This is a ready-to-use checklist to run on an FDA Form 483 or Warning Letter response before it is submitted. It is the quality gate that catches the recurring reasons responses fail: a restated-observation root cause, a training-only CAPA, no scope assessment, no effectiveness check, or a data-integrity finding treated as a documentation slip. Work each item, mark Pass / Fail / NA, and do not submit while any item is Fail. Replace every <<FILL: ...>> placeholder. This content is educational and general, not legal advice; use it with your own regulatory and legal review.
Header
| Field | Entry |
|---|---|
| Inspection / 483 reference | <<FILL>> |
| 483 issue date | <<FILL>> |
| 15-business-day response date | <<FILL>> |
| Response type | Full / Interim |
| Reviewer | <<FILL>> |
| Review date | <<FILL>> |
Section A: Timing and mechanics
| # | Item | Pass / Fail / NA | Notes |
|---|---|---|---|
| A1 | The response (or an interim response) will be submitted within 15 business days of the 483 issue date | ||
| A2 | If interim, it acknowledges every observation, states immediate containment, and commits to dated completion of the rest | ||
| A3 | The response is addressed to the correct office and named contact, through the specified channel | ||
| A4 | A dated, complete archive copy will be retained in the master response file |
Section B: Structure and per-observation completeness
| # | Item (repeat per observation) | Pass / Fail / NA | Notes |
|---|---|---|---|
| B1 | Every numbered observation is addressed separately, in FDA’s order, using the agency’s observation text as the heading | ||
| B2 | Each observation response opens by acknowledging the finding, not by arguing it | ||
| B3 | Each observation has a root cause, corrective action, scope/preventive assessment, effectiveness check, and evidence | ||
| B4 | Attachments are numbered and referenced inline; one observation’s evidence is not buried in another’s |
Section C: Root cause depth
| # | Item (per observation) | Pass / Fail / NA | Notes |
|---|---|---|---|
| C1 | The root cause names a system, procedure, or design element, not an individual’s care (“lack of training”/“human error” are not the stopping point) | ||
| C2 | The root cause is not merely a restatement of the observation | ||
| C3 | A structured method was used to real depth (several “whys” or a fishbone), and alternatives considered were ruled out with evidence | ||
| C4 | A second reader could reach the same cause from the evidence provided |
Section D: Corrective and preventive action
| # | Item (per observation) | Pass / Fail / NA | Notes |
|---|---|---|---|
| D1 | Each action is specific, with a named role owner and a specific calendar date (not “within X months”) | ||
| D2 | The CAPA is not training-only; there is a genuine action that removes the cause, not just retraining | ||
| D3 | Correction, corrective action, and preventive action are distinguished; at least one preventive action addresses the mechanism | ||
| D4 | Scope: the response evidences whether the same cause exists in other systems, processes, or sites | ||
| D5 | Each observation has an effectiveness check with a date and acceptance criteria | ||
| D6 | Timelines are realistic; no aggressive date is committed that is expected to slip; long/system-replacement work has credible interim controls |
Section E: Data integrity findings (if any)
| # | Item | Pass / Fail / NA | Notes |
|---|---|---|---|
| E1 | The finding is characterized as a systemic control gap or intentional falsification, with a documented conclusion | ||
| E2 | The breach is scoped: a retrospective review assessed whether it extends beyond the systems and period the investigator identified | ||
| E3 | Reliability of the affected data is addressed, especially where it supports a submission or released product | ||
| E4 | Legal counsel has reviewed; for serious cases a qualified third-party assessment is proposed rather than awaited |
Section F: Commitment reconciliation and ownership
| # | Item | Pass / Fail / NA | Notes |
|---|---|---|---|
| F1 | Every dated commitment in the response is on the commitment register with owner, date, status, and evidence record | ||
| F2 | Every commitment has a matching CAPA or change-control record in the quality system with the same date | ||
| F3 | The response reads as the company’s own voice; the people who do the work can explain in their own words what changed and why | ||
| F4 | The quality unit will sign; the response is not signed by regulatory affairs alone |
Overall disposition
| Field | Entry |
|---|---|
| Any item Fail? | Yes / No |
| Disposition | Ready to submit / Return for rework |
| Reviewer signoff (name, date) | <<FILL>> |
| Quality signatory signoff (name, date) | <<FILL>> |
Do not submit while any item is Fail. Record the rework and re-review before disposition.
References
Federal Food, Drug, and Cosmetic Act, section 704. 21 CFR 211.22, 211.192. FDA draft guidance, Responding to FDA Form 483 Observations at the Conclusion of a Drug CGMP Inspection (draft, March 2026; confirm status). FDA guidance, Data Integrity and Compliance With Drug CGMP: Questions and Answers (2018).
Filled specimen
The following shows selected items completed for one observation, so you can see the level of scrutiny expected. Illustrative content; replace with your own.
| # | Item | Result | Notes |
|---|---|---|---|
| C1 | Root cause names a system, not a person | Pass | Cause is a change-control gap: testing sequence changed without updating the disposition SOP |
| C2 | Not a restatement | Pass | Does not say “review was not adequate”; names why the hold point was missing |
| D2 | Not training-only | Fail (initial) | Draft said “retrain the reviewer”; reworked to close the change-control gap and add the missing hold point |
| D4 | Scope assessed | Pass | All disposition SOPs checked for the same missing-dependency pattern |
| D5 | Effectiveness check present | Pass | Sample 30 dispositions at 90 days; zero released before all dependent results final |
| F2 | Matched in quality system | Pass | CC-1051 opened with the same 2026-06-15 date |
In this example the checklist caught a training-only CAPA (D2) before submission, sent it back, and the reworked response fixed the actual change-control gap. That single catch is the difference between a response that closes and one that escalates.
Common inspection findings this checklist prevents
- A restated-observation root cause reaching the agency.
- A training-only CAPA with no systemic fix.
- No scope assessment, so the next inspection finds the same gap elsewhere.
- Commitments in the letter with no matching quality-system record.
- A data-integrity finding answered without the systemic-versus-intentional conclusion.
How to adapt this checklist
- Repeat Sections B, C, and D per observation; a three-observation 483 runs those sections three times.
- Point the references at the current guidance status before use.
- Add house-specific items your quality unit requires (for example a specific legal-review signoff for submission-affecting findings).
- Keep the completed checklist as a record in the master response file.