I support the statement on recommendation for attempt at IDS.  As Ignace mentioned, depending on the timing of randomziation / nature of the trial, patients who do not go on to IDS should be excluded or evaluated as a separate group.