The cross-section study of the publication bias of published Meta-analysis on tumour in science citation index expanded database in 2014

Ya-ping DENG,Hui-qiang YU,Yong LIU,Yue LI,Shi-qi CHENG,Hui-lie ZHENG
2017-01-01
Abstract:Objective To investigate the diagnosis of publication bias (PB) of the Meta-analysis which was about the tumor in Science Citation Index-expanded (SCIE) in order to promote the standardization of diagnosis of publication bias in Meta-analysis,and to improve the quality of these articles.Methods Explored the published Meta-analysis literature in SCIE,and then used R 3.3.1 to calculate the results of diagnosis of the publication bias when different diagnostic methods were applied,and used SAS 9.3 to analyze the data.Results About 89.06% published Meta-analysis conducted the work to diagnose the publication bias,and 11.64% of the above Meta-analysis showed there was publication bias,which was lower than 26.42%,the proportion of Meta-analysis which had publication bias detected by the third-party method for diagnosing the publication bias in Meta-analysis.Conclusion The proportion of Meta-analysis which showed there was publication bias is lower,so we suggest that reviewers or editor of the journal can resort to the third-party method to reevaluate whether there is publication bias in Meta-analyses or not.
What problem does this paper attempt to address?