UX Caucus Database Tips

 A few years ago I ran into a search quirk when translating a lengthy search from PubMed to Ovid MEDLINE. My Ovid results were significantly larger and it wasn't immediately obvious what was happening. After much testing and consulting with my colleagues (thanks, Mark MacEachern !), we finally pinned it down to a single MeSH term returning vastly different results between the two platforms .   Ovid - exp Food/ 1,428,580 results PubMed - "Food"[Mesh] 701,490 results We figured out what was causing the difference, but there was nothing in the help documentation that explained why it was happening. Even though the searches appeared to be the same, both MeSH terms were exploded, the platforms appeared to be handling them differently. I had to reach out to Ovid to get an answer for my remaining question. Over the course of a few weeks and many emails, they were finally able to provide a clear explanation: Ovid was including “Drug Terms” in the explode, where PubMed was not

ETA: Many thanks to Whitney Townsend for being a second set of eyes to check my testing strategies! Building on our earlier post, Tip #25: Using Subheadings in PubMed , here are some more fun things to know about PubMed's subheadings...  Did you know that PubMed's subheadings have their own hierarchy? You can see the full list of Subheadings categories in the MeSH database . And just like you find with other MeSH terms, PubMed also automatically explodes subheadings.  For example, when you search "Blood Glucose/analysis"[Mesh] , you are getting all of the relevant lower-level subheadings under "analysis" [Subheading].  This MeSH and subheading combo includes: "Blood Glucose/ analysis "[Mesh] "Blood Glucose/ blood "[Mesh] "Blood Glucose/ isolation and purification "[Mesh] "Blood  Glucose/ cerebrospinal fluid "[Mesh] In the screenshot below, you can see that ORing the other subheadings in doesn't increase or chang