FDAC183 Missing value for CMDOSU, when CMDOSE, CMDOSTXT or CMDOSTOT is provided

Dose of Medication is expected to be provided as a number

Hi Manolya, 

This is an example of potentially invalid data collection design.

Dose of Medication is expected to be provided as a number. Dose Text (CMDOSTXT) variable is designed to capture a range of dose as "number-number" format. For example, '10-19' or '20-40'. However, 'Low' and 'High' are invalid values for --DOSTXT variables. 

I assume that in your study 'statins therapy' represents different inhibitors. Correct SDTM implementation would keep a name of inhibitor in CMTRT and 'statins therapy' info in CMCAT or CMSCAT variables.

Combining all statins therapy into a single term and representing all dosage as 'low intensity', 'medium...' and 'high...' may be enough for your study SAP. However, such approach may potentially limit a use of your study data for any integrated analysis. 

Regards, 

Sergiy

CMDOSTXT

As an addition to Sergiy's answer, take into account that this is an FDA rule, not a CDISC rule.
The SDTM-IG (3.2) states: "Dosing amounts OR a range of dosing information COLLECTED in text form. Units MAY be stored in CMDOSU. EXAMPLE: 200-400, 15-20. Not populated when CMDOSE is populated". For CDISC, the range requirement is not there: it is just an example, nothing more. The CDISC definition can read as "dosing amount collected in text form" so if you collected it as "low", "high", it should go into CMDOSTXT.
My own impression is that the FDA over-interpreted the SDTM-IG here.

CMDOSTXT

Hi both,

Thanks a lot for your input! I am aware that CDISC has different thoughts on things, but for a FDA submission I don't think that errors like this will be accepted...right? Regarding the collection of the data, we only have "statins" pre-specified in the CRF (with no information on inhibitors). Investigators then can selected between different "dosages" collected as "high", "low", etc. So I don't know where to collect those within CM. Is it acceptable to put this information maybe in CMSCAT (like "LOW INTENSITY", "HIGH INTENSITY", etc?) whereas CMCAT could be something like  "STATINS THERAPY" (though we have CMCAT already populated) and CMTRT = "Statins" (as we do not have any other information).

Thanks in advance!

Manolya

FDAC183

Hi Manolya,
Are you sure it was rule FDAC183? Rule FDAC183 does NOT say anything about that CMDOSTXT must have the format "number-dash-number". The exact text is:

"Dose Units (--DOSU) must be populated, when Dose per Administration (--DOSE), Dose Description (--DOSTXT) or Total Daily Dose (--DOSTOT) is provided".
Also in the publications of the FDA, I did not find any such a requirement.

So is the "number-dash-number" requirement a "secret" requirement?

Regarding where to put it in: I would still do it in CMDOSTXT and then write in the SDRG that this error is a false positive. CMSCAT is not the right place for it. If you do not want to put it in CMDOSTXT because of this bug in the software, I would suggest that you create a supplemental qualifier and then put it in SUPPCM.

You will be OK! in terms of FDA submission

Hi Manolya, 

You will be OK! in terms of FDA submission. You need to provide explanation for reported validation finding in Reviewer's Guide. Also, I would recommend to address this quite unusual data collection process for Concomitant Medication in the domain description section 3.3.x. 

Kind Regards, 

Sergiy