Abstracts Often Do Not Accurately Reflect Trial Results, Study Shows

N. Hawkes
DOI: https://doi.org/10.1136/bmj.f179
2013-01-01
BMJ
Abstract:Bias and “spin” are commonplace in the reporting of trials of breast cancer treatments, a Canadian team has found. When trials failed to meet their headline objectives, authors often found positive results among the small print, and the severity of adverse effects was often understated. Authors and journals needed to do better and readers to be alert to such subtle manipulation, say Ian Tannock and colleagues from Princess Margaret Hospital in Toronto. The team searched for phase III trials of treatments of breast cancer between 1995 and 2011, finding 164 that met their criteria. They focused on the abstract—the only part of study …
What problem does this paper attempt to address?