Throughout the mini-series so far there has been an implicit assumption that must be made explicit. Let's start with that today.
When referring to the body of evidence that is being rated, GRADE defines the best available body of evidence as that derived from systematic review. That can be from a pre-existing systematic review or one conducted by or for the guideline panel.
If you need a resource that covers standards on how to conduct systematic review, see the Institute of Medicine (IOM) - now called The National Academies of Sciences, Engineering and Medicine - report Finding What Works in Health Care. In the United States, the IOM report is considered the standard for the conduct of systematic reviews. Another resource is the Cochrane Handbook for Systematic Reviews of Interventions
in general, a systematic review consists of:
•Systematically Developed Clinical Question(s)
•Systematic Search of Multiple Databases
•Systematic Selection of Eligible Studies
•Systematic Quality Assessment of the Selected Studies
•Systematic Data Collection from the Selected Studies
•Systematic Synthesis of the Body of Evidence
It's not a systematic review if even one of the steps is not systematic.
You can evaluate the quality of a completed systematic review using one of several readily available tools. Two in common use are:
The ROBIS tool
The AMSTAR tool
TheEvidenceDoc is a co-developer of another tool used to evaluate the quality of systematic reviews. It is in current use by CHEST guideline panels, by BJC HealthCare, and other organizations. This tool was developed for internal use but because of features like extra space for recording decision details that facilitate resolution between reviewers and updating of systematic reviews, its use has spread. You, too, can access resources for using the DART tool.
TheEvidenceDoc November 14, 2017