Background: since 2007, the Centre for Evidence-Based Chinese Medicine at Beijing University of Chinese Medicine has been working on identifying and translating randomized controlled trials (RCTs) of traditional Chinese medicine (TCM), published in Chinese, for the Cochrane Central Register of Controlled Trials (CENTRAL) via the Complementary and Alternative Medicine Field, using ProCite database software. With the development of a description of Chinese TCM literature, authors put more emphasis on key words for specific drugs and therapies.
Objectives: to search comprehensively for RCTs in TCM published in Chinese, and improve the search strategy to reduce the manpower needed to select eligible studies.
Methods: we compared search strategies that were based on subject words (Strategy 1), and those based on TCM journals (Strategy 2), as being of primary importance. We chose TCM journals from the catalogue of Chinese core journals so that the literature would be of higher quality. The search strategies are found in Table 1. We searched the CNKI database (the most commonly used Chinese literature database) by using the two strategies. We selected the top 500 records for manual screening. We compared the two search methods for a higher inclusion rate. After screening the articles manually, we used TCM words (#12-#46 in Strategy 1) to search on in articles retrieved from Strategy 2, to compare the differences between the articles in Strategy 1 and Strategy 2.
Results: we got 11,184 records using Strategy 1 and 2291 records using Strategy 2, and the top 500 records were selected into group 1 and group 2. A total of 426 records (85.2%) were included in group 1 and 437 records (87.4%) were included in group 2. We saw that 276 records (63.16%) in group 2 used words about TCM (#12-#46 in Strategy 1) as subject words. We found manual screening easier when using strategy 2. We also searched some articles that had exclusive words (#4-#10 in Strategy).
Conclusions: many RCTs in TCM from TCM journals did not use TCM words in the Abstract, and Strategy 2 shows a little higher inclusion rate, so we can use Strategy 2 when submitting Chinese RCTs of TCM to CENTRAL. More high-quality journals should be found if initial retrieval records are not enough. Exclusive words should be considered seriously when searching the CNKI database.
Patient or healthcare consumer involvement: we are working to disseminate good evidence about TCM for Cochrane systematic reviews, as this will help patients or healthcare consumers to obtain the latest medical evidence about TCM.