An FDA inspection ends with a conversation, and sometimes with a document. That document tells you how serious the agency considers your findings.
Form 483. Investigators issue this at the close of an inspection to list observations of possible violations. It is a starting point, not a verdict. Your response, and how quickly you act, shapes what happens next.
Warning letter. This is a formal escalation for significant violations. FDA expects a written response within 15 business days. Without adequate corrective action, the agency can block imports, withhold approvals, or pursue enforcement.
“Write-up.” This is an informal term for either document. Both are public, so competitors, partners, and customers can read them.
Cleaning validation keeps showing up in the findings
Cleaning rarely makes headlines, but it fills inspection reports.
In FY2025, FDA recorded 2,837 drug-related citations. Two cleaning provisions made the top 10: 21 CFR 211.67(a) ranked fifth with 95 citations, and 211.67(b) ranked eighth with 70. Together, they account for 165 citations, nearly 6% of the total.
Warning letters tell the same story. An ECA review of fiscal year 2022 found 211.67 cited in 13 warning letters, ranking sixth. Almost all of those deficiencies also pointed to inadequate cleaning validation.
Sterile manufacturers are not exempt. A 2024 analysis of sterile drug observations flagged cleaning validation program gaps, with special attention to evaluating all worst-case elements.
The companies differ, but the failures repeat. The lessons below come from those repeats.
Five lessons from FDA citations
1. Verify results against acceptance criteria before approval
In one warning letter, a cleaning validation report stated that acceptance criteria were met although some values exceeded the limit. FDA cited the firm under 21 CFR 211.67(b). Reviewers should compare every result with its limit before the report is approved.
2. Include all worst-case conditions in the validation
FDA expects firms to identify and evaluate every worst case. Its letters list higher-toxicity and higher-potency drugs, lower solubility in cleaning solvents, hard-to-clean characteristics, the most difficult swab locations, and maximum hold times before cleaning. A 2025 warning letter cited a manufacturer for not conducting worst-case product cleaning validation, which created cross-contamination risk.
3. Document the rationale for sampling locations and sample counts
In one warning letter, FDA noted that a manufacturer had no justification for its swab locations or the number of swabs taken. Firms should record a risk-based rationale for each sampling decision.
4. Follow written procedures and maintain complete records
One 483 cited insufficient documentation for cleaning. A warning letter cited cleaning procedures that were not adequately documented during validation or verification. In another case, FDA found leftover pooled product, with discoloration, in a mixing tank. Records must show that each cleaning followed the approved procedure.
5. Apply corrective actions across the whole cleaning program
One firm responded to a 483 by committing to update its cleaning validation procedures and perform new validations. FDA found the response inadequate. In another letter, FDA found a response inadequate because it lacked detail on the updated cleaning validation. FDA has asked firms to verify and validate cleaning procedures across all products, processes, and equipment.
These findings share a common thread. Each one involves cleaning evidence that was missing, incomplete, or not reviewed before inspection.
How digital validation addresses these findings
Paper-based cleaning validation relies on manual review, transcribed data, and records kept in separate places. Each lesson above traces back to that reliance.
A digital validation tool (DVT) changes how teams create and check cleaning evidence:
- Acceptance criteria. The system compares results with predefined limits as teams enter data. Out-of-limit values are flagged before approval.
- Worst-case selection. Structured product and equipment data supports a documented worst-case rationale linked to the protocol.
- Sampling rationale. Protocol templates require a rationale for each sampling location and sample count.
- Procedure adherence. Guided execution and electronic signatures record who performed each step, and when.
- Program-wide corrective action. One system shows which products, equipment, and protocols a change affects.
Kneat Gx applies these controls across cleaning and other validation disciplines in a single platform. To see how it works for cleaning validation, visit Kneat.
