Preferred reporting items for systematic reviews and meta-analyses (PRISMA) statement and publication bias

Abstract

Publication bias is a major problem in evidence based medicine. As well as positive outcome studies being preferentially published or followed by full text publication authors are also more likely to publish positive results in English-language journals. This unequal distribution of trials leads to a selection bias in evidence l level studies, like systematic reviews, meta-analysis or health technology assessments followed by a systematic failure of interpretation and in clinical decisions. Publication bias in a systematic review occurs mostly during the selection process and a transparent selection process is necessary to avoid such bias. For systematic reviews/meta-analysis the PRISMA-statement (formerly known as QUOROM) is recommended, as it gives the reader for a better understanding of the selection process. In the future the use of trial registration for minimizing publication bias, mechanisms to allow easier access to the scientific literature and improvement in the peer review process are recommended to overcome publication bias. The use of checklists like PRISMA is likely to improve the reporting quality of a systematic review and provides substantial transparency in the selection process of papers in a systematic review.