PSM (Propensity Score Matching)

Use the synthetic data setup first; it defines data and the columns used below. These are advanced direct R examples.

Match observations on propensity scores, then inspect balance and overlap in the matched sample.

psm_result <- causal_psm(data, "Immunotherapy", "Response",
                        confounders = c("Age", "Sex", "Stage", "ECOG"),
                        outcome_type = "binary",
                        match_method = "nearest",  # "nearest", "optimal", "full"
                        caliper = 0.2,
                        ratio = 1)  # 1:1 matching

Read the result

Check the matched population, caliper, ratio, overlap and balance. Matching changes the comparison sample and does not control unmeasured confounding. A binary-exposure survival HR is a limited matched Cox summary; continuous-exposure primary survival estimation is not implemented.

See effect scales and reporting, diagnostics and the current support reference.