Research was conducted in multiple database (MEDLINE from 1965 to 2011, Cochrane Library, and LILACS), and in crossed references with the surveyed material aiming the identification of articles with the best methodological design. Following the findings, critical evaluation of the contents and classification according to the strenght of evidence were performed. The research was conducted between December 2010 and April 2011. For PubMed, were used the following strategies:

No abstract

No abstract

No abstract