Principled selection of effect modifiers: Comments on 'Matching-adjusted indirect comparisons: Application to time-to-event data'
In this commentary, we raise our concerns about a recent simulation study conducted by Aouni, Gaudel-Dedieu and Sebastien, evaluating the performance of different versions of matching-adjusted indirect comparison (MAIC). The following points are highlighted: (1) making a clear distinction between prognostic and effect-modifying covariates is important; (2) in the anchored setting, MAIC is necessary where there are cross-trial imbalances in effect modifiers; (3) the standard indirect comparison provides greater precision and accuracy than MAIC if there are no effect modifiers in imbalance; (4) while the target estimand of the simulation study is a conditional treatment effect, MAIC targets a marginal or population-average treatment effect; (5) in MAIC, variable selection is a problem of low dimensionality and sparsity-inducing methods like the LASSO may induce bias; and (6) individual studies are underpowered to detect interactions and data-driven approaches do not obviate the necessity for subject matter knowledge when selecting effect modifiers.
READ FULL TEXT