Versions Compared

Key

  • This line was added.
  • This line was removed.
  • Formatting was changed.

...

In this example, the APACHE II score was measured on Day 1 (prior to implementation of treatment), Day 7, and Day 15 or on the day of discharge (whichever was earlier). In the example, as per the SAP, subjects who died on Day 15 or earlier were assigned the highest observed APACHE II score of any of the participants subjects at any time during the trial (worst case imputation for deaths). This imputation method is for example purposes only and not meant to be a recommendation nor considered a common practice. Missing data values of the parameters required for the derivation of the APACHE II score were replaced by the last available assessment. Item level imputation, in this example, is handled by the investigators and/or clinicians at the site and is considered source data, already captured in SDTM. The  --LNKID variable can be used to determine the source or input record used to assess that particular physiological item on APACHE II at that time point. SOC included a variety of supportive therapies that ranged from the administration of supplementary oxygen to full intensive care support, alongside the use of antiviral treatment, convalescent plasma, corticosteroids, antibiotics or other agents.

In the example, the SAP specifies that planned collected visits, including baseline measurements, are to be used in analyses and displays of results. In fact, no Unscheduled Visits unscheduled visits exist in the source data. If the planned physiology assessment is not done or invalid, the investigator may choose to use a previous or repeat assessment of that parameter as the value at that visit for the scoring of APACHE II. The records for 2 subjects are shown in the example datasets. The data for 3 visits is shown for the first subject, with no missing data. The data for 1 visit is shown for the second subject, who died before the Day 15 (Day of Discharge) Visit so the APACHE II total score is imputed to the worst case value of any subjects across any timepoints at that visit. Note that the worst case value is not shown in the sample data.

...