Page History

Submission of the SE dataset is strongly recommended, as it provides information needed by reviewers to place observations in context within the study. As noted in the SE - Description/Overview, the TE and TA datasets should also be submitted, as these define the design and the terms referenced by the SE dataset.

The SE domain allows the submission of data on the timing of the trial elements a subject actually passed through in their participation in the trial. Section 7.2.2, Trial Elements, and Section 7.2.1, Trial Arms, provide additional information on these datasets, which define a trial's planned elements and describe the planned sequences of elements for the arms of the trial.

1. For any particular subject, the dates in the SE table are the dates when the transition events identified in the TE table occurred. Judgment may be needed to match actual events in a subject's experience with the definitions of transition events (i.e., events that mark the start of new elements) in the TE table; actual events may vary from the plan. For instance, in a single-dose pharmacokinetics (PK) study, the transition events might correspond to study drug doses of 5 and 10 mg. If a subject actually received a dose of 7 mg when they were scheduled to receive 5 mg, a decision will have to be made on how to represent this in the SE domain.
2. If the date/time of a transition element was not collected directly, the method used to infer the element start date/time should be explained in the Comments column of the Define-XML document.
3.  
4. Judgment will also have to be used in deciding how to represent a subject's experience if an element does not proceed or end as planned. For instance, the plan might identify a trial element that is to start with the first of a series of 5 daily doses and end after 1 week, when the subject transitions to the next treatment element. If the subject actually started the next treatment epoch (see Section 7.1, Introduction to Trial Design Model Datasets, and Section 7.1.2, Definitions of Trial Design Concepts) after 4 weeks, the sponsor would have to decide whether to represent this as an abnormally long element, or as a normal element plus an unplanned non-treatment element. 
5. If the applicant If the sponsor decides that the subject's experience for a particular period of time cannot be represented with one of the planned elements, then that period of time should be represented as an unplanned element. The value of ETCD for an unplanned element is “UNPLAN” and SEUPDES should be populated with a description of the unplanned element.
6. The values of SESTDTC provide the chronological order of the actual subject elements. SESEQ should be assigned to be consistent with the chronological order. Note that the requirement that SESEQ be consistent with chronological order is more stringent than in most other domains, where --SEQ values need only be unique within subject.
7. When TAETORD is included in the SE domain, it represents the planned order of an element in element in an arm. This should not be confused with the actual order of the elements, which will be represented by their chronological order and SESEQ. TAETORD will not be populated for subject elements that are not planned for the arm to which the subject was assigned. Thus, TAETORD will not be populated for any element with element with an ETCD value of “UNPLAN”. TAETORD also will not be populated if a subject passed through an element thatelement that, although defined in the TE dataset, was out of place for the arm to which the subject was assigned. For example, if a subject in a parallel study of drug A vs. drug B was assigned to receive drug A but received drug B instead, then TAETORD would be left blank for the SE record for their drug B element. If a subject was assigned to receive the sequence of elements A, B, C, D, and instead received A, D, B, C, then the sponsor would have to decide for which of these SE records TAETORD should be populated. The rationale for this decision should be documented in the Comments column of the Define-XML document.
8. For subjects who follow the planned sequence of elements for the arm to which they were assigned, the values of EPOCH in the SE domain will match those associated with the elements for the subject's arm in the TA dataset. The sponsor applicant will have to decide what value, if any, of EPOCH to assign SE records for unplanned elements and in other cases where the subject's actual elements deviate from the plan. The sponsor's methods for such decisions should be documented in the Define-XML document, in the row for EPOCH in the SE dataset table. 
9. Because there are, by definition, no gaps between elements, the value of SEENDTC for one element will always be the same as the value of SESTDTC for the next element.
10. Note that SESTDTC is required, although --STDTC is not required in any other subject-level dataset. The purpose of the dataset is to record the elements a subject actually passed through. If it is known that a subject passed through a particular element, then there must be some information (perhaps imprecise) on when it started. Thus, SESTDTC may not be null, although some records may not have all the components (e.g., year, month, day, hour, minute) of the date/time value collected.
11. The following identifier variables are permissible and may be added as appropriate: --GRPID, --REFID, --SPID.
12. Care should be taken in adding additional timing variables: 
    1. The purpose of --DTC and --DY is to record the date and study day on which data was collected. Elements are generally “derived” in the sense that they are a secondary use of data collected elsewhere; it is not generally useful to know when those date/times were recorded.
    2. --DUR could be added only if the duration of an element was collected, not derived.
    3. It would be inappropriate to add the variables that support time points (--TPT, --TPTNUM, --ELTM, --TPTREF, and --RFTDTC), because the topic of this dataset is elements.