From the Journals

Half of psychiatry, psychology trial abstracts contain spin


 

FROM BMJ EVIDENCE-BASED MEDICINE

About half of papers in a sample of psychiatry and psychology journals have evidence of “spin” in the abstract, according to an analysis published in BMJ Evidence-Based Medicine.

Samuel Jellison, a third-year medical student at Oklahoma State University, Tulsa, and coauthors wrote that the results of randomized, controlled trials should be reported objectively because of their importance for psychiatry clinical practice. However, given that researchers were allowed more latitude in the abstract of a paper to highlight certain results or conclusions, the abstract might not accurately represent the findings of the study.

To evaluate this, the authors investigated the use of spin in abstracts, which they defined as “‘use of specific reporting strategies, from whatever motive, to highlight that the experimental treatment is beneficial, despite a statistically nonsignificant difference for the primary outcome, or to distract the reader from statistically nonsignificant results.”

They analyzed 116 randomized, controlled trials of interventions where there was a nonsignificant primary endpoint and found that 56% of those contained spin in the abstract. Spin was evident in 2% of publication titles, 21% of abstract results sections, and 49.1% of abstract conclusions sections. In 15% of trials, spin was found both in the results and the conclusions.

Spin was more common in trials of pharmacologic treatments, compared with nonpharmacologic treatments. However, the study did not find a higher level of spin in industry-funded studies, and in fact, spin was more common in publicly funded research.

The most common form of spin was focusing on a statistically significant primary or secondary endpoint while omitting one or more nonsignificant primary endpoints. Other spin strategies included claiming noninferiority or equivalence for a statistically nonsignificant endpoint; using phrases such as “trend towards significance”; or focusing on statistically significant subgroup analyses, such as per protocol instead of intention to treat.

The authors observed that most physicians only read article abstracts, and up to one-quarter of editorial decisions are based on the abstract alone.

“Adding spin to the abstract of an article may mislead physicians who are attempting to draw conclusions about a treatment for patients,” they wrote, while calling for efforts to discourage spin in abstracts.

In an interview, Paul S. Nestadt, MD, said the findings were not surprising.

“The proportion [56%] of psychiatry and psychology abstracts which Jellison et al. found to contain spin is similar to that found in broader studies of all biomedical literature in previous reviews,” said Dr. Nestadt, assistant professor in the department of psychiatry and behavioral sciences at Johns Hopkins University, Baltimore. “It is disheartening that attempts to mislead have become ‘standard of care’ in medical literature, but it is a predictable outcome of increasing competition for shrinking grant funding, awarded partly on the basis of publication history in leading journals that maintain clear publication biases toward positive results.

“As the authors point out, we all share a responsibility to call out spin when we see it, whether in our role as reviewer, editor, coauthor, or as the writers themselves.”

Neither Mr. Jellison nor his coauthors reported funding or conflicts of interest.

SOURCE: Jellison S et al. BMJ Evid Based Med. 2019 Aug 5. doi: 10.1136/bmjebm-2019-111176.

Next Article: