Altmetric and Bibliometric Analysis of Obstetrics and Gynecology Research: Influence of Public Engagement on Citation Potential

Abstract

Background

Whether research engagement with social media and other public platforms results in increased citations in obstetrics and gynecology (OB/GYN) remains uncertain. The Altmetric Attention Score (AAS) is a metric of research influence that is based on mentions in social media and public platforms, like newsfeeds and Wikipedia. The correlation between AAS, absolute citation rates and the Relative Citation Ratio (RCR), a novel metric of research engagement that is also based on citation rates, for research in OB/GYN is uncertain.

Objective

To evaluate the correlation between AAS, absolute citation rate, and RCR for articles published in OB/GYN journals from 2004-2019. Our second objective was to identify, characterize, and compare the 100 articles with highest AAS, the 100 top-cited articles, and the 100 articles with highest RCR.

Study Design

We performed a cross-sectional altmetric and bibliometric study of all OB/GYN articles indexed in the National Institutes of Health (NIH) Open Citation Collection (OCC) from 2004-2019. Articles were included if they were published in OB/GYN journals according to InCites Journal Citation Reports indexing. Citations data, including citation numbers and RCRs, were downloaded on 20 March 2021 and merged with altmetric data from the Altmetric Explorer based on each article's unique PubMed identification number. We assessed correlation between AASs and number of citations as well as AASs and RCRs by calculating Pearson correlation coefficient. The 100 articles with highest AAS, the 100 top-cited articles, and the 100 articles with highest RCRs were characterized and compared using means (standard deviations; SD) and mean differences (95% confidence intervals; CI).

Results

There were 156,592 articles published in 82 OB/GYN journals and indexed in the NIH OCC between 2004-2019. The correlation coefficient was 0.18 (95% CI 0.17−0.19) for AASs versus number of citations and 0.10 (95% CI 0.09-0.11) for AASs versus RCRs. There was no overlap among the 100 articles on the highest AAS list and the 100 top-cited list and minimal overlap among the 100 articles on the highest AAS list and the 100 highest RCR list (98 unique articles on each list). Articles with highest AASs generated substantially more engagement with social media and other public platforms compared to top-cited articles (mean [SD] AAS 763.1 (520.8) versus 49.9 [SD 81.6]; mean difference -713.2 [95% CI -819.9 − -606.6]) and highest-RCR articles (mean 116.2 [SD 415.9]; mean difference -661.5 [95% CI -746.2--576.9]). In contrast, the articles with highest AASs generated much fewer citations compared to top-cited articles (mean [SD] 39.7 [SD 47.6] versus 541.8 [SD 312.8]; mean difference 502.0 [95% CI 439.0-565.0]) and highest RCR articles (mean 458.9 [SD 363.5]; mean difference 427.7 [95% CI 353.8-501.6]). Nearly half of articles with highest AASs were basic/translational studies and prominently featured articles about menopause and environmental factors that impact fertility whereas top-cited articles and highest RCR were more likely to be reviews and consensus statements, respectively, and featured articles about the placentation and poly cystic ovarian disease, respectively. Articles with highest AASs were more likely to be published open access.

Conclusion

There appears to be weak short-term correlation between AASs and citation rates. Further study is warranted to ascertain whether there may be long-term correlation between alternative metrics and citation rates in OB/GYN.