Randomized Trials Analyzed as Observational Studies

Ann Intern Med.: October 15, 2013

Randomized trials are not always free of confounding and selection bias. This commentary posits that randomized trials with long follow-up are similar to observational studies and that “intention-to-treat” analysis of such studies may be inadequate. The authors propose use of g-methods to adjust for postrandomization confounding and selection bias. Read More