It's All How You "Spin" It: Interpretive Bias in Research Findings in the Obstetrics and Gynecology Literature.

Abstract

Scientific publications can be subject to varying degrees of interpretive bias, also known as spin. The rate of spin in randomized controlled trials (RCTs) with nonsignificant primary outcomes in the general obstetrics and gynecology literature is unknown. A decade (January 2006 through December 2015) of the tables of contents of Obstetrics & Gynecology and the American Journal of Obstetrics & Gynecology were screened, with 503 RCTs identified. Limiting assessment to only parallel-group RCTs with a nonsignificant primary outcome (P≥.05) resulted in the identification of 194 studies. The abstracts of the articles reported the primary outcome in 93% of studies with 79% containing a precision estimate but only 25% noting an effect size. The extent of any type of spin occurred in 43% of abstracts and 50% of the main text. In articles that contained spin in the abstract, the more common types were: emphasizing statistically significant secondary results despite a nonsignificant primary outcome (40%); interpreting statistically nonsignificant primary results as showing treatment equivalence or comparable effectiveness (37%); and emphasizing the beneficial effect of the treatment despite statistically nonsignificant results (15%). Half of parallel-group RCTs with nonsignificant primary outcomes published in the two leading journals in general obstetrics and gynecology contains some level of spin. As readers of the medical literature, we should be aware of the concept of spin, the diversity and heterogeneity of spin in the reporting of conclusions, and its effects, particularly when discussing articles that may change clinical practice.