Threats to validity in randomized clinical trials

The purposes of this article are to present an overview of randomized clini cal trials (RCTs) and describe some of the methodological problems inherent in using RCTs in nursing research. Many nursing intervention studies are f raught with problems that defy the stringent control criteria required for RCTs, leading to biased estimates of intervention efficacy. Five threats to validity in RCTs are presented, including problems related to (a) differen tial dropout, (b) random assignment, (c) identifying and maintaining an ade quate control condition, (d) nonadherence to research protocols, and (e) as sessment of clinically meaningful change. Three strategies are recommended for addressing some of the problems posed by RCTs and improving inference, (C) 2000 John Wiley & Sons, Inc.