[About meta-analyses].

. Based on the publication in this issue of AI&R of a metanalytic assessment of the degree and the quality of the available evidences on the clinical efficacy of an otherwise diffuse utilization of nutritional integration on a spectrum of fragile patients, the methodological comment is focused on the limitations, biases, misleading implications which are today threatening the role of metanalytical analyses of existing literature, to produce useful-reliable evidences in areas where the representativeness of the source of data is substantially scarce. The exposure to such risk appears to be specifically worrying in the field of nursing care, due to the very limited availability of randomized and other "controlled" prospective original studies with hard end-points.