I have a large subset of consent documents (about 450,000) that have a unique composite key, meaning two keywords are used to uniquely identify the document. Patient's sign two or three consents per encounter. I need to retroactively fill a keyword with the user who processed the document in the third party system. The sending system stopped sending this one field for a period of time, though that was fixed there are thousands of documents missing this value.
I need to use Both the Encounter Number and the Consent Form Type to identify uniqueness, in order to accurately assign the user who processed the document.
Looking around it sounds like my best bet for looking up a dataset based on multiple values is to use a Reverse Autofill Keyset Lookup. But this only works when manually reindexing?
Either way, I still need to update 450,000 documents with the missing keyword.
At the moment I am considering the following as my only option. Please let me know if there is another/better way to do this:
Export my current Encounter Based Autofill Keyset and import into SQL Server
Add a new keyword to the document with an auto incrementing numerical value (for another point of uniqueness)
Run the Existing documents thru workflow to update the auto incrementing numerical value on all documents (can't copy or compare the doc handle)
Write a SQL query to add all of the keywords, from the existing documents, to the data set (including the new numerical unique value)
Use SQL to add the user from the 3rd party system to the data set
Create a new Autofill Keyset in Onbase to import the updated dataset (primary based on new numerical unique value, can't expand on doc handle)
Replace the current Autofill Keyset on this document with the new one
Run all documents thru a workflow process to expand the keyset (populate the missing keyword)
Use SQL queries to Validate the data in Onbase
Remove the new keyword from the documents
Replace the Autofill Keyset on the Document Type with the original (keyset updates from HL7 processor)
Remove values from the new autofill keyset and rename it to "Do not use..."
The danger with this is removing the Current Autofill Keyset assigned to this document type as it is updated from the HL7 processor. Users would not be able to find the new patients during registration. The process would need to take place off hours for the clinics, overnight likely. Otherwise, I would need to change the document type on the entire subset of documents to sort-of offline process them during hours. Another option would be to add a workflow step to keep both keysets in synch during the process. Either would add more work to an already complicated process or require overnight work. I will consider these options until I am ready or come up with a better solution.
As a side note, I will likely add the Document ID (GUID) from the sending system as a hidden keyword on all documents going forward, to simplify this process in the future. We have another project that has a similar situation, and for that project we did include the vendor's Document ID on all Onbase documents to confirm uniqueness between both systems. For now, I will still need to create a new field in Onbase for uniqueness, to tie everything back when the updated dataset is imported.
