Randomized controlled trials (RCTs) are the gold standard test for efficacy and safety of clinical interventions and rightly influence guidelines for evidence-based clinical practice. In drawing inferences from trials, it is important that the trial population represent the clinical population to which such inferences are being applied. The extent to which this is true is known as external validity or generalizability. If a trial population differs from a clinical population, uncertainty exists as to whether trial findings will hold true in clinical practice. The lower the external validity, the greater this uncertainty. This problem is unfortunately very familiar to the nephrology community, since patients with impaired kidney function are routinely excluded from trials (1). There is an increasing recognition of the need to improve the external validity of trials to routine clinical practice (2).