Knowledge Base for Supply App
Prediction & follow-up stage
Table of contents
Prediction
If you want to define any other kind of prediction rules ((e.g, you want to predict the second visit dispensing for all treatment arms and all dose levels), beside the ones you can define for the randomization and titrations, you can enter overriding custom predictions in this table.
You can define any dispensing predictions for a specific patient group, a specific period, from/to specific treatment arms, visits, dose levels and patient evolution categories.
Entering values in “From treatment arms/From visits/From dose levels” and “To treatment arms/To visits/To dose levels” does not imply that un-entered values comprised between “From” and “To” parameters will be predicted. Only the values entered are predicted. For example, entering “visit 2” in “From visits” and “visit 5” in “To visits” will only predict visit 2 to visit 5 directly and not the visits in between.
⚠️ Make sure that the IRT system is able to implement these custom predictions for your trial.
- This data set and the other prediction modes for randomization and titrations are optional in a trial modeling. Buffers are computed for unpredictable demand (following the site resupply algorithm) and if these prediction features are not used, what was unpredicted will remain unpredicted.
- This table is not automatically populated with the predictions coming from randomization and titrations but you can see if they have been modified or not with the blue badges above the table (see screenshot above).
Follow up stage
In the follow up stage tab, the user defines the potential follow up stage with:
- The follow up visits list, in the right order.
- The visit intervals (and corresponding visit window) between the main stage and the follow up stage. This will define how many days pass between patient discontinuation and the first follow up visit.
- The visit intervals (and corresponding visit windows) between the follow up visits.
- The follow up dispensing plan.
