In mid-2025, the FDA handed Novo Nordisk’s Bloomington facility a Form 483 after inspectors found cat hair and human hair in vial stoppers, pest infestations in classified areas, and contamination complaints dating back to 2022 that were never properly investigated.

Cat hair. In a sterile fill-finish operation.

That facility was acquired from Catalent in December 2024 as part of a $16.5 billion deal. The contamination issues predated the acquisition. Novo inherited them. And here is the line from the FDA report that should make every quality leader reading this pause:

“Your investigation failed to determine a root cause for the contamination, assess the potential impact to the rest of the lot, or evaluate whether similar issues may have occurred in upstream batches.”

This is not a cat hair problem. This is a quality systems problem. And it is the same quality systems problem showing up in 483s everywhere right now.

If you read enough 483s and warning letters (and I do), the same five patterns keep surfacing. None of them are new. That’s the point.

1. Quality system failures are systemic, not episodic.

The FDA has shifted from flagging individual procedural gaps to citing the quality unit itself. Under 21 CFR 211.22, the quality unit is responsible for approving or rejecting procedures, and the FDA is now explicitly calling out firms where that unit is either absent, under-resourced, or functionally ignored. Quality system issues now account for over 30% of all warning letter citations. The message from the agency is clear: they are no longer interested in whether you have an SOP. They want to know whether your quality system actually governs your operation.

2. Data integrity is still the number one accelerant.

Missing raw data, deleted electronic records, unexplained gaps, altered lab worksheets, uncontrolled system access. These findings show up across microbiology, environmental monitoring, and analytical labs. Between 2022 and 2024, the FDA conducted 114 unannounced inspections in India alone, and 94 of them resulted in 483 observations. Data integrity was the common thread. If your audit trail cannot prove who did what and when, every quality decision built on that data is suspect.

3. CAPA programs are getting cited for being decorative.

Companies are documenting corrective actions but not implementing them. Or implementing them but never verifying effectiveness. The FDA is specifically calling out firms that treat CAPA as a documentation exercise rather than a root cause resolution system. If your CAPA closure rate looks great but your deviation recurrence rate hasn’t moved, an inspector will notice.

4. CMO/CDMO oversight is now a primary enforcement target.

This is the one that should get your attention if you work with contract manufacturers. The FDA cited AACE Pharmaceuticals after their CMO, Brassica Pharma, was found to have fabricated microbiology records and used ripped, dirty gowning. The critical finding: AACE’s quality unit requalified Brassica in 2023 without evaluating whether corrective actions actually addressed the deficiencies identified in prior audits. Supplier qualification that functions as a checkbox exercise is now a direct path to a warning letter.

5. Training keeps appearing in the root cause column.

Inadequate employee training is consistently identified as a top reason for 483 observations. Not because firms don’t have training programs, but because those programs don’t connect to operational performance. Process variability traced to insufficient training. SOPs not followed because personnel were never effectively qualified on them. Deviations recurring because the corrective action was “retrain” without anyone examining whether the original training was adequate in the first place.

The Connection Nobody Talks About

Here is what ties all five of these together: none of them are truly independent findings. They are symptoms of the same underlying failure.

A weak quality system produces weak SOPs. Weak SOPs produce weak training. Weak training produces operators who deviate from process. Deviations produce investigations. Investigations produce CAPAs. And if the CAPA is “retrain the operator” without fixing the system that produced the bad training in the first place, the cycle restarts.

The FDA knows this. That is why they are shifting from citing individual observations to citing systemic quality culture failures. They are not asking “did you train this person?” They are asking “does your training system produce people who can actually execute?”

That’s a fundamentally different question. And most sites cannot answer it.

What to Do With This

If you’re a quality leader or training manager reading this, here are three things worth doing before your next inspection:

Audit your CAPA recurrence. Pull every CAPA closed in the last 12 months that listed “retraining” as a corrective action. Then check whether the same deviation type recurred. If it did, your CAPA system is generating paperwork, not solutions.

Pressure-test your CMO oversight. When was the last time you audited your contract manufacturers against actual performance data, not just their quality manual? If your supplier requalification process does not include a review of corrective action effectiveness from prior findings, you have the same gap AACE had.

Connect your training metrics to your quality metrics. Deviation trending by training status. Time-to-competency correlated with error rates. CAPA recurrence mapped to training gaps. If these data streams live in separate systems with separate owners, you are blind to the most common root cause in your operation.

That’s it for Issue #1. If this was useful, forward it to someone on your quality team. If you want to talk about any of this in more detail, I’m at bdrapeau@gxpframe.com.

Brian Drapeau
Founder, GxP Frame
gxpframe.com

See you next week!