A hospital-based oncology group analysed the way in which serious adverse effects are reported in articles presenting the results of clinical trials of cancer drugs. The authors examined articles published in 5 top journals in oncology, looking for overly general or subjective wording on adverse effects such as "acceptable", "tolerable" or "manageable". Terms of this nature were found in 43% of the articles identified, especially in the abstract, conclusion and discussion sections. The main purpose of these terms seemed to be to divert attention away from or downplay the adverse effects of the drug in question.

Among the 53 articles containing terms that downplayed the drug’s adverse effects, 14 contained no data on severe adverse effects, 22 had no data on life-threatening adverse events, and 2 had no data on deaths. In articles that included data on the incidence of adverse effects, most of the drugs were described as having a "favourable" or "tolerable" adverse effect profile, yet compared with patients receiving placebo or standard treatment, the patients who received the study drug had higher rates of severe adverse effects (30 out of 39 trials), life-threatening adverse effects (26 out of 31) and fatal adverse effects (34 out of 51).

The descriptions of clinical trial results in scientific publications are often biased, downplaying the adverse effects of the treatment under evaluation. The risk of underestimating a treatment's harms is even greater if you only read the abstract and conclusion.

©Prescrire 1 January 2021

Source: "Adverse effects and clinical trial results in oncology: misleading terms and omissions" Prescrire International 2021; 30 (222): 26-27. Subscribers only.

Download the full article