A OR Model A 95% CI Model A P value
#> 1 1.000, 1.000            NA      1.000   1.000, 1.000              NA
#> 2 0.624, 2.050         0.686      1.130   0.624, 2.050           0.687
#> 3 0.738, 2.385         0.348      1.312   0.731, 2.370           0.364

if (requireNamespace("gtsummary", quietly = TRUE) && requireNamespace("broom.helpers", quietly = TRUE)) {
  leo_logistic_format(res_prs, style = "gtsummary")
}