Forums: Validation Rule Suggestions
Hi!
This and other similar rules, which compare dates to the subject latest disposition date, are quite notorious for their false-positive results. In most cases the issue is related to data collection process, when a study termination event was not captured at all. Your example is another implementation issue.
SD0082 rule has long history and is under discussion now. I believe that the industry can figure out the best solution soon.
Thanks,
Sergiy
Ok, thanks, Sergiy. (and thanks for the quick reply)
We came through a similar situation in dm.xpt: subject's RFSTDTC reported after RFENDTC when they had the same date, but the time was recorded only for RFSDTC.
Our solution was to represent both RFSTDTC and RFENDTC as date (without the time)...
The SEND standard allows you to represent dates in yyyy-MM-dd format (no time info).
Hi Amelia,
I do not think that it’s a good practice to modify collected data. The discussed issue is definitely a programming bug of the validator. You do not need to reduce precision of your data to avoid false-positive warning messages.
Best Regards,
Sergiy
When the disposition date (DSSTDTC) is specified with a precision of date/time (e.g., 2012-12-31T10:59:03) whereas the exposure end date (EXENDTC) is specified only to the date-level (e.g., 2012-12-31), rule SD0082 ("Exposure end date must be less than or equal to latest disposition date") fires.
It appears as though the processing assumes that the EXENDTC without a time specified must be 12am (e.g., 2012-12-31T00:00:00), so it will always appear as before the DSSTDTC.
Can you please change this rule such that when the precisions differ between the two dates, the comparison is done based on the date alone? (Another solution could be setting the time portion to 23:59:59 when there is no time present).