This article discusses the potential biases in meta-analysis, particularly in the selection and location of studies. It highlights the issue of publication bias, where studies with significant results are more likely to be published than those without. It also addresses other biases such as language bias, database bias, citation bias, and multiple publication bias. The article emphasizes the importance of examining these biases through sensitivity analyses and funnel plots. It provides examples of how these biases can affect the results of a meta-analysis, such as in the case of two meta-analyses of low molecular weight heparins in the prevention of thrombosis after surgery. The article also discusses the role of funding sources in publication bias, with studies sponsored by the pharmaceutical industry being less likely to be published. It highlights the importance of including all relevant studies in a meta-analysis and the need for transparency in the selection process. The article concludes that biases in publication, location, and inclusion are a serious problem in meta-analysis and that critical examination of these biases is essential for reliable results. It also suggests that efforts to identify as many controlled trials as possible through manual searches of medical journals can help reduce bias. The article references several studies and reviews that support these findings.This article discusses the potential biases in meta-analysis, particularly in the selection and location of studies. It highlights the issue of publication bias, where studies with significant results are more likely to be published than those without. It also addresses other biases such as language bias, database bias, citation bias, and multiple publication bias. The article emphasizes the importance of examining these biases through sensitivity analyses and funnel plots. It provides examples of how these biases can affect the results of a meta-analysis, such as in the case of two meta-analyses of low molecular weight heparins in the prevention of thrombosis after surgery. The article also discusses the role of funding sources in publication bias, with studies sponsored by the pharmaceutical industry being less likely to be published. It highlights the importance of including all relevant studies in a meta-analysis and the need for transparency in the selection process. The article concludes that biases in publication, location, and inclusion are a serious problem in meta-analysis and that critical examination of these biases is essential for reliable results. It also suggests that efforts to identify as many controlled trials as possible through manual searches of medical journals can help reduce bias. The article references several studies and reviews that support these findings.