This guide provides an overview of the steps in preparaing a systematic review (SR). For more detailed information please consult the sources used to create this guide at the bottom of the page.
The review team is responsible for managing and conducting the SR. In order to minimize bias and produce a high quality SR, multidisciplinary teams of two or more qualified individuals should be organized.
First time reviewers should work with others who are experienced in the process.
The team should consist of the following:
Potential conflicts of interest (COI) should be disclosed for each team member, including financial, professional or intellectual bias. Consideration for exclusion should be given for individuals whose COI may diminish credibility of the review.
It is important to develop a focused, answerable, well-formulated question as this will determine many other components of the review, including the search for studies, data extraction, synthesis, and presentation of findings.
The following factors should be considered when formulating a research question:
Formulating a precise question can be difficult. The PICO mnemonic is a structured format used to improve the scientific rigor of a SR.
Image Source: http://books.nap.edu/openbook.php?record_id=13059&page=68
Conducting a SR is an expensive and time consuming process. Before embarking in the process the team should ensure that another SR does not already exist. It is useful to search the following databases for completed SRs or protocols for those in progress.
Once it has been determined that a SR should be conducted the next step is to prepare a protocol. A protocol is a detailed description of the objectives and methods of the review. It is important for several reasons:
Source: Shamseer (2015)
It is common to amend protocols after the review has started. Common reasons include extending the period of search to include older and newer studies, broadening eligibility criteria, and adding new analyses suggested by the primary analysis. Researchers should not modify the protocol based on knowledge of the results of analyses. IOM - standards for SR paper
The PRISMA-P checklist (below) can be used to develop protocols:
Conducting a SR takes time - anywhere from 6 to 12 months!
A SR has a fixed component and a variable component that increases with the number of citations that must be reviewed. An analysis of time spent showed that it took on average 1,139 hours to conduct a SR. Hours were broken down into the following:
The Cochrane Group provides an example of a timeline for a 12 month review: