Variation in results from randomized, controlled trials: stochastic or systematic?

Conflicting results from RCTs can represent a spectrum of “real” outcomes for specific treatments. Such trials are best evaluated by considering concurrently both the validity of study design as well as the generalizability of patients and interventions involved.