[Evidence-based medicine: can we trust the results of well-designed randomized trials?]

Evidence based medicine assists in clinical decision-making by integrating critically appraised information with patient's values and preferences within an existing clinical context. A fundamental concept in this paradigm is the hierarchy of information. The randomized clinical trial is recognized as one of the designs that is less prone to bias and therefore of higher methodological quality. Clinical guidelines are one of the principal tools that evidence based medicine uses to transfer scientific information to clinical practice and many of their recommendations are based on these type of studies. In this review we present some of the limitations that the results can have, in even well designed and executed randomized clinical trials. We also discuss why valid results in these types of studies could not necessarily be extrapolated to the general population. Although the randomized clinical trial continues to be one of the best methodological designs, we suggest that the reader be careful when interpreting its results.