I think a pragmatic, risk-based approach is appropriate. I would suggest that a review of the files by the monitor is still appropropriate but use judgement in terms of what might be expected for studies that are very old.

It is important to remember though that although the clinical trial directive 2001/83/EC is relatively recent, ICH GCP itself has been in existence with an expectation that it is followed since 1996 (18 years). Going back even further, the basic principles of good clinical practice - including the maintenance of appropriate records - was also documented in the CPMP Note for Guidance 111/3976/88 wich was published in 1988 and adopted in 1990. So the requirement to maintain adequate trial records has been a specific regulatory requirement for at least 24 years. Trials would have to be VERY long-running to pre-date these requirements but I'm guessing there may be some that fit into this category.