Open Access Articles Garner Increased Social Media Attention and Citation Rates Compared With Subscription Access Research Articles: An Altmetrics-Based Analysis

Abstract

Background: To better understand the research impact on social media, alternative web-based metrics (Altmetrics) were developed. Open access (OA) publishing, which allows for widespread distribution of scientific content, has become increasingly common in the medical literature. However, the relationship between OA publishing and social media impact remains unclear. Purpose: To compare social media attention and citation rates between OA and subscription access (SA) research articles within the orthopaedic and sports medicine literature. Study Design: Cross-sectional study. Methods: Articles published as either OA or SA in 5 high-impact hybrid orthopaedic journals between January 2019 and December 2019 were analyzed. The primary outcome was the Altmetric Attention Score (AAS), a validated measure of social media attention. Secondary outcomes included citation rates, article characteristics, and the number of shares on social media. Independent t tests and chi-square analyses were used to compare outcomes between OA and SA articles. A multivariable linear regression analysis was performed to determine the association between article type and AAS while controlling for bibliometric characteristics. Results: A total of 2143 articles (246 OA articles, 11.5%; 1897 SA articles, 88.5%) were included. The mean AAS among all OA articles was 62.4 ± 184.6 (range, 0-2032), whereas the mean AAS among all SA articles was 18.4 ± 109.8 (range, 0-3425), representing a statistically significant difference ( P < .001). The mean citation rate among OA articles was significantly higher (17.0 ± 22.5; range, 0-139) than that of SA articles (8.6 ± 13.4; range, 0-169) ( P < .001). Multivariable linear regression analysis demonstrated that OA status (β = 15.15; P = .044), number of institutions (β = 2.13; P = .023), studies classified as epidemiological investigations (β = 107.40; P < .001), and disclosure of a conflict of interest (β = −11.18; P = .032) were significantly associated with a higher AAS. Conclusion: OA articles resulted in significantly greater AAS and citations in comparison with SA articles. Articles published through the OA option in hybrid journals as well as those with a higher number of institutions, those that disclosed a conflict of interest, and those classified as epidemiological investigations were positively associated with greater AAS in addition to a greater number of citations. The potential for more extensive research dissemination inherent in the OA option may therefore translate into greater reach and social media attention.