Evaluation of spin in the abstracts of systematic reviews and meta-analyses focused on treatments of alcohol use disorder

Abstract

ABSTRACT Background: With 14.4 million U.S. adults diagnosed with alcohol use disorder (AUD) annually, effective treatments for combatting this condition are essential. Clinicians are often guided by systematic reviews and meta-analyses – considered the gold standard of research. Spin, a biased way of reporting results, may lead to misinterpretation of research findings, resulting in suboptimal patient care. Objective: Our primary objective was to investigate the presence of spin in the abstracts of systematic reviews of AUD treatments. Methods: After systematically searching MEDLINE and Embase for systematic reviews of AUD treatments, abstracts were evaluated for the nine most severe types of spin. Additional article characteristics were concurrently extracted and study quality was evaluated. Descriptive statistics of spin were calculated and associations between spin and study characteristics were determined through Fisher’s exact and logistic regression. Results: Among 79 included systematic reviews, 44 instances of spin were identified spanning 43% of our sample (34/79). Of the nine forms of spin, eight were found with a majority of instances being “selective reporting of or overemphasis on efficacy outcomes” (13/44, 29.5% of cases). The majority of articles were rated as critically low quality (51/79, 64.6%). No association was found between the presence of spin and extracted study characteristics. Conclusions: Spin was found in more than 40% of systematic review abstracts that evaluated pharmacotherapies in the treatment of AUD. Coupled with the finding that the majority of systematic reviews on the subject were of low quality, increased awareness of spin among physicians may be warranted.