Systematic review and evidence synthesis
Librarians partner with researchers to conduct evidence syntheses. These are syntheses of all previously conducted research on a topic and represent the highest level of evidence in research. The service is available to U of M researchers.
About the service
Evidence synthesis reviews encompass systematic reviews, scoping reviews, meta-analyses, and evidence gap maps, and more. Unlike other types of reviews, these research methods include a reproducible and transparent methodology. For help differentiating between the various types of review, consult A Typology of Reviews (Grant & Booth, 2009).
Work with librarians to increase the quality of reviews and streamline the process. Librarians help you:
- determine if there are existing reviews on your topic,
- develop a protocol to ensure transparency and rigor,
- create search strategies to identify all relevant studies,
- deliver search results formatted for citation managers and evidence synthesis review tools,
- implement best practices for screening, risk of bias assessment, and data extraction and synthesis,
- write the search methodology, and
- determine best evidence synthesis type for your project.
How librarians can help
As you start your evidence synthesis project, librarians can help at either the consultant or co-author level.
Consultant
As a consultant, a librarian can step in at different points of your evidence synthesis review. Librarians can
- provide background information and resources on the evidence synthesis process,
- recommend databases, protocol registration platforms, and citation management software, and
- suggest edits for your search strategy.
This option is a limited time commitment.
Co-author
Co-authoring is a more substantial commitment, and a librarian will typically devote more than a year to partner with you on your evidence synthesis review.
As a co-author, the librarian will be more hands-on and can
- comment on the protocol,
- select databases and grey literature resources,
- write the search strategy,
- translate searches to syntax of all databases,
- perform searches and export them to citation management software,
- perform deduplication, or train your team on the process,
- set up in article screening software, and
- write a portion of the methods section specific to searching.
Tasks and timelines
An evidence synthesis review will typically require a year or more to complete, and librarians’ availability may vary, so please plan ahead and reach out to us as early as you can.
Below is a detailed chart that breaks down the steps of a traditional evidence synthesis review and the librarian co-author’s potential contributions.
Note: these time estimates may vary depending on the project, and some steps may happen concurrently.
Steps in a traditional evidence synthesis review | Estimated time investment | Potential contribution of librarian co-author |
---|---|---|
1. Assemble evidence synthesis review team and select project manager | Varies | Provide guidance |
2. Identify appropriate review methodology | 2 weeks | Provide guidance |
3. Define research question | 2 weeks | Provide information on appropriate question frameworks (e.g. PICO) |
4. Define inclusion/exclusion criteria | 1 week | Provide guidance |
5. Select databases | 1 week | Suggest appropriate databases |
6. Select grey literature resources | 1 week | Suggest grey literature resources |
7. Write search strategy for primary database | 2-4 weeks | Lead writing of the search strategy |
8. Write and register protocol (written compilation of previous steps) | Varies | Provide comments on protocol and guide protocol registration process |
9. Translate search strategy to syntax 2 of all databases (including grey literature) | 2-4 weeks | Translate search strategy |
10. Search and export results into citation management software | 2 weeks | Perform searches and export results |
11. De-duplicate results | 2-4 weeks | Perform de-duplication, or train your team on the process |
12. Title and abstract screening | 2–3 months* | Recommend article screening software and advise on use of software |
13. Retrieve full-text articles | 1 month* | Train team on full-text article retrieval |
14. Full-text screening | 2–3 months* | Provide guidance |
15. Risk-of-bias assessment | 2–3 months | Provide guidance |
16. Data extraction | 2–3 months | Provide guidance |
17. Meta-analysis or synthesis of results | 2–3 months | Provide guidance |
18. Write the manuscript | 2–3 months | Write information retrieval portion of the methods section |
* Timeframe can vary significantly depending on number of citations identified for screening.