A survey on literature searching of systematic review in traditional Chinese medicine filed

Article type
Authors
Liang D1, Jian H1, Min C2, Guan-Jian L1
1Chinese Cochrane Centre, West China Hospital, Sichuan University, Chengdu, China
2Department of Traditional Chinese Medicine, No. 3 People s Hospital of Chengdu, Chengdu, Sichuan, China
Abstract
Background: The application of the method of systematic review (SR) in traditional Chinese medicine (TCM) filed provide high quality evidence resources for TCM clinical practice, the problems of TCM trials identified from the production of SRs provide valuable information for improving their quality. However, the quality of SRs have high effect on their conclusions, the literature searching is one of the most important links. Objective: To analysis and compare the literature searching status of TCM SRs published in the Cochrane Library (CL) and Chinese Journals, to provide reference for improving literature searching of TCM SRs. Methods: We searched the CL (Issue 4, 2009) and the Chinese Biomedical Database (CBM, from 1978 to July 31, 2009) to identify SRs in TCM field. A self-designed data extraction table was used for data extraction. The SPSS 15.0 software was used for data description and analysis. Results: 341 (CBM: 245; CL: 96) SRs were included. Compared with the CBM reviews, Cochrane reviews searched more databases (media: 4 vs. 6; Z = -7.862, P = 0.000). The top three databases searched by the authors of Cochrane and CBM reviews were MEDLINE (100%), EMBASE (88.4%), CENTRAL (74.7%), and CBM (75.9%), CNKI (70.2%), CENTRAL (60.4%), respectively. 42 CBM reviews didn t report what databases were searched. Of the CBM reviews, 47.7% reported the exact time range covered by the search, 14.4% reported either the beginning or the end date of the time range, and the remaining 37.8% did not report the time range at all. Only 20 CBM reviews reported at least one detailed searching strategy. Fifty CBM reviews reported the gray literature was searched, 134 reported the hand-searching was conducted. Conclusions: There are big problems in literature searching of TCM SRs. These problems include inadequate selection of object databases, poor repeatability of searching process, inadequate searching for gray literature, uncertainty function of hand-searching, etc. Practice guideline need to be developed to standardize the literature searching of TCM SRs.