Article Text

Download PDFPDF

Social media and predictors of traditional citations: insights from the Journal of Neurointerventional Surgery
  1. Ashutosh P Jadhav1,
  2. Andrew F Ducruet2,
  3. Reade de Leacy3,
  4. Kyle M Fargen4
  1. 1 Department of Neurology, University of Pittsburgh, Pittsburgh, Pennsylvania, USA
  2. 2 Department of Neurosurgery, Barrow Neurological Institute, Phoenix, Arizona, USA
  3. 3 Department of Neurosurgery, Mount Sinai Hospital, New York City, New York, USA
  4. 4 Department of Neurological Surgery, Wake Forest University, Winston-Salem, North Carolina, USA
  1. Correspondence to Dr Ashutosh P Jadhav, Department of Neurology, University of Pittsburgh, Pittsburgh, PA 15218, USA; jadhavap{at}

Statistics from

In 1971, a patient with suspected left frontal lobe tumor underwent the first clinical CT scan. The prototype scanner was developed by Godfrey Hounsfield.1 Almost 10 years earlier, Allan Cormack had independently demonstrated that multiple measurements of radiographic attenuation could be reconstructed to form an image of the target tissue. While both Hounsfield and Cormack shared the Nobel Prize Physics and Medicine in 1979 for the discovery of the CT scan, Hounsfield and Cormack never met and furthermore Hounsfield was apparently not familiar with Cormack’s work.1 It is tempting to speculate whether earlier knowledge sharing and collaboration between the scientists would have accelerated the development of this revolutionary technology.

Timely diffusion of medical knowledge has dramatically improved over recent decades and likely itself contributes to the pace at which advances are being made. Estimates suggest that the doubling time of medical knowledge in 1950 was 50 years, decreased to 7 years in 1980 and further decreased to 3.5 years in 2010. In 2020, it is anticipated that the double time will be only 73 days. With the widespread availability of computers and the internet, dissemination of medical knowledge has become faster with fewer barriers, even across sub-specialties.2

Social media engagement is the latest tool used to generate nearly real-time awareness of both peer-reviewed publications as well as freshly-presented conference abstracts and presentations. In addition to sharing scientific developments, social media allows for anecdotal sharing of clinical experience, although such reports are prone to bias. Other applications of social media include the formation of virtual communities using aggregating tools such as ‘hashtags’ on Twitter. A recent analysis of the hashtag #Stroke has demonstrated a significant increase in use over the last 6 years with the highest activity among non-physician groups.3 The long-term impact of social media activity on traditional citations, however, remains unclear.4–7

Beginning in 2015, the Journal of Neurointerventional Surgery (JNIS) acquired social media editors and has been actively promoting published content through social media.8–10 In order to better understand the predictors of future conventional citations based on Web of Science, we analyzed 451 articles published online first in  JNIS  between February 2015 to October 2016. We recorded level of evidence based on prior published methodology.11 We defined ‘clicks’ as the number of times followers clicked the link on JNIS social media posts to directly access the article. Citations were collected on Web of Science 2 years after the original date of online publication of each article.

The strongest predictors of citations at 2  years were the number of clicks, (with a 1% increase in citations for every additional click) and the level of evidence presented by the article, with a 20% increase in citations for every one unit increase in higher level of evidence (table 1). Analysis by subject matter revealed that articles focusing on socioeconomics are associated with the highest number of citations followed by stroke and aneurysm. Twitter impressions and engagement did not predict conventional citations; however, even after controlling for the increasing number of social media followers over time, the number of clicks per Twitter user has increased over time (figure 1). This suggests that both the number of followers as well as the engagement among followers is increasing over time.12 13

Figure 1

Number of clicks per Twitter follower.

Table 1

Significant predictors of standard citations in regression analysis

The interaction between medical advancement and knowledge propagation is complex and continues to evolve. The two components are closely inter-related as faster awareness guides further research. Our analysis of articles published in JNIS suggests that the level of evidence of the publication and the topic of research strongly predicts future citations. The number of clicks also appears to be a strong predictor of future citations and the number of clicks increases as the number of Twitter users also grows. While the direct relationship between social media and traditional citations remains elusive, this emerging platform plays a growing role in rapidly amplifying peer-reviewed scientific literature. This counter-measure is essential in an era where misinformation can be propogated and subsequently threaten scientific integrity.14 15


Carol Aschenbrenner for assistance with statistical analysis.


View Abstract


  • Patient consent for publication Not required.

  • Contributors all authors contributed.

  • Funding The authors have not declared a specific grant for this research from any funding agency in the public, commercial or not-for-profit sectors.

  • Competing interests None declared.

  • Provenance and peer review Commissioned; internally peer reviewed.

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.