How do I translate my search from one database to another?
Answer
When conducting knowledge synthesis, searching in multiple databases is crucial to ensure comprehensive and unbiased results. By searching across multiple sources, you reduce the risk of publication bias, capture a broader range of perspectives, and strengthen the reliability of your synthesis. However, the search will look different from one database to the next. Here are the steps you can follow to ensure that you're doing it right:
- Develop your search in a primary database, ensuring that you have used all the relevant subject headings, keywords and search fields.
Once you are confident that your search is sound in this database, you can move on to the translation stage. If you come across other terms in other databases that you maybe missed in the first, you can go back and make adjustments.
2. Find the equivalent subject headings for each subsequent database.
Different databases have different subject headings. For example, when searching in Medline, you will use the the MeSH term Polycystic Ovary Syndrome. However, in Embase, the EMTREE term is Ovary Polycystic Disease. You can use the attached worksheet at the end of this post to keep track of your subject headings across databases. You may find that one database uses one subject heading to describe a term, while another has broken that term into two or more subject headings. Likewise, you may not be able to find a subject heading in a database as the term does not exist in that thesaurus.
3. Enter your keywords with the correct syntax.
The keywords you use will remain unchanged across databases. However, the way that you enter these keywords will change depending on the interface. You should also be looking for the equivalent search fields. A search for the keyword diabetes in the title and abstract would look like this in different databases:
Medline = diabetes.ti,ab
CINAHL = TI diabetes OR AB diabetes
Scopus = TITLE-ABS(diabetes)
You should also be checking that you are using the correct truncation, wildcards and adjacency operators. If you are using any kind of generative AI or the SR Accelerator Polyglot tool, please make sure that you are double-checking the output, as these tools, while helpful, are known to make mistakes.
Overall, we recommend running your searches and exporting your results on the same day, as it will be easier when documenting the search process in your methodology.