Proposed new score to rate the strength of evidence and its application to large-scale clinical trials of angiotensin-receptor blockers.

BACKGROUND Many large-scale clinical studies appear to be interpreted with bias and with hindsight. To select the best treatment, accurate evaluation of randomized controlled trials (RCTs) and fair comparison of the results using the concept of evidence-based medicine are critical. A scoring system has been developed to rate the scientific strength of evidence and thereby provide guidance for best clinical practice and for large-scale RCTs of angiotensin-receptor blockers (ARBs). METHOD AND RESULTS Positive evidence scores (ESs) were given based on whether specified study endpoints were proven and whether treatments tested made significant progress over current usual therapy. Retrospective ad hoc analyses data were not counted, in order to simplify the system and to avoid hindsight interpretations, even in cases that were medically significant. In fact, in more than half of the large-scale trials with ARBs examined, ad-hoc analyses had been retroactively performed. When such post-hoc analyses were not used, the ESs were positive with candesartan and valsartan for treatment of heart failure, irbesartan and losartan for nephropathy and losartan for hypertension. Applying the ES system, losartan was judged to be an ARB with the strongest evidence covering a wide range of clinical relevance. CONCLUSION ES is useful for quantifying and comparing the strength of evidence obtained in large-scale RCTs of ARBs. Several problems related to rating the evidence obtained from clinical trials were recognized in this study.