Manually flag a study as a duplicate#402

I’ve had a weird number of studies lately that have separate records in PMC and PubMed (with two different PMIDs) and show up as two records in NK, but upon review are clearly the exact same paper (apparently there were some record keeping issues in the aneurysm field in 2013-2014)- is it possible to flag these in NK so they can be processed as a duplicate? I’ve just been making a separate “duplicate study” exclusion, but that’s not ideal for the PRISMA diagram. Obviously the original problem arises from a PubMed/PMC database error that’s hard to account for, but if there’s a way to manually flag duplicates in NK that would help.

This is an interesting problem, and I see the need. Our current solution (just exclude) is simple to use, but as you point out, can give a false impression in PRISMA.

A couple thoughts off the top of my head:

1. We’re introducing the ability to mark “Multiple Reports” (used for e.g. a follow up study published some time after an original study). This is used for PRISMA, as well as ease of data management during screening/extraction. We’d need to be very careful to differentiate “Duplicate Records” from “Multiple Reports” in the UI, otherwise users may misunderstand their intent & end up with even worse data quality!
2. When marking a duplicate, if any extraction has been performed, one record’s extracted data must be discarded. Not a big deal if you’re moving do discard, tho :)

Also, out of curiousity, could you share an example of these records?

I wonder if the duplicate flagging could be done like the full text unavailable option, where you can designate an exclusion option to change how it’s presented in the PRISMA diagram. Then it would be clear that the study is being excluded, and the PRISMA would be correct. I’ll add you to the nest I’ve been working on, the records are all under their own exclusion reason. Thanks!

Nest: https://nested-knowledge.com/gather/configure_exclusion/1930 obviously only one version of each report is under the “duplicate” exclusion reason, but you’ll have all the info to find the other!

That could be an option! Not too intrusive either. We’d probably want to lightly restructure how our PRISMA settings work, if we add another toggle on top of “No FT”, to prevent the page from getting super busy.

And this is interesting. Looks like PubMed indexed twice, once for print publication & once for electronic!

https://pubmed.ncbi.nlm.nih.gov/?term=Rupture+of+giant+vertebrobasilar+aneurysm+following+flow+diversion%3A+mechanical+stretch+as+a+potential+mechanism+for+early+aneurysm+rupture

To clarify the status of this issue - we marked it in progress when we were specifying & designing. It’s now entering development, so we plan to have it released in 1-3 weeks!

Awesome, thanks Karl!

Added in release 1.44.0. We appreciate your feedback (Erin!) and patience on this one.

To use this feature:

• Navigate to the “Related Reports” tab
• Search for the study of interest (author year or title searches will work)
  -Click “Mark Duplicate” to send the record you’re currently on to the duplicate queue (it may be recovered, as with all dupes).

Happy deduplicating!