Objective Scientific journals and authors are frequently judged on ‘impact’. Commonly used traditional metrics are the Impact Factor and H-index. However, both take several years to formulate and have many limitations. Recently, Altmetric—a metric that measures impact in a non-traditional way—has gained popularity. This project aims to describe the relationships between subject matter, citations, downloads and Altmetric within rheumatology. Methods Data from publications in Rheumatology were used. Articles published from 2010 to 2015 were reviewed. Data were analysed using Stata 14.2 (StataCorp, College Station, TX, USA). Correlation between citations, downloads and Altmetric were quantified using linear regression, comparing across disease topics. Relationship between downloads and months since publications were described using negative binomial regression, clustering on individual articles. Results A total of 1460 Basic Science and Clinical Science articles were identified, with the number of citations, downloads and Altmetric scores. There were no correlations between disease topic and downloads (R2 = 0.016, P = 0.03), citations (R2 = 0.011, P = 0.29) or Altmetric (R2 = 0.025, P = 0.02). A statistically significant positive association was seen between the number of citations and downloads (R2 = 0.29, P < 0.001). No correlations were seen between Altmetric and downloads (R2 = 0.028, P < 0.001) or citations (R2 = 0.004, P = 0.445). Conclusion Disease area did not correlate with any of the metrics compared. Correlations were apparent with clear links between downloads and citations. Altmetric identified different articles as high impact compared with citation or download metrics. In conclusion: tweeting about your research does not appear to influence citations.

Objective: to establish if gravidity and parity associate with the development of rheumatoid arthritis (RA), and to establish if this effect is influenced by the time elapsed since pregnancy/childbirth, the number of pregnancies/childbirths, and serological status, through systematically reviewing the literature and undertaking a meta-analysis. Methods: we searched Medline/EMBASE (from 1946-2018) using the terms "rheumatoid arthritis.mp" or "arthritis, rheumatoid/" and "pregnancy.mp" or "pregnancy/" or "parity.mp" or "parity/" or "gravidity.mp" or "gravidity/" (observational study filter applied). Casecontrol/cohort studies that examined the relationship between parity/gravidity and the risk of RA in women were included. Studies reporting effect size data for RA in ever vs. never parous/gravid women as ORs/RRs with 95% confidence intervals were included in a metaanalysis. Other relationships (i.e. risk by pregnancy/childbirth numbers) were analysed descriptively. Results: twenty studies (from 626 articles) met our inclusion criteria, comprising 14 casecontrol (4,799 cases; 11,941 controls) and 6 cohort studies (8,575 cases; 2,368,439 individuals). No significant association was observed in the meta-analysis of studies reporting the risk of RA in ever vs. never parous women (OR 0.91; 95% CI 0.80-1.04) and ever vs. never gravid women (OR 0.86; 95% CI 0.46-1.62). No consistent evidence of a relationship between the number of pregnancies/childbirths and RA risk was seen. No significant association was observed between being pregnant, or in the immediate post-partum period, and the risk of developing RA. Conclusion: our systematic review does not support the concept that gravidity and parity are associated with the risk of RA development.

No abstract

No abstract