Evaluation of the literature: evidence assessment tools for clinicians.

The progressive improvement in the quality of scientific articles has led to an increase in difficulty in reading and interpreting them so that now clinical knowledge and experience must be complemented by methodological, statistical and computer skills. The aim of this article is to offer practitioners the tools, the simplest keys, that will allow them to understand and critically judge the results of scientific studies. The "peer-review" process of a clinical article submitted to a journal is described and the Science Citation Index and the Impact Factor are presented to the reader as essential instruments to evaluate a specific article's impact and the impact of a given journal on the scientific world, respectively. An article should be evaluated on the basis of some key issues which include, at least, an assessment of methodological aspects, a critical analysis of the statistical component and a proper understanding of the clinical impact of the study outcomes. The standard approach for evaluating the quality of individual studies is based on a hierarchical grading system of research design which represents an essential tool to identify the strength of the evidence of an article. Many different biases may affect the reliability of study results. Randomized Control Trials (RCTs) and Systematic Reviews (SRs) are able to minimize the number of biases and thus are at the highest level of the scale of evidence representing the final steps of a treatment's "career." Finally, moving from research to clinical practice, attention on the clinical impact of study's outcomes is of paramount importance as the literature contains studies (including RCTs) that present statistically significant results but which, from the clinical standpoint, are only relatively or not at all significant. Clinical Practice Guidelines represent a useful tool for practitioners assisting the decision-making process when choosing the most appropriate treatment for their patients.