Evidence-based practice: how to evaluate what you read.

Abstract

Systematic reviews and meta-analysis are affected by the studies that are included or excluded. A number of biases are common including excluding unpublished data, non English language publications, and non indexed articles. In addition there can be a differential impact based on the quality and comparability of studies. In this article we review the potential significance of these biases.