Users' guide to detecting misleading claims in research

Editor—Montori et al justifiably draw attention to misleading claims in published clinical trials.¹ This is a serious and common problem.²,³ However, I object to their radical proposal that only the methods and results sections should be read, while the remainder of the paper should be ignored.

The proposal is inconsistent with the title of their paper, “Users' guide to detecting misleading claims in clinical research reports.”¹ How are these misleading claims to be identified if the sections containing them are omitted? Furthermore, anyone capable of critically appraising a trial solely on the basis of the methods and results is unlikely to be fooled by misleading claims in the discussion.

But, more importantly, the proposal would deprive the discerning reader of witnessing the conflict between the results and the unwarranted conclusions. If researchers are willing to disseminate misleading claims then their integrity is brought into question. But if so, then all aspects of the trial—including the methods and results sections—are also brought into question. Misleading claims should be identified and broadcast loudly for they signal doubts about the entire study.

Large scale randomised trials create the ideal conditions for data manipulation. Yet this is merely one of a multitude of problems stemming from a flawed method.³ The remedy? When reading a study, ignore everything except the number of patients recruited. If this is large proceed no further as there is little prospect of encountering any data of genuine benefit to patients.