This question about someone who wants to be put into an isolation chamber got me thinking about the ethics involved in observing someone doing something risky. If a participant approaches you about doing something dangerous, possibly with a high likelihood of death or permanent injury, and wants you to document it as an N-of-1 case study, are you ethically allowed to? For example, a terminally ill patient wants to expose themselves to high doses of a drug that may cure them, but will likely severely impair their cognitive function, lead and wants you to document their cognitive decline. How does one describe this to an IRB? Is it the same as studies of actively using addicts?
Institutional Review Boards (IRBS) are both legal and practical gatekeepers for researchers, https://en.wikipedia.org/wiki/Institutional_review_board. Practically, a social scientist may not be able to publish or receive funding support if their research is not approved. Legally, they also provide some cover. However, they are not guarantors of being ethical, because there is not universal agreement about what is ethical. So perhaps we could ask WHO would consider it more ethical if such a study were approved by an IRB compared to not?
From my experience both inside and outside of hospitals managing IRB submissions, such a study is unlikely to be approved. However, some might still consider it ethical, and there's no practical barrier to observing it. One important feature here is whether the researcher designed, encouraged, or administers the treatment. Researchers are not prohibited from observing harmful behaviors in others, in general.