Differences in interaction and subgroup-specific effects were observed between randomized and nonrandomized studies in three empirical examples

Main and subgroup-specific effects based on observational data were similar to main and subgroup-specific effects in IPDMAs based on RCTs, yet interactions differed.