Description
A special-purpose domain that contains information for each subject's actual and planned visits and contact events.
sv.xpt, Subject Visits — Special-Purpose. One record per actual or planned visit per subject, Tabulation.
Specification
Assumptions
The Subject Visits domain allows the submission of data on the timing of the trial visits for a subject, including both those visits they actually passed through in their participation in the trial and those visits that did not occur.
- Subjects can have 1 and only 1 record per VISITNUM.
- Subjects who screen fail, withdraw, die, or otherwise discontinue study participation will not have records for planned visits subsequent to their final disposition event.
- Planned and unplanned visits with a subject, whether or not they are physical visits to the site, are represented in this domain.
- SVPRESP = "Y" identifies rows for planned visits.
- For planned visits, SVOCCUR indicates whether the visit occurred.
- For unplanned visits, SVPRESP and SVOCCUR are null.
- The identification of an actual visit with a planned visit sometimes calls for judgment. In general, data collection forms are prepared for particular visits, and the fact that data was collected on a form labeled with a planned visit is sufficient to make the association. Occasionally, the association will not be so clear, and the applicant will need to make decisions about how to label actual visits.
- Records for unplanned visits should be included in the SV dataset. For unplanned visits, SVUPDES can be populated with a description of the reason for the unplanned visit. Some judgment may be required to determine what constitutes an unplanned visit. When data are collected outside a planned visit, that act of collecting data may or may not be described as a "visit." The encounter should generally be treated as a visit if data from the encounter are included in any domain for which VISITNUM is included; a record with a missing value for VISITNUM is generally less useful than a record with VISITNUM populated. If the occasion is considered a visit, its date/times must be included in the SV table and a value of VISITNUM must be assigned.
The variable SVCNTMOD is used to record the way in which the visit was conducted. For example, for visits to a clinic, SVCNTMOD = "IN PERSON", visits conducted remotely might have values such as "TELEPHONE", "REMOTE AUDIO VIDEO", or "IVRS".
- The planned study day of visit variable (VISITDY) should not be populated for unplanned visits.
- If SVSTDY is included, it is the actual study day corresponding to SVSTDTC. In studies for which VISITDY has been populated, it may be desirable to populate SVSTDY, as this will facilitate the comparison of planned (VISITDY) and actual (SVSTDY) study days for the start of a visit.
- If SVENDY is included, it is the actual day corresponding to SVENDTC.
- For many studies, all visits are assumed to occur within 1 calendar day, and only 1 date is collected for the visit. In such a case, the values for SVENDTC duplicate values in SVSTDTC. However, if the data for a visit is actually collected over several physical visits and/or over several days, then SVSTDTC and SVENDTC should reflect this fact. Note that it is fairly common for screening data to be collected over several days, but for the data to be treated as belonging to a single planned screening visit, even in studies for which all other visits are single-day visits.
- Differentiating between planned and unplanned visits may be challenging if unplanned assessments (e.g., repeat labs) are performed during the time period of a planned visit.
- Algorithms for populating SVSTDTC and SVENDTC from the dates of assessments performed at a visit may be particularly challenging for screening visits, since baseline values collected at a screening visit are sometimes historical data from tests performed before the subject started screening for the trial. Therefore dates prior to informed consent are not part of the determination of SVSTDTC.
- The following Identifier variables are permissible and may be added as appropriate: --SEQ, --GRPID, --REFID, and --SPID.
- Care should be taken in adding additional timing variables:
- If TAETORD and/or EPOCH are added, then the values must be those at the start of the visit.
- The purpose of --DTC and --DY in other domains with start and end dates (Event and Intervention Domains) is to record the date on which data was collected. For a visit that occurred, it is not necessary to submit the date on which information about the visit was recorded. When SVPRESP = "Y" and SVOCCUR = "N", --DTC and --DY are available for use to represent the date on which it was recorded that the visit did not take place.
- --DUR could be added if the duration of a visit was collected.
- 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 visits.
- --STRF and --ENRF could be used to say whether a visit started and ended before, during, or after the study reference period, although this seems unnecessary.
- --STRTPT, --STTPT, --ENRTPT, and --ENTPT could be used to say that a visit started or ended before or after particular dates, although this seems unnecessary.
- SVOCCUR = "N" records are only to be created for planned visits that were expected to occur before the end of the subject's participation.
Overview
Content Tools