EBMS Tickets

As far as I can tell, the record for this relationship was created in the Drupal 7 system and migrated correctly to the Drupal 9 site. According to the databases for both versions of the site, the relationship was recorded by Bonnie at 12:27:59 on Wednesday February 22, and the two article IDs are 879141 and 895169. I think the next step would be to consult with Bonnie and see what she has to say about why she recorded the relationship. Two things to bear in mind:

• the form for entering the relationship information uses a free-text field for the related article IDs, and it's always possible for the user to enter an incorrect ID

• PubMed has a history of playing fast and loose with PMIDs

Ok good. I just wanted to make sure that this was not an issue with duplicates.

I'll ask Bonnie tomorrow why she labeled it as a duplicate.

Thanks!

Here's a description of the article that PubMed "disappeared": https://aacrjournals.org/cancerdiscovery/article/12/11/2487/709995/Myriad-Cancers-Found-with-One-Blood-DrawMyriad

I have attached the XML we pulled from NLM before they removed it.

Here's another extremely intriguing little discovery. If you enter

myriad cancers found with one blood draw

(which is the title of the "disappeared" article), in the PubMed search box, you get exactly one hit:

You'd have a hard time convincing me that this is a coincidence. We have two facts:

• Bonnie recorded a relationship between the two articles

• PubMed finds one of these articles when you search for the other

I can't believe that these two facts are not connected in some way.

If you ask NLM about the relationship between these two articles, I wouldn't be too surprised if you get a pretty interesting and revealing story.

Very intriguing, but these two studies are still quite different (blood vs saliva) and published 5 years apart so they are definitely not duplicates. But given pubmed's retrieval when the title of the 2022 study is searched, I can see how Bonnie would think these two citations may be related.

I think it is safe to say that this finding is not an indication of any issue with the ebms.

I am glad we investigated because this was such an odd set of citations that I was afraid a non-human error may have occurred. 

Thank you for helping me sort this out.

they are definitely not duplicates

Well, they certainly don't look like duplicates now, but who knows what shell games NLM has played with their internal IDs. Scouring the complete text of the abstract of the only article which showed up in the results set for a search on the blood draw article's title turns up no link whatsoever to the article whose title I entered, so I conclude that NLM's own internal identifiers are what tie these two articles together. Perhaps they recycled a PMID and forgot to clean up all the linkages to what the ID was originally pointing to.

Agreed. 

I asked Bonnie to look at these citations and she doesn't remember any specifics but said it is most likely due to her entering in the wrong ID when making the link. 

Sorry to send us on this wild goose chase but these two citations were just so odd that I wanted to make sure nothing had gone wrong with the related articles data.

I think we can close this ticket.