Skip to Main Content

Evidence Synthesis

Stages of Evidence Synthesis

Evidence Synthesis methodologies follow standardized guidelines that provide details on how to conduct a rigorous synthesis and report on the process and findings. While methods vary, the process can be broken into seven stages, two of which are focused on reporting. Protocol development includes the registration/publication of the protocol. See our subpages for further details on this process.

                                   

Evidence Synthesis Process Overview
stage details
Preparation
  • Identifying the need for evidence: Recognize a problem, conflict, or decision requiring evidence-based solutions, typically arising from a policy issue, strategy decision, or management challenge. This needs assessment will be the first step in forming the research question, which will also impact the type of synthesis you choose.
  • Assess feasibility: Assess the viability of conducting a robust synthesis by evaluating factors like the complexity of the question, availability of relevant research, resources required, and potential impact of findings.
  • Establishing a review team: Form a multidisciplinary team with the necessary subject matter and methods expertise, while ensuring independence of judgment and addressing conflicts of interest.
  • Determining the evidence synthesis type: Taking into consideration the information needed, determine the most suitable methodology. Be sure to review conducting and reporting guidelines in your review and look at published syntheses. 
  • Identify and engage with potential stakeholders: Engagement with stakeholders throughout the synthesis process is important in ensuring that the synthesis addresses relevant questions and that the findings are clearly communicated.
Protocol Development
  • Defining the question and scope: Refine the evidence need into a clear, answerable question using frameworks such as PICO/PECO/PIO. 
  • Developing inclusion/exclusion criteria: Develop detailed criteria by which information sources will be selected, basing these elements on the information needed and types of data you expect to retrieve and use in the synthesis.
  • Developing a search strategy: Create a comprehensive plan for searching literature. Include details on resources to be searched, search methods, final search strings to be employed, and any other methods of gathering information such as stakeholder input or AI systems to supplement systematic searches.
  • Planning the synthesis: Follow established guidelines to develop a detailed protocol that outlines all steps to be taken, including the search strategy, eligibility criteria, data coding/extraction, and synthesis methods.
  • Scoping & Refining: Conduct preliminary searches to assess the size and nature of the evidence base, refine the question, selection criteria, and resource needs based on scoping findings and stakeholder input.
Searching
  • Conducting the search: Execute a comprehensive, reproducible search strategy across diverse sources (e.g., bibliographic databases, grey literature), documenting all search terms, databases, and search dates.
  • Deduplicate citation data: Remove any duplication across search results using automated and manual processes to ensure proper metrics are reported and to speed the screening process.
  • Manage and document resulting citation data: Use software to manage references enhance and clean metadata to prepare it for screening. Tag data to ensure provenance tracking, and in later stages organize data by PICO/PECO concepts for efficient synthesis.
Screening
  • Training & screener agreement: Provide training on the selection criteria for each phase of screening. Ensure consistent application of criteria between screeners using a subset of the citation data. Document all methods and results for reporting.
  • Screen citation titles/abstracts: Apply predefined eligibility criteria to titles, abstracts, and full texts, ideally with two independent reviewers to minimize bias.
  • Screen full-text: Apply detailed criteria to full texts to determine if they meet inclusion criteria, with a brief description of the rationale for inclusion/exclusion.
  • Document decisions: Maintain detailed records of all screening decisions, including inclusions, exclusions, and unclear cases, with justifications for exclusion at the full-text screening level for transparency.
Data Extraction
  • Data coding: Systematically capture relevant study characteristics (metadata) using standardized forms, whether for Systematic Reviews or Systematic Maps.
  • Data extraction: For Systematic Reviews, extract pertinent outcome data from included studies with precision and consistency, considering the synthesis methods to be employed. For Systematic Maps, code studies according to the pre-defined categories, consider the visualizations that will be created and the organizational structure needed for synthesis/analysis.
Synthesis
  • Systematic Reviews: Synthesize data using quantitative (meta-analysis), and/or qualitative methods (narrative synthesis), depending on the evidence base and question.
  • Systematic Maps: Provide an overview of the existing evidence base, often using tables, figures, and visualizations to convey key findings and patterns.
Reporting*
  • *Publish/Register protocol: Register or publish the protocol in a recognized repository to ensure transparency and avoid duplication of efforts.
  • Interpretation and reporting: Interpret the synthesized evidence, acknowledging limitations, strengths, and implications for addressing the review question.
  • Preparing a comprehensive report: Prepare a report adhering to reporting standards (e.g., ROSES for Systematic Reviews, ROSES-Map for Systematic Maps) with transparent documentation of all methods and deviations from the protocol.