May an investigator retain and analyze already collected data about a subject who withdraws from the research or whose participation is terminated by the investigator?
OHRP interprets the HHS regulations at 45 CFR part 46 as allowing investigators to retain and analyze already collected data relating to any subject who chooses to withdraw from a research study or whose participation is terminated by an investigator without regard to the subject's consent, provided such analysis falls within the scope of the analysis described in the IRB-approved protocol. This is the case even if that data includes identifiable private information about the subject.
Of course, in some cases, researchers may still choose to discard such data:
For research not subject to regulation and review by FDA, investigators, in consultation with the funding agency, certainly can choose to honor a research subject's request that the investigator destroy the subject's data or that the investigator exclude the subject's data from any analysis. Nothing in this document is intended to discourage such a practice. For example, an investigator studying social networks in a community may agree to omit all of the data they have collected from a subject of the study at the request of that subject.
(The clause about the FDA is due to the FDA's concern that withdrawals can skew the findings of clinical trials.)
This guidance strikes me as helpful. When reading sample consent forms, e.g., Cornell's, I am often left wondering what is meant by the boilerplate, "you are free to withdraw at any time," especially when it comes to interviews. If a researcher does a great interview and writes a dissertation chapter around it, can the narrator show up at the dissertation defense and pull the information? (This is apparently the case with undergraduate research at Bard College.)
Fortunately, OHRP says no.