Skip to contents

Subgroup-difference (interaction) effects

Source: R/effects.R
interactions.Rd

Reports the CAMS treatment-by-subgroup interaction \(\gamma\). For binary CAMS, \(\mu_B(\pi)-\mu_A(\pi)=\gamma\) for every information fraction \(\pi\), so this quantity is not evaluated at the overall IF. "Prediction" adds new-trial interaction heterogeneity. Named interactions() to match modifiers() and so it does not mask stats::interaction().

Usage

interactions(
  fit,
  at = NULL,
  ifrac = NULL,
  type = c("Overall", "Prediction"),
  prob = NULL,
  max_support = NULL,
  summary = TRUE,
  contrast_sep = NULL,
  exponentiate = NULL,
  shrinkage = FALSE,
  contribution
)

Arguments

fit

A interactmeta from cams().

at, ifrac, prob, max_support, summary, contrast_sep, exponentiate, shrinkage,

contribution As in modifiers(). at / ifrac are unused for the pooled interaction \(\gamma\); they are retained for API compatibility and are used when shrinkage = TRUE.

type

One or both of "Overall" and "Prediction" when shrinkage = FALSE. For CAMS, "Overall" is the fixed interaction \(\gamma\); "Prediction" adds new-trial random effects. For swada(), "Prediction" is the SWADA predictive interval (common study weights on \(g_j\) plus interaction \(\hat\tau_g^2\)); DA / AD accept "Overall" only.

Value

A interactmeta_interaction_effects object analogous to modifiers(), with an interaction column instead of subgroup.

See also

modifiers(), shrinkage(), combined(), forestplot()