Skip to contents

Function for Odds Ratio Table and Plot.

Usage

oddsratio(
  data,
  explanatory,
  outcome,
  outcomeLevel,
  diagnosticPredictor = NULL,
  predictorLevel,
  usePenalized = FALSE,
  showNomogram = FALSE,
  showExplanations = FALSE
)

Arguments

data

The data as a data frame.

explanatory

The explanatory variables to be used in the analysis.

outcome

The outcome variable to be used in the analysis.

outcomeLevel

Specify which outcome level represents the positive case for likelihood ratio calculations. If not specified, the function will use the second level alphabetically.

diagnosticPredictor

Specify the predictor to drive likelihood ratios; must be binary. Defaults to the first explanatory variable.

predictorLevel

Specify which level of the diagnostic predictor represents the positive case.

usePenalized

Use Firth penalized likelihood logistic regression. This is recommended when there is separation (zero cells), small sample sizes, or low events-per-variable.

showNomogram

Display an interactive nomogram for converting pre-test to post-test probabilities using likelihood ratios calculated from the data.

showExplanations

Display educational explanations for each analysis type to help interpret odds ratios, risk ratios, diagnostic test performance, ROC analysis, and likelihood ratios.

Value

A results object containing:

results$todoa html
results$texta html
results$text2a html
results$plotan image
results$oddsRatioExplanationa html
results$riskMeasuresExplanationa html
results$diagnosticTestExplanationa html
results$plot_nomograman image
results$diagnosticMetricsa html
results$nomograma html
results$nomogramAnalysisExplanationa html

Examples

# Basic odds ratio analysis with binary outcome
library(ClinicoPath)
data('histopathology')

# Example 1: Simple odds ratio analysis
result1 <- oddsratio(
    data = histopathology,
    explanatory = c("New Test", "Rater 1"),
    outcome = "Golden Standart",
    outcomeLevel = "1"
)
#> Error in oddsratio(data = histopathology, explanatory = c("New Test",     "Rater 1"), outcome = "Golden Standart", outcomeLevel = "1"): argument "predictorLevel" is missing, with no default

# Example 2: Comprehensive analysis with nomogram
result2 <- oddsratio(
    data = histopathology,
    explanatory = c("New Test", "Rater 1", "Rater 2"),
    outcome = "Golden Standart",
    outcomeLevel = "1",
    showNomogram = TRUE,
    showSummaries = TRUE,
    showExplanations = TRUE
)
#> Error in oddsratio(data = histopathology, explanatory = c("New Test",     "Rater 1", "Rater 2"), outcome = "Golden Standart", outcomeLevel = "1",     showNomogram = TRUE, showSummaries = TRUE, showExplanations = TRUE): unused argument (showSummaries = TRUE)

# Example 3: Analysis with continuous and categorical predictors
# Assuming additional variables in dataset
result3 <- oddsratio(
    data = clinical_data,
    explanatory = c("age", "gender", "treatment"),
    outcome = "mortality",
    outcomeLevel = "Dead",
    showSummaries = TRUE
)
#> Error in oddsratio(data = clinical_data, explanatory = c("age", "gender",     "treatment"), outcome = "mortality", outcomeLevel = "Dead",     showSummaries = TRUE): unused argument (showSummaries = TRUE)