Skip to Main Content

Evidence Synthesis Methodologies in the Health Sciences

Defining Evidence Synthesis

Evidence Synthesis International defines it as:

"[T]he interpretation of individual studies within the context of global knowledge for a given topic. ... Essential to all evidence syntheses is the use of explicit and transparent methodology in the formation of the questions they address. The transparent methodology encompasses how studies are identified, selected, appraised, analyzed, and the strength of the evidence assessed to answer the questioned posed" (Evidence Synthesis International, 2015).

While systematic reviews are the most well-known type of evidence synthesis, there are many others, including scoping and umbrella reviews. However, all evidence syntheses share at least some requirements: a well-defined research question in an appropriate framework, adherence to methodological standards, a thorough and well-documented search, and transparent reporting.

 

Evidence Syntheses vs. Narrative Literature Reviews

A typical narrative literature review lacks formal methodology and reporting standards, including around the quality, comprehensiveness, and transparency of its search strategy. The table below illustrates some differences between a systematic review, one of the most rigorous types of evidence synthesis, and a narrative literature review.

Narrative Literature Reviews vs. Systematic Reviews.
  Narrative Literature Review Systematic Review
Review Question/Topic Topics may be broad in scope; the goal of the review may be to place one's own research within the existing body of knowledge, or to gather information that supports a particular viewpoint. Starts with a well-defined research question to be answered by the review. Reviews are conducted with the aim of finding all existing evidence in an unbiased, transparent, and reproducible way.
Searching for Studies Searches may be ad hoc and based on what the author is already familiar with. Searches are not exhaustive or fully comprehensive. Attempts are made to find all existing published and unpublished literature on the research question. The process is well-documented and reported.
Study Selection Often lack clear reasons for why studies were included or excluded from the review. Reasons for including or excluding studies are explicit and informed by the research question.
           Assessing the Quality of Included Studies   Often do not consider study quality or potential biases in study design. Systematically assesses risk of bias of individual studies and overall quality of the evidence, including sources of heterogeneity between study results.
Synthesis of Existing Research     Conclusions are more qualitative and may not be based on study quality. Bases conclusion on quality of the studies and provide recommendations for practice or to address knowledge gaps.

Table created by Cornell University Library's Evidence Synthesis Service.

Steps in Evidence Synthesis

The steps required for an evidence synthesis project depend on the type of review you select. You can learn more about those options on the Plan Your Project page of this guide. The flowchart below shows the steps involved in a systematic review and meta-analysis, many of which would also apply to other review types.

table-based flow chart of the steps of a systematic review and meta-analysis

(Tsafnat et al., 2014)

Learn about Systematic Reviews From Our Workshop

This recorded workshop is focused specifically on systematic reviews, but many of the principles apply to other forms of evidence synthesis.