additional Medline field

JWT
PubMed citations are online several weeks or more before their keyword fields are populated. If these fields are completed, Zotero brings them in as tags automatically in a controlled vocabulary, MeSH -- very helpful. But if the import is too early, the tagging cannot be done. If Zotero could also bring in the Medline "Date Completed" (DCOM) field, blank until the item is indexed by Medline, we could at least create a collection from which we could later requery PubMed to get complete records.

There must be a simpler way, but the need is there.
  • adamsmith
    we certainly won't add a Zotero field for this. If you have a suggestion that would work for an existing field or other option (that wouldn't impact other people's workflows) I'm certainly open to adding implementing that.
  • JWT
    I certainly don't want to impact other's workflows and don't understand why it would. It would just add the DCOM field into the info panel. If that field is blank we would know in Zotero that the record is incomplete (untagged by MeSH) when it was imported into the library. If it had a date (automatically brought in on import), we would know it is complete. If the user wanted to complete the tagging he could search on DCOM, produce a PMID list and feed that back into PubMed for an updated collection after merging duplicates. Done periodically in a big batch that's not a ton of work, and it doesn't change the Zotero function or other workflows at all that I can see. What am I missing?

    By the way, thanks for a super product.
  • aurimas
    edited October 10, 2014
    What's planned in this regard is to enable Zotero to go and fetch additional metadata for a given item, which would allow you to automatically populate the tags at a later point. The details of this have not been worked out yet, so I can't say if we would be able to do such updates periodically or only given user input.

    As adamsmith said though, I don't think a DCOM field will be added and I don't see where we would add it to the current metadata. Besides, adding such fields opens up a whole can of worms with people demanding that other referencing-unrelated fields be added.

    Given above (and, say, we would only make Zotero fetch metadata when requested), you would still need to figure out a way to distinguish items that were imported before MeSH. We could tag items imported from PubMed without MeSH terms with some special tag. Not sure that the majority of users would be happy about such a tag appearing in their library though.
  • JWT
    Thanks aurimas. Your approach would work fine for me and is better than what I proposed. PubMed has a lot of users and Zotero is emerging as a key adjunct. There is large continuing investment in MeSH, yet Zotero users must either wait
    significantly for complete metadata or forego the indexing. Clearly this is not a problem caused by Zotero, but being able to update, automatic or not, would be a big plus.
  • adamsmith
    It would just add the DCOM field into the info panel. If that field is blank we would know in Zotero that the record is incomplete
    just to restate what aurimas is saying: there are so many potential fields in so many disciplines that we need to be fairly conservative in what we add. We will add PMID and PMCID in some way, not least since they're required for referencing, but once we start adding fields related to the status of items in specific databases, there will be no end of it.
    We could tag items imported from PubMed without MeSH terms with some special tag. Not sure that the majority of users would be happy about such a tag appearing in their library though.
    we could consider to set such a tag only via hidden, translator specific pref, turned off by default. But that might also take hidden prefing too far. Dan is not a fan.
  • DWL-SDCA
    This is a more nuanced problem than it seems.

    Some articles from some journals JAMA, New Engl. J. Med. have MeSH terms pre-filled. The pre-filled terms will likely change and additional terms will likely be added as NLM indexers do their work. Some records are completed in a few days others in a few weeks.

    However, the status change that will occur when the publisher provides volume, issue , page range, and date is dependent upon the publisher's schedule for sending updates to NCBI. Some immediately send updated metadata for ePub ahead of print articles. Other publishers may have long delays after the issues have been assembled. Some publishers have a short lead-time from ePub status to production while others may have a queue of a year or two.

    PubMed / Medline records have at least two status labels.

    [PubMed -in progress]
    These items have not yet had the final MeSH terms added nor has the record been edited to meet formatting standards -- titles have not yet been placed in sentence case; author names that have been provided in character sets or improper coding have not been moved from "[No authors listed]" to a listing of author's complete names; and other adjustments to various fields.

    Once the editing of the record is complete the status is changed to
    [PubMed - indexed for MEDLINE]

    However, some items will never be assigned MeSH terms. The explanation for that will need to wait for another day.
  • aurimas
    Thanks for that info, DWL-SDCA. Excited to hear that things are more complicated than we had hoped, as always. This whole metadata fetching will have to be very well thought out. I think we might need a way to keep track of per-field (or at least per-item) metadata quality. Probably won't happen soon.

This is an old discussion that has not been active in a long time. Before commenting here, you should strongly consider starting a new discussion instead. If you think the content of this discussion is still relevant, you can link to it from your new discussion.

Sign In or Register to comment.