Issue Number | 3170 |
---|---|
Summary | Write script to patch import information for NCT00319748 |
Created | 2010-06-02 14:58:05 |
Issue Type | Improvement |
Submitted By | Kline, Bob (NIH/NCI) [C] |
Assigned To | Kline, Bob (NIH/NCI) [C] |
Status | Closed |
Resolved | 2010-06-03 11:23:31 |
Resolution | Fixed |
Path | /home/bkline/backups/jira/ocecdr/issue.107498 |
BZISSUE::4857
BZDATETIME::2010-06-02 14:58:05
BZCREATOR::Bob Kline
BZASSIGNEE::Bob Kline
BZQACONTACT::William Osei-Poku
[From email thread]
Hi Bob,
How can we get 484380 to be updated with the correct status (of Completed)? It is currently completed in CTgov but active on cancer.gov. When it was removed from duplicate table yesterday, it rather created a new CTGov protocol – 674727 but with an Active status. So there are currently three trials with the same NCT ID but none of them has the correct status from CTgov. Two of them are CTgov trials (one of the two was just created this morning - 674727. We will be blocking it). The other is an InScope protocol which is blocked - 490083.
Thanks,
William
{Bob's response]
This sound awfully familiar. What is it you would like me to do? Manually force the ctgov_import table to direct updates for NCT00319748 so they get stuffed in CDR484380 instead of CDR674727?
[William replies]
Yes. That would be the preferred solution.
Thanks!
BZDATETIME::2010-06-02 15:25:48
BZCOMMENTOR::Bob Kline
BZCOMMENT::1
I wrote a script to fix the row by hand and imported the updates.
!8128 Wed Jun 02 15:22:28 2010: Updated CDR484380 from NCT00319748
Please review the results carefully.
BZDATETIME::2010-06-03 10:24:48
BZCOMMENTOR::Bob Kline
BZCOMMENT::2
[Response to offline question from Lakshmi]
> I am totally confused. How could a new CTGOV protocol record be
created
> 674727 be created with the wrong status?
Here's what happened. The ctgov_import table had an older version of the trial document from NLM, marked as a duplicate. CIAT went in and changed the 'duplicate' disposition to 'import requested' (which also sets the cdr_id column to NULL). The next morning we got the latest information on the trial from NLM, but the status was not "Recruiting" or "Not yet recruiting" and we didn't have a value in the cdr_id column, so we didn't save what we had just received from NLM. We tried to drop the row from the table (which you had asked us to do back in 2004 for closed trials which would otherwise have shown up on the review page) but that failed, because there were actually rows in the related ctgov_import_event table (apparently the trial was imported before it was decided it was a duplicate) and the foreign key constraint prevented the drop. So the trial information was imported anyway, but with the older version of the document. In this particular case, if CIAT had also set the 'force' flag for the trial, we would have picked up the newer information, overriding what we do by default for closed trials. This all matches the documentation at http://bach.nci.nih.gov/DownloadCTGovProtocols.html.
BZDATETIME::2010-06-03 11:23:31
BZCOMMENTOR::William Osei-Poku
BZCOMMENT::3
(In reply to comment #1)
> I wrote a script to fix the row by hand and imported the
updates.
>
> !8128 Wed Jun 02 15:22:28 2010: Updated CDR484380 from
NCT00319748
>
> Please review the results carefully.
Reviewed. Looks Good. Thank you!
Closing issue.
Elapsed: 0:00:00.000752