| Process Description |
Relevant Links [Under construction] |
|
Protocol
All evidence synthesis projects should develop and follow a protocol. Protocols should be registered publicly if the project is intended for publication in any venue.
- Confirm the scope, viability, and justification for the project
- Develop a protocol (research plan) explaining and providing rational for the process to be used in every step of the project
- Register it publicly before records are collected for screening
- If seeking publication, identify potential publication journal and review their guidelines during protocol development
|
Protocol
|
|
Search
- Develop a comprehensive search strategy for at least 3 databases & relevant additional methods
- Use relevant controlled vocabulary & keywords, appropriately tagged & connected
- Report the search exactly as run
- For publishing a systematic or scoping review, you must use additional methods for identifying sources to enhance the comprehensiveness of your search
|
Search
|
|
Screen (Select)
- All database records are exported out of databases, combined, duplicates removes, and screened in two phases (title/abstract and full text) using pre-defined selection criteria
- Produce an accurate PRISMA flow diagram mapping this process
- In systematic or scoping review for publication, each record / article is screened by more than one person, blinded to each other's decisions (Pilot test & calibrate the screening process with all screeners). A predetermined process is used to resolve differences in decisions.
|
Screen (Select)
- Select and use a citation management tool
- Select and use duplicate removal/screening/review management software
- PRISMA flow diagram resources & guidance
- Review the screening process generally
- Review the screening process for publication
|
|
Extract Data
- Use a data collection form developed or identified during the protocol phase and pilot tested/calibrated across all data
- In a systematic or scoping review for publication, data is extracted from each article by more than one person, blinded to each other's work, & with a predetermined process for reconciling differences (pilot test & calibrate the data extraction process with all team members in this role). A predetermined process is used to resolve differences in decisions.
|
Extract Data
- Consider a review management software with data extraction support or a data extraction tool (forms in Word or Excel can also be used)
- Guidance from Cochrane on data extraction for systematic reviews
- Guidelines from JBI plus additional details on data extraction for scoping reviews
- Additional assistance and sample forms from Himmelfarb Library
|
|
Quality Assessment
This process is (generally) not done for scoping reviews or mapping reviews.
- Use an existing risk of bias or quality assessment tool
- In a systematic review for publication, quality assessment of each article is done by more than one person, blinded to each other's work, & with a predetermined process for reconciling differences. Pilot test & calibrate the quality assessment process with all team members in this role
|
Quality Assessment
|
|
Synthesis
- Synthesis is done by closely examining study characteristics, grouping them based on which studies are similar enough for synthesis, and then applying the most appropriate synthesis methods for the data extracted
- In a systematic review for publication, you must use a process such as GRADE to synthesize your assessment of the quality of the overall body of evidence.
|
Synthesis
|
|
Publishing
- If you are considering publication, identify your intended publication journal as early as possible (preferably in the protocol phase). Check the journal's author guidelines for any expectations related to the type of review your are seeking to publish
|
Publishing
|
