Out-of-Specification (OOS) Investigation for Pharmaceutical Quality Control

E1.1  A failing result, invalidated on resampling alone

“Your firm failed to adequately investigate failing microbiological results in finished products. For example, your (b)(4) lot (b)(4) obtained a failing result...You invalidated this failing result based on results of resampling and released the lot.”FDA, Warning Letter 721916 to Medical Products Laboratories, Inc., April 9, 2026. fda.gov

Requirement This is Module B2's invalidation standard, violated in the plainest possible way. FDA's own guidance requires observation and documentation of a specific test event before a result may be invalidated — a passing resample, by itself, documents nothing about what caused the original failing result. The firm invalidated first and, on this record, never supplied the assignable cause the standard actually requires.

E1.2  The same pattern, twice, on stability data

“In two separate incidents, you invalidated OOS stability results for subpotent assay without adequate scientific justification. You relied on resampling data from (b)(4), lots (b)(4), to invalidate the initial stability OOS result with no identified laboratory error or rationale.”FDA, Warning Letter 721916 to Medical Products Laboratories, Inc., April 9, 2026.

Requirement FDA cited the microbiological finding under 21 CFR 211.192 and the stability finding under 21 CFR 211.166(a) — two separate regulations, one repeated failure pattern. Twice was enough for FDA to treat this as systemic rather than a single lapse.

E1.3  The remedy

Requirement FDA required a retrospective independent review of all invalidated failing/OOS results for in-market, within-expiry U.S. products — not just the specific lots the inspection happened to examine. Every invalidation decision the firm had made, across its entire in-market portfolio, was now subject to review.

---

E2.1  Nine chances, no conclusive cause

“...your investigations failed to adequately identify the root causes of contamination in your drug products and the source of B. cereus contamination within your manufacturing environment.”FDA, Warning Letter 724347 to Medline Inc., May 28, 2026. fda.gov

Requirement Nine separate, documented investigations over more than two years, and the same organism kept appearing. This is Module D1's warning made concrete: a "probable" explanation, offered and apparently accepted nine separate times, is not the same thing as an actual, conclusive answer.

E2.2  What FDA said inadequate investigation actually costs

“Inadequate investigations can lead to unidentified root causes, ineffective CAPA, and recurring problems that may pose a patient safety hazard.”FDA, Warning Letter 724347 to Medline Inc., May 28, 2026.

Requirement FDA also noted that finished product microbiological testing cannot be relied upon as sole justification to release drug product batches, since contamination is not uniformly distributed — a batch can pass finished-product testing purely by chance, even while the underlying contamination source remains active and unaddressed.

E2.3  The remedy

Requirement FDA required an independent retrospective review of all critical investigations for the last three years, together with an independent assessment and remediation plan for the firm's CAPA program as a whole — a program-level fix, not just a correction to the nine investigations already on record.

---

E3.1  Deviations cancelled, not investigated

“...your firm cancelled numerous deviations without investigating the root cause or assessing product impact.”FDA/CBER, Warning Letter 728681 to Genzyme Ireland Limited, June 22, 2026. fda.gov

Requirement Cancelling a deviation is itself a disposition decision, and 21 CFR 211.192's investigation duty applies to it the same way it applies to a formal OOS result. A retrospective review found that 36 of 74 cancelled deviations actually required investigation that had never been performed — not a rare exception, but close to half the sample reviewed.

E3.2  An unsupported "no product impact"

Requirement For one specific deviation, QE-1679805, FDA found the firm's "no product impact" conclusion "not adequately supported due to discrepancies," requiring a revised investigation addressing root causes. This is Module D2's warning in its most literal form: the conclusion itself may not have been wrong, but it was asserted rather than earned, and FDA's review caught the difference.

E3.3  A compounding records problem

Requirement FDA separately cited incomplete laboratory records under 21 CFR 211.194(a): multiple non-viable particulate excursions and repeated test failures that were not documented for review. Without that documentation, no one — not the firm's own quality unit, and not FDA on inspection — could actually check the reasoning behind the firm's conclusions after the fact.

---

E4.1  Three letters, three pillars, one remedy pattern

Practice Every remedy in this table looks backward across a defined period, not just forward from the date of the letter. None of the three firms was allowed to treat the specific instance FDA happened to find as the whole problem — each was required to go looking for how far the same failure pattern actually reached.

E4.2  211.192's scope requirement, enforced rather than voluntary

Requirement Arc A introduced 211.192's requirement that an investigation extend to other batches and products sharing the same cause. All three remedies in this table are that same requirement, applied after the fact: when a firm doesn't do this scoping voluntarily and promptly, FDA's remedy is to require it anyway — typically across a wider window than the firm would have needed to cover if it had done the work at the time.