Predictors of citations in the urological literature

Abstract

To assess the factors associated with increased citation rates in the urological literature by reviewing articles published in the four major urological journals to help authors improve the impact of their work. A random sample of 200 original research articles published between January and June 2004 was analysed from The Journal of Urology, Urology, European Urology and BJU International. Study information was abstracted by two independent reviewers and citation counts within 4 years of publication were collected using Web of Science(TM) . Study characteristics and citation rates were analysed using median and interquartile ranges (IQRs), and logistic regression analysis was used to evaluate which factors predicted greater citation rates. The overall median number of citations per published article was 6.0 (IQR 3-12). After univariate analysis, we found that study design, study topic, continent of origin and sample size were associated with greater median citation rates. In a multivariate linear regression model, study design and study topic (oncology) predicted increased citation rates. Randomized controlled trials were cited a median of 13.5 times and were the strongest predictor of citation rates with an odds ratio of 115.5 (95% confidence interval 9.4-1419.6). Citation rates are associated with study design and study topic in the urological literature. Authors may improve the impact of their work by designing clinical studies with greater methodological safeguards against bias.