How do N-of-1 trials account for long term variations? Like maybe a treatment have some unknown long term side-effects that "contaminate" the effects observed during later rounds of placebo. We could take average over large population because it's reasonable to assume that one person is unrelated to another but it's probably not reasonable to assume later treatment rounds are independent from earlier rounds.
@TheRealBuilder