1418. Predictors of Citations in Neurosurgical Research

Authors: Chesney Oravec, MD; Casey Frey, BS; Benjamin Berwick, MS; Lukas Vilella, BS; Carol Aschenbrenner, MA; Stacey Wolfe, MD; Kyle Fargen, MD, MPH (Winston Salem, NC)

Introduction: The number of citations an article receives is an important measure of impact for published research. There is limited published data on predictors of citations in neurosurgery research. We sought to analyze predictors of citations for neurosurgical articles. Methods: All articles published in 14 neurosurgical journals in the year 2015 were examined and data collected about their features. The number of citations for each article was tallied using both Web of Science (WoS) and Google Scholar (GS) 2.5 years after their publication in print. Negative binomial regression was then performed to determine the relationship between article features and citation counts for scientific articles. Results: A total of 3,923 articles were analyzed comprising 2,867 scientific articles (72.6%) and 1,056 non-scientific (editorial, commentary, etc) articles (27.4%). At 2.5 years, scientific articles had a median[IQR] number of citations per article of 3.0[6.0] and 7.0[9.0] found in WoS and GS, respectively; non-scientific articles had accumulated median 0.0[2.0] in both WOS and GS. Articles with the study topic spine had the highest citation count at 4.0[5.0] and 8.0[10.0] in WoS and GS, respectively. Significant predictors of citation count in scientific articles were level of evidence, number of centers, number of authors and impact factor. Conclusion: This is the largest investigation analyzing predictors of citations in the neurosurgical literature. Factors found to be most influential on citation rates in scientific articles included the study’s level of evidence, number of participating centers, number of authors, and the publishing journal’s impact factor.