Cost-effectiveness frontier

Identify the set of non-dominated solutions based on cost and impact.

Usage

frontier(x, convex_hull = FALSE, threshold = Inf)

Arguments

x

Data frame containing `cost` and `impact` columns.

convex_hull

Logical. If TRUE, restricts results to the convex cost-effectiveness frontier by removing extendedly dominated strategies (those with non-monotonic ICERs). If FALSE, returns the full set of non-dominated solutions, which may be more useful for exploring how cost-effective options change across different budget levels.

threshold

Optional numeric value describing the allowed marginal cost per unit of impact gained. After identifying the frontier, solutions with ICERs above this value are discarded. We assume we start with the cheapest dominant solution and iterate checking ICERs moving up in solution cost from there.

Value

A data.frame with dominant, ICER compliant solutions

Details

Solutions are filtered so that no remaining option has higher cost and lower or equal impact than another. After finding these dominant solutions an optional `threshold` removes with ICER above a given threshold.