| Issue Number | 5173 |
|---|---|
| Summary | [Term] Punctuation display option for Modify Drug Terms Eligible For Refresh From EVS report? |
| Created | 2022-11-01 11:59:52 |
| Issue Type | Inquiry |
| Submitted By | Osei-Poku, William (NIH/NCI) [C] |
| Assigned To | Kline, Bob (NIH/NCI) [C] |
| Status | Closed |
| Resolved | 2023-03-16 12:53:50 |
| Resolution | Duplicate |
| Path | /home/bkline/backups/jira/ocecdr/issue.331445 |
Note: This may need further discussions before implementation so I am not including the acceptance criteria yet until it is clear how this request will be implemented.
As a CDR Terminology editor, I want to be able to run the Drug Terms Eligible For Refresh From the EVS report with and without reported punctuation changes so that I can more effectively review and process them.
Currently punctuation changes are reported as part of this report. However, in many cases, the changes are the result of differences in style and therefore not needed. Examples:
"T-cell"
"T cell"
"B-cell"
"B cell"
"T-lymphocyte"
"T lymphocyte"
"B-lymphocyte"
"B lymphocyte"
Instead of requesting the software to ignore these differences in a blanket way or keeping a list of hard-coded terms or punctuations to ignore, it will be helpful to rather be able to run the report to show or not show the punctuation changes.
Also, worthy of note is the new request for a report to mark certain terms for later review OCECDR-5171. However, this new request is for a different use case but could be potentially consolidated into OCECDR-5171
Fundamentally, the report uses term name normalization to help answer three questions (I say "help" because there are other factors which kick in for some of these questions, such as the Reviewed status for a name):
How do we determine which terms are different between the CDR and the EVS, and therefore should appear on the report?
Which names/definitions do we show for the terms selected in answering the first question?
Which names/definitions do we store when we update a term in the CDR using values from the EVS?
My understanding of what this ticket is requesting is that it will only affect the answer to the second of these three questions (what do we show for the selected terms?). This understanding is based on the language in the ticket's description which says "... it will be helpful to ... show or not show [emphasis added] the punctuation changes." Is this understanding correct?
To implement this version of the report, we will need to revisit the options for normalization described in the original ticket, as well as the followup questions. The JIRA mechanism for linking directly to comments is broken, so you will need to bring up the original ticket, scroll to the bottom, click the three vertical dots and select Show original comment, scroll back down to the bottom and then click Load 10 older comments until that button is gone (you can hold down the shift key when clicking that button to fetch them all). Then do a browser search for the comment marked "Wednesday, 24 Nov 2021 11:36 AM" and "Thursday, 2 Dec 2021 03:59 PM" and decide which normalization option you want the software to use for punctuation, as well as any of the questions in the second comment which would change in light of the different approach to normalization for the alternate version of the report (for example, question #7; look for answers which say things along the lines of "this question is less meaningful (or is moot) given the decision to use the most conservative approach to normalization").
I propose adding this alternate version of the report as a separate menu option on the Admin Term Report menu.
It would probably be helpful to know what advantages are anticipated from having a version of the report which doesn't always show you why the terms don't match between the CDR and the EVS. Can you provide a description of those advantages?
I would like to be able to see a report where the changes are substantive, not just hyphens/no hyphens (ususally having to do with B and T cells).
But you understand that the decision as to which terms to be shown will still use the existing algorithm for normalizing names and definitions, so you'll have the same terms on the report, but you won't know why.
I would like to have a report that gives me the option of seeing the subset that does NOT just have hyphen differences. I'm not sure what I won't know.
Please go back and read my original comment on this ticket, in which I described the specifics of how I understood the request for this ticket. William confirmed that this understanding was correct in this morning's weekly PDQ status meeting. So the logic of which terms are selected for the report will not change in this alternate version of the software.
Ignoring for a moment the unresolved parts of the specific requirements, can you explain what the issue title means?
Mary and I met and agree that the logic report will not change.
But you understand that the decision as to which terms to be shown will still use the existing algorithm for normalizing names and definitions, so you'll have the same terms on the report, but you won't know why.
Can you clarify why the "new" report will still show the same terms?
The request says "...it will be helpful to rather be able to run the report to show or not show the punctuation changes." So it's just concerned with what gets displayed for the terms which are eligible to be refreshed. It doesn't say anything about changing the set of terms which are listed as eligible. Make sense?
Ok. Understood. I thought by asking to change the set of terms that are eligible, we would also be asking to change how to determine which terms are different between the CDR and EVS (which we wanted to avoid). But it is looking like we may have to do that.
Closing this ticket because the new interface developed in OCECDR-5202 serves the same purpose.
Elapsed: 0:00:00.001357