Factors associated with online media attention to research: a cohort study of articles evaluating cancer treatments

Romana Haneef; Philippe Ravaud; Gabriel Baron; Lina Ghosn; Isabelle Boutron

doi:10.1186/s41073-017-0033-z

Factors associated with online media attention to research: a cohort study of articles evaluating cancer treatments

Res Integr Peer Rev. 2017 Jul 1:2:9. doi: 10.1186/s41073-017-0033-z. eCollection 2017.

Authors

Romana Haneef^{1

2

3}, Philippe Ravaud^{1

2

3

4

5}, Gabriel Baron³, Lina Ghosn², Isabelle Boutron^{1

2

3

4}

Affiliations

¹ INSERM, UMR 1153 Epidemiology and Biostatistics Sorbonne Paris Cité Center (CRESS), METHODS team, University of Paris Descartes, Centre d'Épidémiologie Clinique, AP-HP (Assistance Publique des Hôpitaux de Paris), Hôpital Hôtel Dieu, Paris, France.
² 2Paris Descartes University, Sorbonne Paris Cité, Faculté de Médecine, Paris, France.
³ Centre d'Épidémiologie Clinique, AP-HP (Assistance Publique des Hôpitaux de Paris), Hôpital Hôtel Dieu, Paris, France.
⁴ French Cochrane Center, Paris, France.
⁵ 5Department of Epidemiology, Columbia University Mailman School of Public Health, New York, NY USA.

Abstract

Background: New metrics have been developed to assess the impact of research and provide an indication of online media attention and data dissemination. We aimed to describe online media attention of articles evaluating cancer treatments and identify the factors associated with high online media attention.

Methods: We systematically searched MEDLINE via PubMed on March 1, 2015 for articles published during the first 6 months of 2014 in oncology and medical journals with a diverse range of impact factors, from 3.9 to 54.4, and selected a sample of articles evaluating a cancer treatment regardless of study design. Altmetric Explorer was used to identify online media attention of selected articles. The primary outcome was media attention an article received online as measured by Altmetric score (i.e., number of mentions in online news outlets, science blogs and social media). Regression analysis was performed to investigate the factors associated with high media attention, and regression coefficients represent the logarithm of ratio of mean (RoM) values of Altmetric score per unit change in the covariate.

Results: Among 792 articles, 218 (27.5%) received no online media attention (Altmetric score = 0). The median [Q1-Q3] Altmetric score was 2.0 [0.0-8.0], range 0.0-428.0. On multivariate analysis, factors associated with high Altmetric score were presence of a press release (RoM = 10.14, 95%CI [4.91-20.96]), open access to the article (RoM = 1.48, 95%CI [1.02-2.16]), and journal impact factor (RoM = 1.10, 95%CI [1.07-1.12]. As compared with observational studies, systematic reviews were not associated with high Altmetric score (RoM = 1.46, 95%CI [0.74-2.86]; P = 0.27), nor were RCTs (RoM = 0.65, 95%CI [0.41-1.02]; P = 0.059) and phase I/II non-RCTs (RoM = 0.58, 95%CI [0.33-1.05]; P = 0.07). The articles with abstract conclusions favouring study treatments were not associated with high Altmetric score (RoM = 0.97, 95%CI [0.60-1.58]; P = 0.91).

Conclusions: Most important factors associated with high online media attention were the presence of a press release and the journal impact factor. There was no evidence that study design with high level of evidence and type of abstract conclusion were associated with high online media attention.

Keywords: Altmetric score; Cancer treatment; Journal impact factor; Media attention; Open access; Press release.