Beta-Binomial Diagnostic Accuracy Models — betabinomialdiagnostic • ClinicoPath

Beta-binomial models for overdispersed diagnostic accuracy data. Handles heterogeneity in sensitivity and specificity across studies or centers using beta-binomial distributions with flexible correlation structures.

Usage

betabinomialdiagnostic(
  test_result,
  disease_status,
  study_id = NULL,
  covariates = NULL,
  overdispersion_model = "beta_binomial",
  correlation_structure = "independent",
  estimation_method = "maximum_likelihood",
  heterogeneity_test = TRUE,
  tau_squared_method = "dersimonian_laird",
  confidence_level = 0.95,
  alpha_prior = 1,
  beta_prior = 1,
  alpha_prior_spec = 1,
  beta_prior_spec = 1,
  mcmc_iterations = 10000,
  burnin_iterations = 2000,
  thin_factor = 5,
  convergence_diagnostics = TRUE,
  prediction_intervals = TRUE,
  forest_plot = TRUE,
  summary_roc_curve = TRUE,
  residual_plots = FALSE,
  influence_diagnostics = FALSE,
  publication_bias = FALSE,
  subgroup_analysis = FALSE,
  robust_variance = FALSE,
  finite_sample_correction = TRUE
)

Arguments

test_result: Test result variable
disease_status: Gold standard disease status
study_id: Clustering variable for multi-center studies
covariates: Covariates for explaining between-study heterogeneity
overdispersion_model: Statistical model for overdispersed diagnostic data
correlation_structure: Assumed correlation structure
estimation_method: Method for estimating model parameters
heterogeneity_test: Whether to test for heterogeneity
tau_squared_method: Estimator for τ² (heterogeneity variance)
confidence_level: Confidence level for parameter estimates
alpha_prior: Prior alpha for sensitivity in Bayesian models
beta_prior: Prior beta for sensitivity in Bayesian models
alpha_prior_spec: Prior alpha for specificity in Bayesian models
beta_prior_spec: Prior beta for specificity in Bayesian models
mcmc_iterations: MCMC iterations for posterior sampling
burnin_iterations: MCMC burn-in iterations
thin_factor: Keep every nth MCMC sample
convergence_diagnostics: Whether to check MCMC convergence
prediction_intervals: Whether to compute prediction intervals
forest_plot: Whether to create forest plots
summary_roc_curve: Whether to plot summary ROC curve
residual_plots: Whether to create residual diagnostic plots
influence_diagnostics: Whether to perform influence analysis
publication_bias: Whether to assess publication bias
subgroup_analysis: Whether to conduct subgroup analyses
robust_variance: Whether to use robust standard errors
finite_sample_correction: Whether to apply small sample corrections

Value

A results object containing:

`results$instructions`					a html
`results$modelSummary`					a table
`results$overdispersionParameters`					a table
`results$heterogeneityTests`					a table
`results$studyLevelResults`					a table
`results$pooledEstimates`					a table
`results$correlationResults`					a table
`results$convergenceDiagnostics`					a table
`results$subgroupAnalysis`					a table
`results$influenceAnalysis`					a table
`results$publicationBiasTests`					a table
`results$forestPlotSensitivity`					an image
`results$forestPlotSpecificity`					an image
`results$summaryROCPlot`					an image
`results$residualPlots`					an image
`results$convergencePlots`					an image
`results$funnelPlot`					an image
`results$methodExplanation`					a html

Tables can be converted to data frames with asDF or as.data.frame. For example:

results$modelSummary$asDF

as.data.frame(results$modelSummary)

Examples

# Example: Beta-binomial diagnostic accuracy analysis
betabinomialdiagnostic(
    data = diagnostic_data,
    test_result = test_outcome,
    disease_status = true_disease,
    study_id = study_center,
    overdispersion_model = "beta_binomial"
)