The language restrictions and handsearching: a comparative analysis of meta analysis of interventions published in Chinese academic journals and Cochrane Library

Article type
Authors
Chen YL1, Li X1, Wei D1, Yang SF1, Yuan B1, Sun LN1, Yao L1, Wang Q1, Zhang PZ1, Yang KH1
1Evidence-Based Medicine Center of Lanzhou University, China
Abstract
Background: The validity of a systematic review is partially based on a comprehensive literature search. It is recommended that a search strategy include the databases, the hand-searching of the bibliographies of selected articles, conference proceedings, and abstracts; and personal communications with researchers. Selective search can affect the quality of the systematic review.

Objectives: To investigate the language restrictions and handsearching of systematic reviews between Chinese academic journals and Cochrane library.

Methods: We searched Chinese Biomedical Literature database (CBM) which is considered Chinese MEDLINE from January 2009 to December 2011 and all meta-analysis of interventions were included. We also searched Cochrane Database of Systematic Reviews (Issued 1, 2012) and 10% random samples of meta-analysis of interventions were included from 2009 to 2011. Each meta-analysis was independently identified and evaluated by 2 reviewers.

Results: Finally 967 Chinese meta-analysises were included. 674 (70%) didn’t report the language restrictions. 138 (14%) had no language restrictions; 155 (16%) had language restrictions, of which 31 (20%) were related to traditional Chinese medicine; 598 (62%) didn’t report the handsearching; 369 studies (38%) made the handsearching, of which 83 (23%) of the 269 studies were related to traditional Chinese medicine. One hundred and thirty-seven Cochrane meta-analysises were included. One (0.74%) had language restrictions which were published in 2010 and limited to English, French, Spanish, Italian and German; 86 (63%) had no language restrictions; 50 (36%) didn’t report the language restrictions; 79 (58%) of the 137 studies didn’t report the handsearching; 58 (42%) made the handsearching.

Conclusions: The systematic review shouldn’t exclude any language studies and the Cochrane Handbook recommends the handsearching. The rate of language restrictions of Chinese meta-analysises was 18 times than English, of which the studies related to the traditional Chinese medicine accounted for 1/5. The handsearching rate of Chinese systematic reviews was slightly low than English.