I have a database with a number of references I had already checked for duplicates and screened and marked to the abstraction stage. After discussion with research collaborators, a decision was made to run another related search and include more references. Around 180 new references were put into the original EPPI database. I then screened for duplicates, with the majority of new duplicates being already linked to a reference in the original database. After finding the duplicates (perhaps stupidly), I marked these new references as duplicates accordingly and started abstract and title screening on the new references. I then realised that for the references that were in EPPI already (and thus had been screened to abstraction stage), if a duplicate had been found for them in the new round, all the marked/screening data for these was gone (including the uploaded PDFs). Effectively, the old references that had new duplicates checked against them were now at title and abstract stage of screening.

I did not realise that marking a new set of duplicates in an already existing database would delete data, especially because I didn’t change the master. Is this information recoverable, and is there a way to prevent this from occuring?

Regards