The presence of spin in systematic reviews focused on diabetic neuropathy: A cross-sectional analysis

Spin-the misrepresentation of a study's actual results-has the potential to alter a clinician's interpretation of the study's findings and therefore could affect patient care. Studies have shown spin frequently occurs in abstracts of systematic reviews from a variety of other medical disorders and specialties.

Our primary aim was to evaluate whether the nine most severe types of spin occurred in systematic review abstracts' concerning diabetic neuropathy treatments. Secondly, we aimed to determine whether spin presence was associated with the methodological quality of a systematic review.

A search of MEDLINE and Embase collected 1297 articles focused on diabetic neuropathy treatments, of which we included 114 systematic reviews for spin assessment. Each included study was evaluated for the nine most severe types of spin as defined by Yachitz et al. The methodological quality of a systematic review was determined by using the AMSTAR-2 instrument. All screening and data extraction were conducted in a masked, duplicate fashion. Since the final sample size of 114 was not sufficiently powered to do multivariable logistic regression, we calculated unadjusted odds ratios which evaluated relationships between spin presence within abstracts and study characteristics.

From the 114 articles reviewed, spin was present in 7.9% of the studies (9/114), with spin type 5: "conclusion claims the beneficial effect of the experimental treatment despite the high risk of bias in the included primary studies" as the most frequent in our study. Spin types 1, 2, 6, and 8 were not identified. No association was observed between the study characteristics and spin presence, including the methodological quality of a systematic review.

Overall, spin is infrequently observed in abstracts of systematic reviews covering diabetic neuropathy treatments. When comparing our results to other fields of medicine, the field of diabetic neuropathy research publishes systematic reviews whose abstracts mostly portray the findings of the review's full-text to reflect the results adequately.