Summary of predicted log hazard ratios — summary.ecmeta

Summary statistics for true log hazard ratios (HRs) predicted using ecmeta prediction functions.

# S3 method for ecmeta_jags_prediction
summary(
  object,
  exponentiate = FALSE,
  quantiles = c(0.025, 0.25, 0.5, 0.75, 0.975),
  ...
)

# S3 method for ecmeta_ml_prediction
summary(
  object,
  exponentiate = FALSE,
  quantiles = c(0.025, 0.25, 0.5, 0.75, 0.975),
  ...
)

Arguments

object	A object of the appropriate class.
exponentiate	Logical indicating whether or not to exponentiate the the log HR. Defaults to `FALSE`; if `TRUE`, then HRs are summarized.
...	Additional arguments to pass to `coda::summary.mcmc.list()`.

Value

A data.frame with one row for each log HR comparison: treatment vs. internal control (loghr_trt_ic), treatment vs. external control (loghr_trt_ec), and internal control vs. external control (loghr_ic_ec).

The columns contain summary statistics and quantiles. Statistics reported are the mean (mean), standard deviation (sd), and naive (ignoring autocorrelation across iterations) standard error of the mean. Additional columns contain quantiles of the distribution as specified by quantiles.

If a Bayesian model was used to draw from the posterior distribution of loghr_trt_ec and multiple chains were used, then the column rhat is included as well. This is the R-hat convergence diagnostic with values substantially above 1 indicating a lack of convergence. Note that rhat is only provided for loghr_trt_ec. Diagnostics for loghr_ic_ec should be based on mu and sigma from ecmeta() and loghr_trt_ic is deterministically related to loghr_trt_ec and loghr_ic_ec.