rmda: Risk Model Decision Analysis for Clinical Treatment Policies

An R package by Marshall Brown providing tools to evaluate the clinical utility of risk prediction instruments. It estimates and displays decision curves to compare risk-based treatment policies against 'treat all' and 'treat none' strategies using data from prospective cohort or case-control studies. The package includes functions for bootstrap confidence intervals and cross-validated curve estimation.

Use Cases

• Evaluate the population-level net benefit of a risk-based treatment policy based on decision curve analysis.
• Compare multiple risk prediction instruments for a binary clinical outcome based on their standardized net benefit.
• Assess a risk model's utility for recommending treatment opt-out for low-risk patients, as described for version 1.4.
• Calculate cross-validated decision curves using the provided k-fold cross-validation wrapper function.

Strengths

• Provides specific statistical methods like decision curve analysis and bootstrap confidence intervals.
• Supports analysis for both prospective cohort and case-control study designs, as mentioned in the description.
• Includes version 1.4 functionality for alternative decision framing where treatment is the standard-of-care.

Limitations

• Row count and dataset scale are unknown, which may limit suitability assessment.
• Column-level documentation is absent; field semantics must be inferred after download.
• Last update date is unknown; freshness unverified.

Provenance

Source

Marshall Brown

Collection Method

Software package for statistical analysis.