In this paper, we develop and implement a general sensitivity analysis methodology for drawing inference about subgroup effects in a two-arm randomized trial when subgroup status is only known for a non-random sample in one of the trial arms. The methodology is developed in the context of the MADIT II study, a randomized trial designed to evaluate the effectiveness of implantable defibrillators on survival.
Scharfstein, Daniel O.; Onicescu, Georgiana; and Goodman, Steven, "ANALYSIS OF SUBGROUP EFFECTS IN RANDOMIZED TRIALS WHEN SUBGROUP MEMBERSHIP IS INFORMATIVELY MISSING: APPLICATION TO THE MADIT II STUDY" (June 2008). Johns Hopkins University, Dept. of Biostatistics Working Papers. Working Paper 170.