Randomized Controlled Trials in Nephrology: State of the Evidence and Critiquing the Evidence Joshua A. Samuels and Donald A. Molony The randomized controlled trial (RCT) remains the ‘‘gold standard’’ for the evaluation of therapies. Despite some progress dur- ing the past decade, the number and quality of published RCTs addressing the core issues for patients with CKD and with renal diseases, in general, lag behind other areas in internal medicine. The paucity of robust evidence results in fewer patients receiv- ing evidence-based therapies in nephrology and fewer rigorous systematic reviews to inform nephrology practice and health care policy. Because trials of lower methodologic rigor continue to be published, the evidence-based practitioner must evaluate new evidence from the medical literature carefully before incorporating that evidence into their clinical practice. The types of errors that may limit the validity or applicability of evidence from RCTs is outlined. A detailed discussion of the most important design elements for the conduct of a high-quality RCT is described in the text. These considerations are placed into the context of critical appraisal tools. These tools allow the clinician to efficiently assess the quality of published RCTs and to determine how the new RCT evidence should change current best practice. Q 2012 by the National Kidney Foundation, Inc. All rights reserved. Key Words: Randomized trial, Clinical trial, Evidence-based medicine T here are 2 main types of clinical investigation: obser- vational studies and experimental (clinical) trials. Be- cause exposure is controlled in an experimental clinical trial, many of the biases that can potentially distort the truth emerging from observational studies are either ab- sent from or greatly diminished in a randomized con- trolled trial. Thus, randomized controlled trials (RCT) are generally considered the gold standard when interro- gating a question dealing with a therapeutic intervention. For an RCT to best inform therapy choices, it must be con- ducted in a manner that is methodologically rigorous and must address the clinical question in a fashion that is most likely to yield clinically meaningful and applicable re- sults. We shall turn first to a general discussion of the state of RCT evidence within nephrology. As a precautionary measure, we shall consider the important gaps in knowl- edge that may evolve out of the current practices of sup- port and conduct of many of the published RCTs in nephrology. We will then illustrate how even the most rig- orous critical appraisal might not expose the potential for these distortions or gaps in knowledge but that any inter- rogation of the literature must begin with critical ap- praisal of the individual studies, and only those studies of high methodological rigor should be considered in the vetting of totality of the evidence. We will discuss in detail the rationale for and the steps involved in critical appraisal of RCTs to provide the practitioner with the core evidence-based medicine (EBM) tools for evaluating evidence on best therapies. State of Evidence in Nephrology Although it has been widely recognized for more than 2 decades that the best way to determine the most effec- tive and safest treatment of any particular disease is to conduct randomized controlled trials, 1 far fewer robust RCTs have been performed to address the core questions in nephrology than in other fields of medicine. 2-4 In 2000, Campbell and colleagues performed an extensive search and evaluation of the medical literature for 6 core topics in ESRD and identified only 39 clinical trials that were of sufficient rigor to meet the Cochrane stan- dards for inclusion in a systematic review. More re- cently, a search of the Cochrane register of clinical trials using keywords or titles for a number of common diseases illustrates the persistence of this state of affairs. The results of this search performed on the Cochrane 1- 2012 data set are shown in Table 1. Strippoli and col- leagues more formally analyzed the comparative state of RCT evidence in nephrology, first reporting their find- ings in 2004 and then recently updating these. 3,4 In 2004, they observed that the absolute number of RCTs was the lowest for any specialty of internal medicine; the per- centage of citations of the nephrology literature that rep- resented RCT versus observational studies was only 1.15%. 3 They reviewed a sampling of the RCTs for qual- ity and found that the overall quality of RCTs in ne- phrology was poor when using major criteria as quality filters. Eight-nine percent lacked clear allocation concealment, 92% of RCTs failed to blind assessors of the outcomes, and more than 50% failed to perform an intention-to-treat analysis. The importance of each of these elements to the integrity of the data will be dis- cussed later in the text. During the past decade, there has been some improvement in the number of RCTs re- ported in nephrology; however, the increase has been at a slower rate than in all other specialties in medicine, and the differential has actually widened. 4 These trends are illustrated in Figure 1. From Division of Pediatric Nephrology and Hypertension, University of Texas Medical School at Houston, Houston, TX; and Division of Renal Diseases and Hypertension, University of Texas Medical School at Houston, Houston, TX. Address correspondence to Joshua Samuels, MD, MPH, Division of Pediat- ric Nephrology and Hypertension, University of Texas Medical School at Houston, 6431 Fannin Street, MSB 3-121, Houston, TX 77030. E-mail: Joshua.A.Samuels@uth.tmc.edu Ó 2012 by the National Kidney Foundation, Inc. All rights reserved. 1548-5595/$36.00 doi:10.1053/j.ackd.2012.01.009 Advances in Chronic Kidney Disease, Vol 19, No 1 (January), 2012: pp 40-46 40