Evidence Synthesis Methods Interest Group

Hello everyone,

I'm working as part of a team on a systematic review of hearing aids and quality of life. Originally, a different librarian (at a different institution than mine) was going to work on this project and drafted the initial search strategies. They ran the strategies in EMBASE (Ovid) and Medline (Ovid) initially in June.

I was re-running these two searches when I noticed that, in EMBASE, certain search lines were yielding less results in September than in June.

Here are some examples from 2 different lines:

12 exp optimism/ (7252 results in June; 7091 results in September)

19 (memory$ or memori$).mp. (542937 results in June; 541956 results in September)

These aren't huge differences, but normally results go up as more records are added.   Does anyone know why results might go down instead? I used the exact syntax given from the previously run search.  I'm mostly just wanting to make sure I'm not risking losing any potentially relevant results for the researchers.

------------------------------
Elizabeth Sanders
Research, Engagement, and Learning Librarian
Lamar University, Mary and John Gray Library
She/Her/Hers
------------------------------

Hello Elizabeth,

I ran into a similar issue a few years ago when updating a search for a project I was working on.  Essentially, I performed the search in February 2020 and then again in September 2020, and the September searched retrieved 43 fewer results overall.  To add to the strangeness, when I applied a date limiter to the September search to only retrieve the new items since the February search, it said there were around 200 new items in that time period.  I reached out to Embase at the time, and they told me that an Embase Release had been put out shortly after my first search which was causing the decrease in total results in September.  They didn't elaborate too much, but I took it to mean that they had changed and updated some of their indexing, which was causing this reduction.  As a side note, I don't know if it would matter or not, but I was using embase.com rather than ovid.

I don't know if that helps at all, but I thought I would let you know you are not alone. :)

-Anna

------------------------------
Anna Biszaha
Research & Education Librarian
She/Her/Hers
------------------------------