Postoperative adjuvant chemotherapy of gastric cancer: Scrutiny into the clinical evidence based on quality assessment of medical literature of randomized controlled trials

The aim of this study was to scrutinize the evidence of adjuvant chemotherapy of gastric cancer by assessing the quality of the medical literature of randomized controlled trials (RCTs). A quality assessment (QA) scoring system was devised with the three parameters-control of bias, quality of report, and quality of design-which consisted 19 items. We searched for all the publications of the RCTs, from 1969 to 2007, with surgery-only arm, and their associated meta-analyses to score. Among the 26 RCTs, quality of three articles were graded as (2+), 10 articles as (1+), and 13 articles as (-). Recently published studies had overall better quality of report, but not necessarily better quality of design. Three studies demonstrating a positive survival benefit of adjuvant chemotherapy had a grade (1+). Hierarchical clustering revealed that the 26 articles were grouped into three major branches associated with study quality and a multi-institutional setting. We also obtained a statistically significant set of ten items (P < 0.001) that could differentiate articles of good (1-2+) and low quality (-) through supervised two-way hierarchical clustering. Finally, the level of recommendation for adjuvant chemotherapy in gastric cancer was to be a "B" according to the Scottish Intercollegiate Guidelines Network (SIGN) System. QA of medical literature should be an essential consideration for medical-related decision-making and the formation of evidence-based guidelines. Multidisciplinary discussion to develop and refine trial design is important for procuring better quality of RCTs of adjuvant chemotherapy of gastric cancer.