Several recent major randomized clinical trials (RCTs) using renal outcomes resulted in conflicting results. We searched MEDLINE via PubMed with the search request '(dialysis OR end-stage renal disease) and creatinine' in six major general journals and two leading journals of nephrology; 123 articles were found; 17/123 were relevant RCTs. Some disagreement among surrogate endpoints in 11/15 articles (missing data in two RCTs) and between surrogate and hard renal endpoints in 10/13; the intervention effects were in the opposite direction in 4/15, mostly in patients with cardiovascular disease, but discrepancies and conflicting results were also found among renal trials. Among our selected RCTs, 14/17 used composite endpoints: vital and renal endpoints were mixed in 11/14 trials, the components of the composite endpoints were of similar importance in 0/14 trials and of similar frequency in 1/11 trials, the intervention was likely to have a similar effect on the components in 4/14 trials, and the relative risk reduction was similar for the different components in 2/10 trials. None of the trials fulfiled all conditions of validation. Based on this analysis, we believe that the current use of renal endpoints in RCTs must be reviewed and their conditions of validation defined.
© 2011 The Authors Fundamental and Clinical Pharmacology © 2011 Société Française de Pharmacologie et de Thérapeutique.