PID in Investigator Site File

Hi all,
I still do not feel it is clear what identifiable data is permitted to be held within the ISF, and what must be retained separately.
For example, must CRF data (containing subject ID) be retained separately from the subject list (identity – subject ID)? And if this is the case, what are the separation requirements? Can, for example, ID lists simply be retained in an envelope, and clearly marked as identifiable data, and then stored alongside CRF data within an ISF?
The same question applies for source data. Where source data is first recorded for the purposes of a research study (ie, captured in a source data sheet), and where the source data is not relevant to the clinical care of the participant (& therefore need not retained in the medical records), can the source data sheet be held in the ISF? Does it matter if a source data sheet contains PID?
It was my understanding that as this is being retained by the investigator site under secure conditions (eg, access controls in place, etc), then it would not matter if identifiable data was retained within the ISF.
I note that GCP R2 states ‘[I]Documents containing direct identifiers of trial subjects, such as the subject medical files, the identification code list and informed consent forms, should be [B]maintained separately by[/B] and under the control of [B]the investigator/institution’[/B], [/I]yet it is not clear exactly what this entails. This statement has been removed from GCP R3 - there is no mention of separation of documents containing identifying information.
Grateful for any advice or guidance.

Your understanding is still right; it doesn't really matter. The principle to consider is confidentiality, which practically means "the sponsor shouldn't hold, or have uncontrolled access to, identified participant data."

[QUOTE=Research_Gov;6925]
I note that GCP R2 states ‘[I]Documents containing direct identifiers of trial subjects, such as the subject medical files, the identification code list and informed consent forms, should be [B]maintained separately by[/B] and under the control of [B]the investigator/institution’[/B][/I][/QUOTE]

Regarding the above quote -- I was unable to locate this within GCP R2. It seems that it's actually from an [URL="https://www.ema.europa.eu/en/documents/scientific-guideline/guideline-content-management-and-archiving-clinical-trial-master-file-paper-andor-electronic_en.pdf"]EMA guidance document[/URL] (4.1.3), and reading this in context it's actually just the same principle as above; don't put identified information into sponsor repositories.

----

As an aside, note that "ISF" (or, on the sponsor side, TMF) is not so much considered to refer to a singular location, rather as the collective [investigator] records for the trial. Identified records are one example of this (as per the EMA guidance quoted, even if kept separately they are a part of the ISF) .

A different example: your equipment maintenance records, even if filed in a central location, are part of the trial records and are therefore part of the ISF. This is much more practical than replicating the same records in trial-specific binders for multiple trials, but still compliant. Hope this helps with understanding.