Altmetrics Attention Scores for Randomized Controlled Trials in Total Joint Arthroplasty Are Reflective of High Scientific Quality: An Altmetrics-Based Methodological Quality and Bias Analysis. (December 2020)

• Record Type: Journal Article
• Authors: Kunze, Kyle N.; Richardson, Michelle; Bernstein, David N.; Premkumar, Ajay; Piuzzi, Nicolas S.; McLawhorn, Alexander S.
• Is Part Of: Journal of the AAOS. Volume 4:Number 12(2020)
• Publication Date: 2020-12
• Subjects: Orthopedic surgery -- Periodicals; 617.4705
• DOI: 10.5435/JAAOSGlobal-D-20-00187
• Languages: English
• ISSNs: 2474-7661

Abstract

Abstract : Introduction: The Altmetric Attention Score (AAS) has been associated with citation rates across medical and surgical disciplines. However, factors that drive high AAS remain poorly understood and there remains multiple pitfalls to correlating these metrics alone with the quality of a study. The purpose of the current study was to determine the relationship between methodologic and study biases and the AAS in randomized controlled trials (RCTs) published in total joint arthroplasty journals. Methods: All RCTs from 2016 published in The Journal of Arthroplasty, The Bone and Joint Journal, The Journal of Bone and Joint Surgery, Clinical Orthopedics and Related Research, The Journal of Knee Surgery, Hip International, and Acta Orthopaedica were extracted. Methodologic bias was graded with the JADAD scale, whereas study bias was graded with the Cochrane risk of bias tool for RCTs. Publication characteristics, social media attention (Facebook, Twitter, and Mendeley), AAS, citation rates, and bias were analyzed. Results: A total of 42 articles were identified. The mean (±SD) citations and AAS per RCT was 17.8 ± 16.5 (range, 0 to 78) and 8.0 ± 15.4 (range, 0 to 64), respectively. The mean JADAD score was 2.6 ± 0.94. No statistically significant differences were observed in the JADAD score or total number of study biases when compared across the seven journals ( P = 0.57 and P = 0.27). Higher JADAD scores were significantly associated with higher AAS scores (β = 6.7, P = 0.006) but not citation rate ( P = 0.16). The mean number of study biases was 2.0 ± 0.93 (range, 0 to 4). A greater total number of study biases was significantly with higher AAS scores (β = −8.0, P < 0.001) but not citation rate ( P = 0.10). The AAS was a significant and positive predictor of citation rate (β = 0.43, P = 0.019). Conclusion: High methodologic quality and limited study bias markedly contribute to the AAS of RCTs in the total joint arthroplasty literature. The AAS may be used as a proxy measure of scientific quality for RCTs, although readers should still critically appraise these articles before making changes to clinical practice.