[Development and analyses of evaluation criteria for meta-analysis-relationship between renin-angiotensin system inhibitor and diabetes mellitus].

In recent years, evidence-based medicine is practiced in medical care, and meta-analysis is positioned as the highest quality research paper. However, there are no unified criteria to evaluate meta-analysis. Thus, we developed and evaluated objective criteria of meta-analysis paper. Meta-analysis was evaluated from the standpoint of format and quality. Format score was developed based on QUOROM statement. Quality scores were evaluated in each step of meta-analysis. We evaluated meta-analysis for advantageous or adverse effect in the relationship between Renin-angiotensin system (RAS) inhibitor and diabetes mellitus. Research articles either in English or Japanese were searched in PubMed and the Cochrane Library in July 2009. Fifty-five studies were evaluated for format and quality score. Format score was 64.5+/-25.3% (Mean+/-S.D.), and quality score was 57.1+/-21.6%. After plotting each score along the X-Y coordinate, format score was correlated with quality score (R2=0.702, p<0.001). In studies with quality score of 70% or higher, RAS inhibitor prevented new-onset diabetes mellitus. Angiotensin converting enzyme (ACE) inhibitor reduced risk of myocardial infarction, however, angiotensin receptor blockers (ARB) increased the incidence in odds ratio 1.21 (95%CI: 1.04-1.40). We evaluated meta-analysis by format and quality score, and these factors showed correlation. Therefore, assessment of quality score leads to prompt and reliable evaluation of meta-analysis. According to articles evaluated by quality score, RAS inhibitor showed effectiveness on diabetes mellitus. However, careful attention must be given in ARB therapy on patients with myocardial infarction risk.