Double-blind, randomized, controlled trials (DB-RCT), if well-designed and -conducted, are widely considered the gold standard in medical research for purposes of establishing causal efficacy. Their logic is compelling: by balancing out all confounding variables through the research design, DB-RCTs are thought to reveal whether a proposed treatment—by virtue of its characteristic constituents—causes therapeutic effects. Many studies on psychedelic-assisted therapy (PAT) follow this ostensible gold standard and use a DB-RCT design. However, several authors have already noted that conducting psychedelic DB-RCTs is particularly challenging: due to the psychoactive effects of psychedelics, participant awareness of condition assignment is likely; this awareness may then interact with response expectancy and experimenter behavior, introducing systematic bias into the trial. For this reason, these authors have suggested ways to rescue DB-RCTs for PAT. The present paper takes a different direction. It argues that we should abandon the DBRCT design as the assumed gold standard in PAT research, because its logic is largely undermined by the intervention(s) in question, and the design in its standard form neglects potentially important aspects of PAT (i.e., extra-pharmacological factors and their interaction(s) with the psychedelic). Doing so opens the door to a more holistic study of PAT, in which DB-RCTs are still useful for certain ends, but no longer considered to produce results that are per se superior to those of other research designs, but complementary.
