The Worldwide Influence of Social Media on Cervical Spine Literature

DISCUSSION

Although citation rates and impact factors have traditionally been used to assess the influence of academic publications, social media outlets, such as Twitter and Facebook, are playing an increasing role in the dissemination of medical information. In order to take the influence of these “less formal” sources into account, innovative measures such as the Altmetric score have been developed. Altmetric rankings for general spine-related literature have previously been assessed by Parrish et al.12 However, while lower back pain and problems related to the lumbar spine may be considered a relatively “mainstream” topic in modern medicine, the cervical spine may represent more of a specialized or “niche” subset of literature with a potentially narrower scope of both contributors and readers. Our analysis of literature related to the cervical spine revealed a number of differences based on which metric was used to identify articles.

In terms of the articles returned by our search of the Altmetric database, Spine, European Spine Journal, and The Spine Journal published the most articles related to the cervical spine. In their broader review of the Altmetric database for articles related to the spine in general, Parrish et al identified the same 3 journals as the most prolific contributors of spine-related literature.12 The tendency of relevant articles to be published in spine-specific journals, rather than general medical or orthopedic publications, highlights the specificity of cervical spine surgery. It also further emphasizes the importance of understanding how spine-related literature is disseminated to the general public, especially given the relatively narrow readership of many such journals. Interestingly, only one of the top cervical spine research publishers score among the top AAS scores, with top scores belonging to Journal of Neurosurgery: Spine, Spine, and the Journal of Manipulative & Physiological Therapeutics. Wang et al performed a similar analysis of articles published in neurosurgical journals and determined that the Journal of Neurosurgery had the most publications among their list of the top 100 articles ranked by AAS score.

Twitter, Facebook, and web-based news sources contributed the most social media mentions per article, with Twitter in particular demonstrating by far the highest rate of mentions at 4.4, compared to 0.5 and 0.3 for Facebook and news sources, respectively. Parrish et al similarly found that Twitter was especially influential in contributing to the Altmetric score of spine-related articles.12 Studies have demonstrated that a number of social media campaigns by academic journals and groups have been successful in increasing the readership of their articles.16,17 Furthermore, an observational study of the top 50 trauma and orthopedic surgery journals, ranked by 2016 impact factor, demonstrated significant correlations between traditional measures of academic influence (ie, impact factor) and social media activity, particularly in terms of Twitter engagement by journals, Twitter followers and mentions on Twitter.18

A number of previous studies have demonstrated relationships between Altmetric scores and more traditional metrics such as citation count.18–20 Twitter mentions, in particular, are directly correlated with citation count, though they accumulate at a significantly higher rate than citations.18,19 In a study of radiation oncology literature, Paradis et al found that not only was a Twitter activity associated with the number of times an article was cited but also that articles with a Twitter “buzz” of 10 or more tweets prior to publication had 3.6 more citations than those that did not.21

However, only 8 of the articles identified by our analysis overlapped between the top 100 identified by AAS and the top 100 most cited articles. This discrepancy underlines the difference in the articles that are highlighted and elevated by “alternative” sources, such as social media, compared to those that are frequently cited by other peer-reviewed, academic publications. Interestingly, a systematic review of orthopedic randomized clinical trials demonstrated that higher numbers of social media mentions (mostly Twitter) were associated with less risk of bias, in addition to higher journal impact factor and higher author h-index values.19

Social media can significantly impact the type and content of medical information that is readily available to the public. An analysis of colorectal cancer information available on Twitter by Park et al determined that although medical professionals and institutions produced a relatively low proportion of total tweets on the topic (2.0% and 0.6%), 84.5 % of the information posted was medically accurate.22 Providing further hope for the information on social media, Park et al found that the most frequently retweeted tweets were significantly more accurate than randomly selected tweets. Conversely, a study of the reliability of health information disseminated on Twitter found that over half of the Arabic langauge, health-related tweets posted by professionals over a 5-day span provided information that was not true.23

In addition to differences in the content of the articles prioritized by social media and traditional citation rates, variation also exists with regard to the type of publications that are prioritized. Our analysis of cervical spine literature revealed that the top AAS articles included more prospective studies and qualitative review articles, while the most cited articles contained more retrospective studies and systematic reviews. This highlights an interesting association given that prospective studies or randomized controlled trials are perceived as a higher level and preferred methodology of research. Similar to our findings, studies involving journals from other medical fields have reported that over 25% of the research conducted utilized the retrospective chart review research design.24 It must be noted that with the appropriate design, retrospective cohort studies may offer substantial validity and reliability as well.25 With AAS promoting a higher level of research methodology as compared to citation frequency, this suggests that Altmetric scores may facilitate the promotion of research that is more generalizable and reliable. As medicine continues to become globalized and increasingly connected, understanding regional trends in literature contribution and readership is more important than ever. In their 2016 study of “The 100 most Influential Articles in Cervical Spine Surgery”, Skovrlj et al list the United States, Japan, and Canada as the origin of the most articles in their top 100 most-cited list, with 65, 16, and 6 top 100 articles, respectively.6 The United States has consistently dominated in terms of publication volume and citation rate, both for the cervical spine and spine literature in general, with Japan, China, Germany, and Canada frequently cited as other major contributors.14,26–28 Additionally, research productivity has been tied to economic status, with high-income countries producing the vast majority of spine-related publications, followed by middle-income, and distantly by low-income countries.14 While this economic inequality in research productivity may not be surprising, it provides further reason to understand less traditional channels for the dissemination of spine-related literature at the global level, including social media outlets such as Twitter.

Our analysis of the Altmetric database revealed that the United States, the United Kingdom, and Spain contributed the most Twitter mentions of literature related to the cervical spine. Given its well-documented role in producing spine-related literature, it is unsurprising that the United States would also be a leader in terms of social media activity related to the cervical spine. Interestingly, however the trends we observed in other countries tweeting about cervical spine literature are consistent neither with previous data regarding regional contributions to spine literature nor with trends in Twitter usage by country. As of July 2020, the United States had by far the largest Twitter audience at 62.6 million users and Japan the second most at 49.1 million.29 However, the United Kingdom was fifth in Twitter participation at 15.3 million, and Spain was 13th with only 7.1 million users. Furthermore, while the United States and the United Kingdom are both predominantly English-speaking countries, Spain is not. A recent review of health-related posts on Twitter indicated that the average writing level of these tweets was equivalent to approximately a 6- to 9-year-old level, which is substantially lower than the reading level of the average user.30 This relatively simple level of writing may explain the accessibility of Twitter posts even if they are not written in the user’s first language. Our results regarding global Twitter mentions of cervical spine literature are especially interesting to consider in this context, since they highlight a unique proclivity for social media–based dissemination of cervical spine literature in the United Kingdom and Spain that do not appear to be based solely on either the research productivity or the Twitter engagement of these countries.

Social media facilitates the rapid dissemination of research articles at a large scale.23 Possible pitfalls associated with lower barriers for dissemination include misinterpretation of research findings and increased circulation of pseudoscience.23 Within the context of cervical spine literature, our analysis noted that top AAS articles included more prospective studies and qualitative reviews vs retrospective studies, with AAS scores correlating with research that is generalizable, valid, and reliable. However, there was a noted disconnect between the top 100 most-cited articles and the top 100 articles by AAS score. The findings of our study would suggest social media can be a tool that can promote increased awareness of quality cervical spine literature but would caution readers to be aware that articles with higher AAS scores may tend to overrepresent studies of specific content type and methodological method when compared with peer-reviewed journals. Another pitfall for consumers of research spread via social media to navigate is to ensure that study findings commented on by social media participants correctly represent the article shared. Without peer review, self-policing is critical to combat the spread of pseudoscience.