Background:Systematic reviews provide the most reliable evidence for certifying some treatment’s effect and safety, and enable decision making for clinical practice.Therefore, it is vital to retrieve the relevant literature comprehensively and objectively as far as possible.

Objectives:To help systematic reviewers in China select the best biomedical databases from which to retrieve relevant trials according to comparison of coverage, search feature, and so on.

Methods:Three foreign biomedical databases (MEDLINE, EMBASE and the Cochrane Central Register of Controlled Trials (CENTRAL)) and four Chinese biomedical databases (China Biology Medicine (CBM), CNKI, VIP and WANFANG) were selected. Data about coverage, search features, records downloading, and presence were found through their websites. The deadline for data collection was the end of April 2014. Based on the data collected, the number of journals from the inaugural issue and indexed cover-to-cover were calculated.

Results: In MEDLINE there were about 5695 indexed journals, and about 5728 in EMBASE; about 2000 are nonredundant. The kinds of indexed journals in CENTRAL are relatively comprehensive, but it updates four times a year. For the four Chinese databases, the way CBM retrieves is similar to the three foreign databases. CBM indexed fewer articles than CNKI, but indexed the most journals (1784, 91%). It provides MeSH searching and more terms than other databases, and has an output of the tagged texts up to 500 records per file. CNKI and WANFANG provide English interfaces, which is convenient for English searching, and CNKI has a function of 'Cross-Language Search', which automatically translates English into Chinese.

Conclusions:At the very least Chinese systematic reviewers should search MEDLINE, EMBASE, CENTRAL and CBM. CBM is the preferred database for systematic reviewers to retrieve relevant studies in Chinese, while CNKI is recommended for non-Chinese-speaking researchers due to its English interface and 'Cross-Language Search' function.