SDTM Exposure as Collected (EC)

Description TOBA-693 - Getting issue details... STATUS

An interventions domain that contains information about protocol-specified study product administrations, as collected.

Specification

Assumptions

1. EC Definition
1. The Exposure as Collected domain model reflects protocol-specified study product administrations, as collected.
   1. EC should be used in all cases where collected exposure information cannot or should not be directly represented in EX. For example, administrations collected in mL but protocol-specified unit is mg, or administrations collected in LOZENGE but protocol-specified unit is mg. Product accountability details (e.g., amount dispensed, amount returned) are represented in DA and not in EC.
   2. Collected exposure data are in most cases represented in a combination of one or more of EC, DA, or FA domains. If the entire EC dataset is an exact duplicate of the entire EX dataset, then EC is optional and at the applicant's discretion.
2. Collected exposure log data points descriptive of administrations typically reflect amounts at the product-level (e.g., number of tablets, number of mL).
2. Product Description
1. ECTRT is applicant defined and should reflect how the protocol-specified study product is known or referred to in data collection. In an open-label study, ECTRT should store the product name. In a masked study, if product is collected and known as Product A to the subject or administrator, then ECTRT = "PRODUCT A". If in a masked study the product is not known by a synonym and the data are to be exchanged between applicants, partners and/or regulatory agency(s), then assign ECTRT the value of "MASKED".
3. ECMOOD is permissible; when implemented, it must be populated for all records.
1. Values of ECMOOD, to date include:
   1. "SCHEDULED" (for collected subject-level intended dose records)
   2. "PERFORMED" (for collected subject-level actual dose records)
2. Qualifier variables should be populated with equal granularity across Scheduled and Performed records when known. For example, if ECDOSU and ECDOSFRQ are known at scheduling and administration, then the variables would be populated on both records. If ECLOC is determined at the time of administration, then it would be populated on the performed record only.
3. Appropriate Timing variable(s) should be populated. Note: Details on Scheduled records may describe timing at a higher level than Performed records.
4. ECOCCUR is generally not applicable for Scheduled records.
5. An activity may be rescheduled or modified multiple times before being performed. Representation of Scheduled records is dependent on the collected, available data. If each rescheduled or modified activity is collected, then multiple Scheduled records may be represented. If only the final Scheduled activity is collected, then it would be the only scheduled record represented.
4. Doses Not Taken, Not Given, or Missed
1. The record qualifier --OCCUR, with value of "N", is available in domains based on the Interventions and Events General Observation Classes as the standard way to represent whether an intervention or event did not happen. In the EC domain, ECOCCUR value of "N" indicates a dose was not taken, not given, or missed. For example, if 0 tablets are taken within a timeframe or 0 mL is infused at a visit, then ECOCCUR = "N" is the standard representation of the collected doses not taken, not given, or missed. Dose amount variables (e.g., ECDOSE, ECDOSTXT) must not be set to zero (0) as an alternative method for indicating doses not taken, not given, or missed.
2. The population of Qualifier variables (e.g., Grouping, Record, Variable) and additional Timing variables (e.g., date of collection, visit, time point) for records representing information collected about doses not taken, not given, or missed should be populated with equal granularity as administered records, when known and/or applicable. Qualifiers that indicate dose amount (e.g., ECDOSE, ECDOSTXT) may be populated with positive (non-zero) values in cases where the applicant feels it is necessary and/or appropriate to represent specific dose amounts not taken, not given, or missed.
5. Timing Variables
1. Timing variables in the EC domain should reflect administrations by the intervals they were collected (e.g., constant-dosing intervals, visits, targeted dates like first dose, last dose).
2. For administrations considered given at a point in time (e.g., oral tablet, pre-filled syringe injection), where only an administration date/time is collected, ECSTDTC should be copied to ECENDTC.
6. The degree of summarization of records from EC to EX is applicant defined to support study purpose and analysis. When the relationship between EC and EX records can be described in RELREC, then it should be defined.
7. Additional Interventions Qualifiers
   1. --DOSTOT is under evaluation for potential deprecation and replacement with a mechanism to describe total dose over any interval of time (e.g., day, week, month). Applicants considering ECDOSTOT may want to consider using other dose amount variables (ECDOSE or ECDOSTXT) in combination with frequency (ECDOSFRQ) and timing variables to represent the data.
   2.  Any Identifier variables, Timing variables, or Findings general-observation-class qualifiers may be added to the EC domain, but the following qualifiers would generally not be used in EC:  --STAT, --REASND, --VAMT, and --VAMTU.