PACBOARD (Probabilistic Analysis Check dashBOARD)
PACBOARD is a web-based tool that helps check and validate health economic models even when their full code isn't available. It uses interactive visuals and statistical modeling to detect errors, explore how inputs affect outcomes, and predict results, making health economic models more transparent and trustworthy.
At a glance
Use when
Validating health economic models when source code is unavailable; enhancing transparency in model reporting; supporting peer review or regulatory assessment of economic models.
Avoid when
Full model code is accessible and direct validation is possible; input data are sparse or not representative of model uncertainty; non-probabilistic models without PSA results are used.
Inputs
Health economic model parameters and outcomes (e.g., costs, effects, utilities, probabilities), typically in tabular or structured data format for probabilistic sensitivity analysis.
Outputs
Interactive visualizations, error detection reports, metamodel fit statistics, predictions of health economic outcomes (e.g., incremental net monetary benefit), and validation test results.
How it works
PACBOARD enables validation of health economic models through standardized tests, interactive visualizations, and metamodeling techniques. It analyzes model parameters and outcomes to detect anomalies (e.g., negative costs, utilities >1), explores parameter-outcome relationships via metamodels, and predicts incremental net monetary benefit. The tool was tested on two mock models with induced errors, successfully identifying all errors and producing accurate metamodel-based predictions compared to original model outputs.
- HTA domains
- Clinical Effectiveness, Costs & Economic Evaluation
- Categories
- AppraisalModel ValidationTransparency
- Assumptions
- The input data represent a sufficient sample from the probabilistic sensitivity analysis of the underlying health economic model; relationships between parameters and outcomes can be approximated using metamodeling techniques.
- Strengths
- Enables validation without access to model code; provides transparent, interactive exploration of model behavior; automatically detects common input errors; supports independent scrutiny of health economic models.
- Limitations
- Relies on quality and completeness of input data; accuracy of metamodel predictions depends on appropriateness of statistical modeling assumptions; limited to models with available probabilistic analysis outputs.
- Also known as
- PACBOARD, Probabilistic Analysis Check dashBOARD
Questions this answers
- › Are the model inputs (e.g., costs, utilities) within valid ranges?
- › How do changes in model parameters affect key outcomes like incremental net monetary benefit?
- › Can a metamodel accurately predict outcomes of the original health economic model?
- › What is the relationship between uncertain parameters and model outputs?
- › Has the model been systematically validated for common errors?
- › Can the model be independently scrutinized without access to its source code?
References & sources
Similar by meaning
Beta record. Generated from the primary source via AI extraction and independent audit, pending final human review.