Definition and rationale for placebo composition: Cross-sectional analysis of randomized trials and protocols published in high-impact medical journals.

BACKGROUND/AIMS: Inadequate description of trial interventions in publications has been repeatedly reported, a problem that extends to the description of placebo controls. Without describing placebo contents, it cannot be assumed that a placebo is inert. Pharmacologically active placebos complicate accurate estimation and interpretation of efficacy and safety data. In this study, we sought to assess whether placebo contents are described in study protocols and publications of trials published in high-impact medical journals.

METHODS: We identified all placebo-controlled randomized clinical trials (RCTs) published in 2016 in Annals of Internal Medicine , The BMJ , the Journal of the American Medical Association (JAMA) , The Lancet , and the New England Journal of Medicine (NEJM) . We included all trials with publicly available study protocols. From journal publications and associated study protocols, we searched and recorded: description of placebo contents; the amount of each placebo ingredient; and investigators' stated rationale for selection of placebo ingredients.

RESULTS: We included 113 placebo-controlled RCTs. Of the 113 trials, placebo content was described in 22 (19.5%) journal publications and 51 (45.1%) study protocols. The amount of each placebo ingredient was described in 15 (13.3%) journal publications and 47 (41.6%) study protocols. None of the journal publications explained the rationale for the choice of placebo ingredients, whereas a rationale was provided in 4 (3.5%) study protocols. The stated rationales were to ensure the placebo was visually indistinguishable from the experimental intervention ( N  = 3) and ensure comparability with a previous study ( N  = 1).

CONCLUSION: There is no accessible record of the composition of placebos for approximately half of high-impact RCTs, even with access to study protocols. This impedes reproducibility and raises unanswerable questions about what effects-beneficial or harmful-the placebo may have had on trial participants, potentially confounding an accurate assessment of the experimental intervention's safety and efficacy. Considering that study protocols are unabridged, detailed documents describing the trial design and methodology, the fact that less than half of the study protocols described the placebo contents raises concerns about clinical trial transparency. To improve the reproducibility and potential of placebo-controlled RCTs to provide reliable evidence on the efficacy and safety profile of drugs and other experimental interventions, more detail regarding placebo contents must be included in trial documents.