Hierarchical Bayesian models for multi-center diagnostic studies. Implements bivariate meta-analysis of diagnostic accuracy with between-study heterogeneity and correlation structures.
Usage
hierarchicalbayes(
tp,
fp,
fn,
tn,
study_id,
study_covariates = NULL,
model_type = "bivariate_normal",
correlation_model = "unstructured",
prior_specification = "weakly_informative",
mcmc_chains = 4,
mcmc_iterations = 20000,
warmup_iterations = 10000,
thin_interval = 1,
adapt_delta = 0.95,
max_treedepth = 15,
credible_level = 0.95,
prediction_interval = TRUE,
cross_validation = FALSE,
posterior_predictive = TRUE,
convergence_diagnostics = TRUE,
model_comparison = TRUE,
forest_plots = TRUE,
sroc_curves = TRUE,
trace_plots = FALSE,
density_plots = TRUE,
pairs_plots = FALSE,
shrinkage_plots = FALSE,
meta_regression = FALSE,
outlier_detection = TRUE,
sensitivity_analysis = FALSE
)Arguments
- tp
True positive counts by study/center
- fp
False positive counts by study/center
- fn
False negative counts by study/center
- tn
True negative counts by study/center
- study_id
Unique identifier for each study/center
- study_covariates
Covariates explaining between-study heterogeneity
- model_type
Statistical model for hierarchical meta-analysis
- correlation_model
Assumed correlation between diagnostic measures
- prior_specification
Method for specifying prior distributions
- mcmc_chains
Parallel chains for convergence assessment
- mcmc_iterations
Total iterations including warmup
- warmup_iterations
Burn-in period for MCMC sampling
- thin_interval
Thinning to reduce autocorrelation
- adapt_delta
Controls step size adaptation in HMC
- max_treedepth
Controls exploration in NUTS algorithm
- credible_level
Level for Bayesian credible intervals
- prediction_interval
Whether to compute predictive distributions
- cross_validation
Whether to perform cross-validation
- posterior_predictive
Whether to perform posterior predictive checking
- convergence_diagnostics
Whether to assess MCMC convergence
- model_comparison
Whether to compute model comparison metrics
- forest_plots
Whether to create forest plots
- sroc_curves
Whether to plot SROC curves
- trace_plots
Whether to create MCMC trace plots
- density_plots
Whether to create posterior density plots
- pairs_plots
Whether to create parameter correlation plots
- shrinkage_plots
Whether to plot shrinkage effects
- meta_regression
Whether to conduct meta-regression
- outlier_detection
Whether to perform outlier analysis
- sensitivity_analysis
Whether to perform sensitivity analysis
Value
A results object containing:
results$instructions | a html | ||||
results$modelSpecification | a table | ||||
results$studySummary | a table | ||||
results$pooledEstimates | a table | ||||
results$hierarchicalParameters | a table | ||||
results$heterogeneityAssessment | a table | ||||
results$correlationAnalysis | a table | ||||
results$metaRegressionResults | a table | ||||
results$outlierAnalysis | a table | ||||
results$convergenceDiagnostics | a table | ||||
results$modelComparison | a table | ||||
results$posteriorPredictive | a table | ||||
results$forestPlotSensitivity | an image | ||||
results$forestPlotSpecificity | an image | ||||
results$srocCurvePlot | an image | ||||
results$tracePlots | an image | ||||
results$densityPlots | an image | ||||
results$pairsPlots | an image | ||||
results$shrinkagePlots | an image | ||||
results$methodExplanation | a html |
Tables can be converted to data frames with asDF or as.data.frame. For example:
results$modelSpecification$asDF
as.data.frame(results$modelSpecification)