SD1048

Could you provide any reference?

Hi!

Could you provide any reference to current IG about " is opposite to the intent of --DOSTOT. --DOSTOT is to be used when individual --DOSE is NOT collected. "

However SDTM v.1.2  (2.2.1 Table 2.2.1 on p.8) says that --DOSTOT "Used when dosing is collected as Total Daily Dose."

I agree with you that this check should be removed now.

I think that current SDTM and

I think that current SDTM and SDTM IG documents do not provide clear definition and details for an intend use of --DOSTOT variable. Therefore there are different interpretations: 1. --DOSTOT is some kind of derived variable that summarizes --DOSE or --DOSTXT values during 24-hrs period (if –DOSTOT is populated, then either of –DOSE or --DOSTXT is expected to be present) 2. --DOSTOT is an alternative to –DOSE and –DOSTXT variables (if --DOSTOT is populated, then both --DOSE and --DOSTOT are expected to be not populated) 3. No conditional relationship between --DOSTOT and (--DOSE/ --DOSTXT)

Definitions are there but not exhaustive

The IG introduces ambiguity, or rather it doesn't do much to clear up ambiguity. However the SDTM Model Document on which the IG depends is more direct (page 8):

--DOSE / Dose / Num / Record Qualifier / Amount of --TRT given. Not populated when --DOSTXT is populated.
--DOSTXT / Dose Description / Char / Record Qualifier / Dosing information collected in text form. Examples: <1 per day, 200-400. Not populated when --DOSE is populated.
--DOSTOT / Total Daily Dose / Num / Record Qualifier / Total daily dose of --TRT using the units in --DOSU. Used when dosing is collected as Total Daily Dose.

So according to this, if --DOSE is populated, --DOSTXT is not populated and vice versa, and when you are COLLECTING dose as Total Daily Dose you use --DOSTOT.

This rules out scenario #1; --DOSTOT isn't supposed to be derived. In fact the SDTM Model was changed from version 1.1 to 1.2 to explicitly add the statement about "collecting" --DOSTOT; the earlier models had indicated that it could be "collected or derived". But the "derived" bit was explicitly & purposefully removed in 1.2.

Scenario #2 is probably good & expected practice but isn't mandatory.

Scenario #3 is actually the only one supported by the documentation. The Model fails to define an explicit relationship between --DOSTOT and the other 2 variables, although it is usually good practice not to COLLECT the same information in two different ways. But we cannot point to a documentable rule about that in the current Model or IG. In fact the user notes for CMDOSTOT introduce a bit more confusion by stating, "CMDOSTOT should be used in addition to CMDOSE, and not in place of it." This text is absent in the notes for EXDOSTOT and SUDOSTOT.

So the defensible rules that we are left with are about the mutually-exclusive nature of --DOSE and --DOSETXT and the expectation that --DOSTOT is collected, not derived. That's it; we don't have a SDTM rule about the relationship between (--DOSE, --DOSETXT) and --DOSTOT.

Randall Austin

Hi Randall, I completely

Hi Randall,

I completely agree with you.

Reagrds,

Sergiy