ci: Confidence intervals for CEA estimates

In uo-cmor/cea: Functions for Trial-Based Cost-Effectiveness Analysis

Description

Generate confidence intervals for mean incremental QALYs, costs, net monetary benefit (INMB), and net health benefit (INHB) from a fitted CEA regression model.

Usage

ci(x, outcomes = "INMB", conf = 0.9, type, wtp, estimand = "ATE", ...)

## S3 method for class 'cea_estimate'
ci(
  x,
  outcomes = "INMB",
  conf = 0.9,
  type = if (sim == "parametric") "perc" else "bca",
  wtp,
  estimand = "ATE",
  method = "delta",
  R,
  sim = "parametric",
  ...
)

## S3 method for class 'cea_boot'
ci(
  x,
  outcomes = "INMB",
  conf = 0.9,
  type = if (x$sim == "parametric") "perc" else "bca",
  wtp,
  ...
)

Arguments

x	cea_estimate or cea_boot object. The fitted CEA regression model or bootstrap resampling from the fitted model.
outcomes	A character vector indicating the outcomes to be calculated. Possible values are "QALYs", "Costs", "INMB", or "INHB".
conf	Confidence level of the required intervals.
type	Which type of intervals are required. Possible values are "norm", "basic", "perc", or "bca".
wtp	Willingness-to-pay level for calculation of INMB & INHB.
estimand	String scalar. Whether to calculate the average treatment effect (ATE), average treatment effect on the treated (ATT), or average treatment effect on the controls (ATC). Only used for non-linear models.
...	Passed to boot_cea.
method	Which method to use. Available methods are 'boot' (bootstrap) and 'delta' (delta approximation) are implemented.
R	The number of bootstrap replicates.
sim	A character vector indicating the type of simulation required. Possible values are "ordinary" (the default), "parametric", "balanced", or "permutation".

Details

The 'boot' method uses boot and boot.ci from the boot package. The 'delta' method uses the delta approximation (Oehlert 1992).

References

Oehlert GW. A note on the delta method. Am Stat 1992;46(1):27-9. https://doi.org/10.2307/2684406

uo-cmor/cea documentation built on Dec. 23, 2021, 2:01 p.m.