Skip to contents

Simulated treatment comparison via G-computation

Source: R/stc.R
stc.Rd

Perform an unanchored simulated treatment comparison (STC) with marginal standardization. Fits an outcome regression to the single IPD treatment arm, standardizes its predictions over the comparator-population covariate distribution, and contrasts that outcome with the reported comparator-arm outcome in the same population. It does not standardize either treatment to the index population.

Usage

stc(
  data,
  link = NULL,
  conf_level = 0.95,
  distribution = "weibull",
  n_boot = 200L,
  seed = NULL,
  rmst_horizon = NULL
)

Arguments

data

An mlumr_data object from combine_data(). Integration points from add_integration() are required whenever the outcome model uses a nonlinear link (binomial, Poisson, normal-log, or survival). Substitution of aggregate means is exact only for a normal identity-link model.

Link function. For binomial: "logit" (default), "probit", or "cloglog". For normal: "identity" (default) or "log". For poisson: "log" (default). Ignored for survival. If NULL, uses the canonical default.

conf_level

Confidence level for the interval (default 0.95)

distribution

For family = "survival": the parametric distribution of the survival G-computation (default "weibull"), fitted with flexsurv; the estimand is the restricted mean survival time difference. "mspline" and "pexp" have no parametric analogue and fall back to a Weibull fit with a warning, recorded as approximated = TRUE. Survival STC supports right-censored data without delayed entry.

n_boot

For family = "survival" only: number of bootstrap resamples for the RMST-difference standard error (default 200; 0 gives a point estimate with no interval). Other families use the delta method.

seed

For family = "survival" only: optional integer seed for the bootstrap, making the standard error reproducible. The global random number stream is restored on exit. Ignored for other families.

rmst_horizon

For family = "survival" only: the restriction time of the RMST difference. Defaults to the largest observed time across both arms; set it explicitly to match an mlumr() fit, whose default can differ. A horizon beyond the observed range extrapolates and warns.

Value

An object of class mlumr_stc. Its separation component records the outcome of the binomial separation check: status is "not_separated" when the exact check ran and found none, "unknown" when only the fitted-value screen ran (it cannot see quasi-complete separation), and "not_applicable" for other outcome models. A separated fit, and a Poisson fit with no events, are refused rather than returned.

Details

For binomial outcomes the result carries the link-scale effect, the standardized and observed event probabilities, the risk difference and the log risk ratio, with delta-method Wald intervals; the observed comparator proportion gets the exact Clopper-Pearson interval, as in naive(). An observed comparator arm with zero or all events uses the pseudo-count (r + 0.5) / (n + 1) in transformed measures. For Poisson outcomes $rd is a rate difference per unit exposure and $estimate the log rate ratio, with a 0.5 continuity correction for a zero comparator count.

Scale note: $estimate (and the binomial $log_rr) is on the link / log scale, where the null is 0. To compare against the natural-scale risk ratio or rate ratio from marginal_effects() (where the null is 1), exponentiate it (e.g. exp(result$estimate)).

Normal-family weighting note: across multiple AgD rows, the normal STC comparator-population prediction and observed mean use sample-size (outcome_n) weights, matching the Bayesian ML-UMR comparator-population estimand. outcome_n is required when there is more than one row; a single row has weight one. The observed comparator-mean variance combines independent, mutually exclusive strata as sum(w^2 * se^2) using normalized population weights.

A GLM is fitted to the IPD, its predictions are averaged over the comparator covariate distribution (the integration points, or the AgD means for the identity-link normal case) on the response scale, and the average is contrasted with the reported comparator outcome, as in Ren et al.'s unanchored STC. Standard errors are first-order delta-method values, conditional on the integration grid and the reported comparator summaries. The estimand is E_B[m_A(X)] - E_B[Y_B] in the comparator population ($rd for binomial, $md for normal); $estimate is the link-scale contrast of the two standardized quantities. Survival STC contrasts the index RMST standardized to the comparator covariates with the RMST of an intercept-only flexsurv::flexsurvreg() fit to the pseudo-IPD. The effect is not transported to the index population; mlumr(model = "relaxed") estimates comparator-specific covariate effects for any other target.

References

Ren S, Ren S, Welton NJ, Strong M (2024). Advancing unanchored simulated treatment comparisons: A novel implementation and simulation study. Research Synthesis Methods, 15(4), 657-670. doi:10.1002/jrsm.1718

Remiro-Azocar A, Heath A, Baio G (2022). Parametric G-computation for compatible indirect treatment comparisons with limited individual patient data. Research Synthesis Methods, 13(6), 716-744. doi:10.1002/jrsm.1565

Examples

if (FALSE) { # \dontrun{
result <- stc(dat)
print(result)
} # }