Pitfalls of multisite randomized clinical trials of efficacy and effectiveness.

This article discusses common pitfalls of multisite randomized clinical trials of efficacy and effectiveness. Issues considered include (1) premature initiation, (2) ineffective study structure, (3) too much or too little communication among researchers, (4) neglect of site differences, (5) use of "naked" p values, (6) premature closure, and (7) overreliance on the interpretation and memory of individual researchers. If future researchers are aware of these common pitfalls, they may be able to avoid them.