You are viewing an old version of this page. View the current version.

Compare with Current View Page History

« Previous Version 2 Next »

This is the Wiki edition of the proposed SDTMIG General Assumption entitled "Disease Milestones and Disease Milestone Timing Variables".

This draft set of assumptions are intended to be appended to the end of SDTMIG Section 4.1.4, Actual and Relative Time Assumptions.

A “Disease Milestone” is an event or activity that can be anticipated in the course of a disease, but whose timing is not controlled by the study schedule. The types of Disease Milestones for a study are defined in the study-level Trial Disease Milestones dataset (TM). For a particular study, they may be events or activities that would have occurred before the study, such as diagnosis of the disease under study, or events or activities anticipated to occur during the study that are not scheduled activities. Disease Milestones during the study are often disease-related events that trigger the collection of data outside of scheduled visits.

4.1.4.11.1 Disease Milestone Name (MIDS)

The occurrence of a Disease Milestone will be recorded in the special-purpose Subject Disease Milestones domain (SM) and in the appropriate Event, Intervention, or Findings domains. In the general-observation-class records for the Disease Milestone, the MIDS variable will be populated with the name of the Disease Milestone. The names of disease milestone are composed of a character string that depends on the disease milestone type, and, if the type of disease milestone is one which may occur multiple times, a chronological sequence number for this disease milestone among others of the same time for the subject. The character string used in the same of a disease milestone is usually a short form of the type of disease milestone. For example, if the type of disease milestone were EPISODE OF DISEASE UNDER STUDY, the values of MIDS for instances of this type of event could EPISODE1, EPISODE2, etc.

In a general observation class record for a finding, event, or intervention which is a disease milestone, MIDS will be populated, but RELMIDS and MIDSDTC will not be populated; the usual timing variables (e.g., --DTC, --STDTC,-- ENDTC) provide timing for this observation and will be used to derive the dates and study days in the Subject Disease Milestones domain.

4.1.4.11.1 Timing relative to a Disease Milestone (MIDS, RELMIDS, MIDSDTC)

Observations made in conjunction with the Disease Milestone use the Disease Milestones Timing variables MIDS, RELMIDS and MIDSDTC to describe the timing of the observation. 

  • MIDS is populated with the name of a Disease Milestone that appears in the SM domain for this subject. MIDS is the “anchor” for describing the timing of the observation relative to the disease milestone. In this sense, its function is similar to –REFTPT for time points.
  • RELMIDS is populated with a textual description of the temporal relationship between the observation and the Disease Milestone named in MIDS. Controlled vocabulary has not yet been developed for RELMIDS, but is likely to include terms such as IMMEDIATELY BEFORE, AT START OF, DURING, AT END OF, and SHORTLY AFTER.
  • MIDSDTC is populated with the date/time of the Disease Milestone, as recorded in the SM domain. Its function is similar to –RFTDTC for time points.

In some cases, data collected in conjunction with a Disease Milestone will not have included the collection of a separate date for the related observation. This is particularly common for pre-study Disease Milestones, but may occur with on-study Disease Milestones as well. In such cases, MIDSDTC provides a related date/time in records that would not otherwise contain any date. In records that do contain date/time(s) of the observation, MIDSDTC allows easy comparison of the date(s) of the observation to the (start) date of the Disease Milestone. In such cases, it functions much like the reference time point date/time (--RFTDTC) in observations at time points.

When a Disease Milestone is an event or intervention, some data triggered by the Disease Milestone may be modeled as Findings About the Disease Milestone (i.e., FAOBJ is the Disease Milestone). In such cases, RELMIDS should be used to describe the temporal relationship between the Disease Milestone and the subject of the question being asked in the finding, rather than as describing when the question was asked. 

  • When the subject of the question is the Disease Milestone itself, RELMIDS may be populated with a value such as “ENTIRE EVENT” or “ENTIRE TREATMENT.”
  • When the subject of the question is a question about the occurrence of some activity or event related to the Disease Milestone, RELMIDS acts like an evaluation interval, describing the period of time over which the question is focused.
    • For questions about a possible cause of an event or about the indication for a treatment, RELMIDS would have a value such as “WEEK PRIOR” or “IMMEDIATELY BEFORE” or even just “BEFORE.”
    • RELMIDS would be “DURING” for questions about things that may have occurred while an Event or Intervention Disease Milestone was in progress.
    • For sequelae of a Disease Milestone, RELMIDS would have a value such as “AT DISCHARGE” or “WEEK AFTER” or simply “AFTER.”
4.1.4.11.1 Use of Disease Milestone Timing Variables with other Timing Variables

The Disease Milestone timing variables provide timing relative to an activity or event that has been identified, for the particular study, as a Disease Milestone. Their use does not preclude the use of variables that collect actual date/times or timing relative to the study schedule.

  • The use of actual date/times is unaffected. The Disease Milestone Timing variables may provide timing information in cases where actual date/times are unavailable, particularly for pre-study Disease Milestones. When the question text for an observation references a Disease Milestone, but a separate date for the observation is not collected, the Disease Milestone Timing variables should populated but the actual date/s should not be imputed by populating them with the date of the Disease Milestone. Examples of such questions: Disease stage at initial diagnosis of disease under study, Treatment for most recent disease episode.
  • Study-day variables should be populated wherever complete actual date/times are populated. This includes negative study days for pre-study observations.
  • EPOCH and TAETORD may be populated for on-study observations associated with Disease Milestones. However, pre-study Disease Milestones, by definition, do not have an associated EPOCH or TAETORD.
  • Visit variables are expected in many findings domains, but findings triggered by the occurrence of a study milestone may not occur at a scheduled visit.
    •  Findings associated with pre-study Disease Milestones are often collected at a screening visit, although the test was not performed at that visit.
    • For findings associated with on-study Disease Milestones but not conducted at a scheduled visit, practices for populating VISITNUM as for an unscheduled visit should be followed.
  • The use of time-point variables with Disease Milestone variables may occur in cases where a Disease Milestone triggers treatment, and time points relative to treatment are part of the study schedule. For instance, a migraine trial may call for assessments of symptom severity at prescribed times after treatment of the migraine. If the migraine episodes were treated as Disease Milestones, then the Disease Milestone timing variables might be populated in the exposure and symptom-severity records. If the study planned to treat multiple migraine episodes, the MIDS variable would provide a convenient way to determine the episode with which data were associated.
  • An evaluation interval variable (--EVLINT or --EVLTXT) could be used in conjunction with Disease Milestone variables. For instance, patient-reported outcome instruments might be administered at the time of a Disease Milestone, and the questions in the instrument might include an evaluation interval.
  • The timing variables for start and end of an event or intervention relative to the study reference period (--STRF and –ENRF) or relative to a reference time point (--STRTPT and --STTPT, --ENRTPT and --ENTPT) could be used in conjunction with Disease Milestone variables. For example, a concomitant medication could be collected in association with a Disease Milestone, so that the Disease Milestone timing variables were populated, but relative timing variables could be used for the start or end of the concomitant medication.
  • The timing variables for start and end of a planned assessment interval might be populated for an assessment triggered by a Disease Milestone, if applicable. For example, the occurrence of a particular event might trigger both a treatment and Holter monitoring for 24 hours after the treatment.
  • No labels