The first phase of screening involves reading through the titles and/or abstracts of records that have been retrieved by the database searches in order to determine:
Sometimes reading the title gives enough information to conclude that a record is clearly irrelevant and the record can be rejected, but often it's necessary to read into the abstract.
If the information in the title and/or abstract indicates that the source is irrelevant or would not fit the eligibility criteria, the record can be rejected.
If the title and/or abstract indicates that the source could fit the review's eligibility criteria (or if there is not really enough information to fairly judge) the full text of the source should be sought for screening. For example, in the search flow diagram (PRISMA diagram) to the right, the authors' database search retrieved a total of 3034 records. After removing duplicates, they screened 1563 records based on title/abstract and rejected 1385 because they appeared irrelevant, leaving 178 full texts to review.
After title/abstract screening, the majority of the database search results will likely have been rejected as irrelevant, leaving a smaller pool of potentially relevant sources. With the information found in the database records, the full text of the sources can be sought for review (see Accessing Full Texts for details about how to locate full texts through NYU Libraries).
During "full text review", the full text of each paper that made it past title/abstract screening is read and screened according to all the elements of the inclusion/exclusion criteria. This second phase of screening allows the author to ultimately identify the papers that will be included in the appraisal, analysis and synthesis phases of the review.
During full text review, it is customary to indicated the reason why an article is being excluded (e.g,. the reported outcomes don't meet the inclusion criteria). For example, in the search flow diagram to the right, 178 articles were read in full, with 168 being excluded after screening for the reasons indicated in the diagram. They then added 12 studies that were identified based on backward/forward citation searching, resulting in 22 that were synthesized as the data sources in their review.
Callanan, F., Bradshaw, C., Tuohy, T., Noonan, M., Murphy, S., & Grealish, A. (2025). An integrative review of how midwives are screening and assessing for trauma in women within perinatal services. PloS One, 20(7), e0327253. https://doi.org/10.1371/journal.pone.0327253
There are many possible workflows related to screening and managing records for an integrative review.
Ideally, evidence synthesis projects like integrative reviews should be conducted in a rigorous and transparent way (especially if they will be published or distributed more widely). To report a review according to PRISMA standards, it is necessary to be more systematic and careful with record keeping.
Less systematic workflow:
More systematic workflow (suitable for reporting via PRISMA):
Citation management software can be used to:
Covidence Knowledge Base: Importing references in Covidence
Citation management software helps you take notes, organize research, import citations from electronic databases and catalogs, format reference lists in standard styles (MLA, APA etc.); insert footnotes and endnotes directly into your papers.
Come to a Library Workshop or view this comparison list.
If you just need some quick citation formatting without having to download software, check out Zotero Bib!
Covidence Knowledge Base: Importing references in Covidence