[The meta-analysis on summary data].

The scientific and medical information is nowadays abundant and easily because of advances in information systems and changes in the number of research. This multiplicity requires a synthesis if one wishes to achieve a specific answer to a question. The meta-analysis is a method to perform such work by combining the results of several studies to synthesize objective as a specific protocol and well reproducible. It also allows to quantify the overall result for all studies considered, and thus obtain a more precise and more acceptable generalization. The meta-analysis on summary data presented here, as opposed to those based on individual data are more frequently used because they make it easier to aim for exhaustivity, essential for synthesis unbiased. The work of synthesis should be detailed in a protocol to ensure reproducibility. It occurs in 4 stages: research studies, selection, data extraction and analysis. Its feasibility is mainly limited to the existence of sufficient numbers of studies that may be included. This method is widely used in medical research since the 1970s, both being also care research. It has become essential for the development of recommendations. The scope is limited only to that of previous studies. The evaluation of the effect of drugs or other therapies but also epidemiology, screening, diagnosis or nursing research are areas where the meta-analysis can be applied.