Additive component network meta-analysis

Usage

network_component_meta(
  object,
  inactive = NULL,
  sep_components = "+",
  C_matrix = NULL,
  level = 0.95
)

Arguments

object: A fitted nma_mars object.
inactive: Optional component with no active effect. Rows containing the inactive component are retained, but the inactive component column is removed from the additive model.
sep_components: Separator used to split treatment names into components. Defaults to "+".
C_matrix: Optional treatment-by-component matrix. Row names must match object$treatments.
level: Confidence level for component and fitted-combination intervals.

Value

A cnma_mars object with component effects, fitted treatment combinations, the C matrix, and additive model diagnostics.

Examples

# \donttest{
cnma_fit <- network_component_meta(nma_fit)
cnma_fit$component_effects
#>   component    estimate  std_error   ci_lower      ci_upper    z_value
#> A         A -0.17134503 0.08696754 -0.3417983 -0.0008917851 -1.9702182
#> B         B  0.02339181 0.08283477 -0.1389613  0.1857449707  0.2823912
#> C         C  0.14795322 0.09951148 -0.0470857  0.3429921279  1.4867955
#>      p_value
#> A 0.04881337
#> B 0.77764353
#> C 0.13706883
# }