New feature in concept sets Lock/snapshot

General
1

Background:

The lock feature is an important one because when concept sets are used in they are often used for studies or publications, and it was recommended by users that there be a means of locking concepts against edits as this ensures that definitions remain consistent for study duration and if prevents accidental changes that can compromise validity.

The snapshot is a way to maintain historical versions that won’t be impacted by modifications (for those concepts that are not locked) or by vocabulary updates as new codes may be added to a concept set and it can be challenging to track such changes that can impact results. So, let’s say a study is being done and a concept set is approved for that study. At that time a snapshot is saved to maintain that historical version. Then 6 months pass, and, for some reason, a new analysis needs to be run. The study lead then can reference the snapshot to see if anything has changed with a recent vocabulary update so that they can ensure they are running the same codes against their data of interest.

Lock feature:

Would apply a clear lock (identified in the concept set library to a concept set such that edits cannot be made without the concept set being unlocked which can only be done by the author or administrator.

This is helpful for users to see clearly that a concept is in locked state and can’t be changed this may be because it is being used for a study or a regulatory purpose.

Once something is locked there would be an option to unlock if needed with the ability to snapshot at unlock as well.

Snapshot feature:

Would allow the author or administrator to click a button to save a record of the content of a concept set (included concepts and source codes) at a particular point in time to maintain a record such that if the concept set is modified or a vocabulary update is applied there is a record of the prior version.

For times when users need a historical reference to concept set contents to include expression, included concepts and source codes for historical reference for studies or regulatory submissions.

These will be independent features that can be used in conjunction with one another where appropriate.

One option is to allow for comparison of snapshots to one another for differences. Is this something that would be helpful for the broader community?

@Brad_Rechkemmer

2

Hi @arobis01

How are you?
Have you had a chance to look at the Version and Compare features in concept sets in Atlas? With a little bit of a trick, I think, most of your thoughts could be implemented via the Comment section, while you’d still need to adjust the user’s privileges if you want a clear lock, which is also possible.

For the snapshot thing, it’s of huge importance, especially given the impact of the vocabulary updates on concept sets. But there is the ideological question - we either produce the snapshots and inevitably overload the concept set expressions, or rely on the reference system (vocabulary) and ensure its consistency, versioning, and leverage snapshots and comparisons this way. You can now produce the snapshots from the historical vocabulary versions, and the system works pretty well. Except, you may lose or earn some concepts if you change the composition of your vocabulary bundle from Athena - this is something farmed out to users right now.

And I fully support the idea that tools like Atlas would have a better toolkit to address this problem, such as:

  • Compare concept sets between the versions
  • Compare definitions of concept sets and source concepts
  • Compare with different vocabulary version/composition
2 Likes