Convert arm-level study data to contrast-level network meta-analysis data

Usage

network_pairwise_data(
  data,
  study_id,
  treatment,
  event = NULL,
  n = NULL,
  mean = NULL,
  sd = NULL,
  effect_measure = c("OR", "RR", "RD", "MD"),
  continuity = 0.5
)

Arguments

data: Arm-level data frame.
study_id, treatment: Column names for study and treatment labels.
event, n: Column names for binary events and sample sizes.
mean, sd: Column names for continuous means and standard deviations.
effect_measure: One of "OR", "RR", "RD", or "MD".
continuity: Continuity correction for binary log odds/risk ratios with zero cells.

Value

A contrast-level data frame with columns study, treatment_1, treatment_2, effect, and variance.

Examples

arm_dat <- data.frame(
  study = c("S1", "S1", "S2", "S2"),
  treatment = c("A", "B", "A", "C"),
  events = c(10, 14, 12, 9),
  n = c(100, 100, 90, 95)
)
network_pairwise_data(
  arm_dat,
  study_id = "study",
  treatment = "treatment",
  event = "events",
  n = "n",
  effect_measure = "OR"
)
#>   study treatment_1 treatment_2     effect  variance effect_measure
#> 1    S1           A           B  0.3819346 0.1941676             OR
#> 2    S2           A           C -0.3853205 0.2188929             OR