Importing search results
One file per database search, with its source, date and search string. The formats that work, and how the imports become your PRISMA-S search log.
One file per search
Upload the exported results of each database search as its own import, and give it:
- the source (the database or registry);
- the date you ran the search;
- the search string, exactly as you ran it.
Together, the imports are your search log: the paper reports them in a PRISMA-S appendix, and the PRISMA flow diagram counts the records identified per source.
Formats
| Format | Typical source | Notes |
|---|---|---|
| RIS (.ris) | Most databases and reference managers | The most common choice |
| BibTeX (.bib) | ACM, arXiv, reference managers | Citation keys are kept; statuses can be imported as already included |
| PubMed/MEDLINE (.nbib) | PubMed | |
| Web of Science | Web of Science | Plain-text export |
| CSV | Spreadsheets, Scopus | Columns are matched by name (title, abstract, authors, year, DOI…) |
| EndNote XML | EndNote |
Export everything the database offers: abstracts and DOIs make screening easier and duplicate detection more reliable.
After an import
- The record count appears in the import log. Records found in several searches are merged in the next step, Duplicates.
- You can import more searches at any time; new records join the queue for screening.
Reproducing a published review
Imports can carry statuses: a BibTeX file of studies a published review included can be imported as already included, to rebuild that review's later stages. This is how the TinyML example in the README reproduces part of a published IEEE Access review.