[Statistics as support tool and not as a decision tool].

The use of statistics in medical articles has risen a lot during the last decades, however it is used in a thoughtless manner in many instances. Today, Statistics is the only tool that allows the medical researcher to obtain results and benefits from those studies the relationships of which can not be interpreted from a determinist perspective, because it is a branch of applied mathematics objective of which is to manage and quantify the uncertainty of the available information, to support decision taking. The objective of this article is to review the basic statistical concepts that every doctor should know to be able to perform and/or detect quality research, as well as to underline the most frequent errors committed when interpreting statistical results. We review the general concepts about data synthesis and differentiation of the different types of measurements, hypothesis testing and errors that can be committed doing it, the real meaning of the "p" value, differentiation between statistically significant and clinically relevant results, the importance of confidence intervals as a measure of significance and clinical relevance, the confusion generated between two concepts that are different as standard deviation and standard error, and the criteria that govern the selection of the adequate statistical tests to evaluate relationships between variables.