A regulatory binder is not a filing job. It is a calendar. Nearly everything in it either expires on a date or becomes wrong on a date, and the binder that passes an inspection is the one where somebody knew every one of those dates before the monitor did. This post lists the expiry clocks in a typical site's binder, what the FDA actually cited last year, and a way to stay ahead of both that does not depend on one coordinator's memory.
In short
- FDA BIMO FY2024: 609 clinical investigator inspections, 484 closed with no action, 110 with voluntary action and 15 with official action.
- The top observation categories were 1572 and protocol compliance, inadequate case histories, protocol deviations, investigational product accountability, safety reporting and consent.
- GCP training refreshes at least every three years under the NIH policy most sites adopt; a Michigan physician licence renews every three years; a CLIA certificate every two.
- 21 CFR 312.62(c): records are kept two years past approval of the marketing application, or two years past discontinuation.
What the FDA cited in FY2024
The FDA publishes its Bioresearch Monitoring inspection metrics every year. For fiscal year 2024 it reports 609 clinical investigator inspections with a final classification, of which 484 were No Action Indicated, 110 Voluntary Action Indicated and 15 Official Action Indicated. Fifteen sites in the country got the letter nobody wants. A hundred and ten got a list.
The most common observations, in the FDA's own summary, were failure to comply with Form FDA 1572 requirements and protocol compliance; inadequate or inaccurate case histories and study records; failure to follow the investigational plan; inadequate accountability of the investigational product; safety reporting; and informed consent issues. In the FDA's more granular FY2024 trend analysis, the protocol compliance theme breaks down into things like missed or out-of-window visits, assessments and labs, consent not obtained or re-obtained on time, eligibility criteria not met, and unqualified or undelegated personnel performing study procedures.
Read that list as a binder. A 1572 that still names a sub-investigator who left in March. A delegation log that says the coordinator does vitals when the medical assistant has been doing them since spring. A consent form version 3.0 signed in June when version 4.0 was approved in May. None of these is a bad site. Each is a date that passed while everyone was busy.
The expiry clocks in a regulatory binder
Every site's list is a little different, but these are the ones that run on a calendar rather than on an event.
| Document | What resets it | Typical interval |
|---|---|---|
| Investigator and sub-investigator CVs | Site SOP, sponsor requirement | Every 2 years, signed and dated |
| Medical licence (Michigan) | LARA renewal | Every 3 years |
| GCP training | NIH policy, sponsor requirement | At least every 3 years |
| Protocol-specific training | Each amendment | Per amendment |
| CLIA certificate | CMS renewal | Every 2 years |
| Laboratory normal ranges and accreditation | Lab issues new ranges or certificate | Per issue; check yearly |
| Financial disclosure (Form 3455) | Study start, changes, one year after end | Per event |
| IRB approval | Continuing review | Usually yearly |
| Form FDA 1572 | Any change to sites, labs, sub-investigators | Per change |
| Delegation of authority log | Staff join, leave or change duties | Per change |
Three of those rows carry a Michigan date. A physician licence in Michigan renews on a three-year cycle, with 150 hours of continuing education across the cycle, under the Board of Medicine rules published by LARA. A CLIA certificate, which every site drawing and testing blood needs, is issued by CMS for two years and the renewal notice arrives about six months out. And the GCP interval most sponsors expect, at least every three years, comes from the NIH policy that took effect on 1 January 2017 for NIH-funded trials and has since become the default everywhere.
How to run the binder as a calendar
- List every dated item across every study and every person. Not per study: per person, then per study. Dr Patel's GCP certificate expires once, and it expires for all six studies she is on.
- Write down the date it becomes a problem, not the date it expires. A CV that is 24 months old on the day of a site initiation visit is a finding. Set the reminder at 22 months.
- Escalate. A reminder that goes once to a coordinator's inbox is a reminder that gets buried. Sixty days out to the coordinator; thirty to the coordinator and the regulatory lead; fourteen to the site director.
- Tie the delegation log to the 1572 and the training records. Nobody is delegated a task on a study they are not named on, on a protocol version they have not been trained on. Check the three together every time one of them changes.
- Keep the missing-document list per study, as status. What is missing from this study's file, today. That is a list a coordinator can act on. "The binder is compliant" is a sentence nobody should write down.
- Know the retention date before you archive. 21 CFR 312.62(c) says two years after the marketing application is approved for that indication, or two years after the study is discontinued and FDA is told. The clinical trial agreement often says longer. The longer one wins.
The mistake most sites make at step 1
Building the list per study. Every study binder has its own tab for CVs and licences, so the same physician's licence gets copied into six binders and expires in six places on the same day. The site that tracks per person sees one date. The site that tracks per binder sees six, or, more usually, sees none of them until a monitor does.
The second mistake is trusting the reminder that lives in one person's calendar. The Society for Clinical Research Sites reported in 2023 that patient-facing turnover at sites now runs between 35% and 61% a year. When the regulatory coordinator leaves, the reminders leave with her.
What ICH E6(R3) changes
ICH adopted E6(R3), the revised Good Clinical Practice guideline, on 6 January 2025, and the FDA issued it as final guidance in September 2025. The essential documents list that used to be section 8 now lives in Annex 1, and the guideline leans harder on proportionate, risk-based oversight and on records being fit for purpose rather than exhaustively filed. What it does not do is remove a single expiry date. A licence still lapses. A consent version still supersedes the last one. The calendar is the same; the guideline just expects you to be able to show you were running it.
What we build for this
Our regulatory document tracking for research sites is that calendar, run by a system instead of a memory. It tracks every expiration-dated item across studies and staff, escalates reminders on the schedule above, flags what is missing per study and per person as status, and assembles the packet when a new study or a new staff member is added. The binder, paper or electronic, stays the system of record; the tracker reads dates and reminds people. It never certifies compliance, and it never touches participant data.
Build the per-person list first, this week, in a spreadsheet if that is what you have. If every date on it is more than ninety days out, you are fine for now. If it is not, talk to us.
