Give survivorship interventions a shared control arm

The structural weakness of this literature is not that it is wrong but that it is small and fragmented. Single-centre trials of forty to a hundred people, each with its own comparator and its own questionnaire, are what a field without a sponsor produces. They are rarely large enough to change practice and almost never comparable with one another.

Platform trials solved an equivalent problem in treatment: a shared protocol, a shared control arm, prespecified rules for adding and dropping interventions, and a common endpoint set. Applied to survivorship, the obvious domains are fatigue, physical function, cognition and return to work, each with several candidate interventions and none with a dominant one. The efficiency gain comes from the shared control: a trial comparing four interventions against one control arm uses roughly half the participants of four separate two-arm trials.

This only works if the core outcome set work is done first or alongside, which is why the two proposals are linked. It also needs a funder willing to commit to a decade of infrastructure rather than a sequence of three-year grants, which is the part that has not happened.