Difficulties inherent in the interpretation and performance of randomized controlled clinical trials

Abstract

Randomized controlled clinical trials have become the backbone of evidence-based clinical practice that all clinicians strive to achieve. This method of determining the efficacy of a treatment approach has become the basis of such processes as meta-analysis, best-evidence synthesis and the Cochrane Collaboration methods of determining the value of therapeutic approaches in all fields of health care. Initially, this research approach was developed to test the efficacy of medications, but increasingly there has been a demand that nonmedication treatment approaches, such as manipulation, exercise, injections and other invasive tests and surgery, be subjected to similar studies. This demand has required that considerable thought and innovation be brought to light in such areas as ethics, research methodology and design as well as statistical analysis. Issues such as referral bias and the ability to generalize results, as well as defining such issues as clinical relevance of an outcome, have become the subject of a great deal of discussion.