12 March 2009
The ‘Guidelines and Guidance’ section of PLoS Medicine, an open-access journal, was launched with the goal of encouraging the highest possible standards in medical research and reporting. Articles that raise awareness of emerging and novel methodological approaches, provide community consensus standards for reporting or carrying out particular types of research, and provide easily digestible guides on conducting commonly used statistical methods or methodological techniques are published in this section. These include articles such as STROBE [Von Elm et al. (2007) PLoS Med 4(10):e296 ; Vandenbroucke et al. (2007) PLoS Med 2007 4(10):e297] and STREGA [Little et al. (2009) PLoS Med 6(2):e22] (see also previous news).
A new guidelines and guidance paper published last week by PLoS Medicine on systematic reviews of genetic association studies aims to describe the important components of the methodology for undertaking a systematic review. The increase in published literature on gene-disease associations has led to an increase in efforts to collate and synthesise this epidemiological evidence. Many systematic reviews and/or meta-analyses are published and are often not explicit in outlining the methods used or even how included studies were identified. The principal aim of a well-conducted systematic review (and meta-analysis) of genetic association studies should be to reliably establish the presence and magnitude of a gene-disease association. Systematic reviews are designed as rigorous research studies, with predefined methods and allow the objective appraisal, critique and synthesis of relevant study data. This is a marked difference from traditional (narrative) reviews which are generally subjective, rarely state how studies were included or assessed for quality, and often support and reinforce the author’s view.
In this paper [Sagoo et al. (2009) PLoS Med 6(3):e1000028 doi:10.1371], the authors (who include PHG Foundation funded staff) outline the general process involved in a systematic review including the need for minimising bias, as well as highlighting some of the issues encountered in the human genome epidemiology field. The search process for eligible articles is described, and a non-statistical overview of meta-analysis presented, along with summary points that highlight the key methodological points for a systematic review of genetic association studies.