Background: Well-designed and properly conducted RCT provides high
quality ‘raw materials’ for conducting systematic reviews, health technology
assessment and decision analysis reports. Poorly designed and
reported trials usually exaggerate the treatment effects which will mislead
clinical decision making. Objectives: Only a few Chinese medical journals
have recommended CONSORT in their ‘instructions for authors’ or ‘guide
for authors’. This study aims to evaluate the reporting quality of
randomized controlled trials (RCTs) published in the five leading Chinese
medical journals indexed by MEDLINE. Methods: We identified RCTs
published from 2004 to January 2007 in five leading Chinese medical
journals by searching three important Chinese databases systematically,
namely CNKI (China National Knowledge Infrastructure/Chinese Academic
Journals full text Database), VIP (a full text database of China) and CBM
disc (China Biomedicine Database Disc) and assessed the quality of each
RCT by using the Consolidated Standards for Reporting of Trials (CONSORT)
and the 5-point Jadad scale. Results: One-hundred and forty-two
RCTs were included. Based on the items in the revised CONSORT
statement, 130 (91.55%) of the 142 RCTs mentioned ‘‘randomization’’
in the title or abstract, but only 38 (26.76%) RCTs described the method to
generate the random sequence; only 6 RCTs had adequate allocation
concealment; 24 (17.61%) RCTs mentioned ‘‘masking’’, but only 7
described the process of masking. Three out of 40 items were reported
clearly in all included trials, while five items were not mentioned at all. The
quality of RCTs was also low as assessed by the Jadad scale with only 22
RCTs were high-quality research (>= 3 points). Conclusions: The reporting
quality of RCTs published in the five leading Chinese medical journals is
low. Chinese journals should adopt the CONSORT statement to improve
the reporting quality of Chinese randomized controlled trials.