What is in Arc C
- The full-scale investigation: what Phase II actually covers — the manufacturing investigation, hypothesis testing, and what "prior occurrence" review really means
- MHRA's Phase III: disposition synthesis, formalized — the genuine, citable third phase Arc A flagged, built out in full
- When Phase II doesn't find a cause: the R&D-escalation practice — the other "Phase III," compared directly against MHRA's version
- The disposition decision — release, reject, or reprocess, and what has to justify each one
This arc goes inside the Phase II investigation Arc A introduced and resolves the two "Phase III" concepts flagged there. Arc D takes up conclusive root cause and CAPA impact assessment directly. Arc E closes with enforcement case studies.
The full-scale investigation: what Phase II actually covers
Arc A introduced Phase II's documentation requirements. This module works through what actually produces that documentation — the investigation itself, not just the paperwork at the end of it.
C1.1 The manufacturing investigation, running alongside the lab review
Requirement A genuine Phase II investigation reaches beyond the laboratory bench and into the manufacturing record for the batch in question: batch production and control records, environmental monitoring data covering the relevant process steps, equipment cleaning and maintenance logs, and the training and qualification records of the personnel involved. This runs concurrently with any remaining laboratory-side review, not after it — a manufacturing cause does not wait for the lab to finish looking at itself first.
C1.2 Hypothesis testing, not a search for the easiest explanation
Requirement MHRA's own OOS/OOT investigation guidance frames Phase II specifically around written and approved hypothesis testing: a specific, documented hypothesis about what caused the result, tested systematically, rather than an informal search that stops at the first plausible-sounding story. This may involve statistical analysis, resampling under the standard Arc B described, retesting by a different analyst, or — for microbiological investigations — organism identification and environmental trend analysis.
C1.3 Prior occurrence: has this happened before?
Requirement One of FDA's specific Phase II documentation requirements, introduced in Arc A, is a review for prior occurrence of the same problem. A failure that has never happened before and one that has recurred across several batches point toward different kinds of causes — and, eventually, different kinds of corrective action. Skipping this check risks treating a systemic, recurring problem as an isolated event.
Arc A listed these specifically, and they are worth holding in view through the rest of this arc: a clear statement of the reason for the investigation; a summary of the manufacturing-process aspects that may have caused the problem; the results of a documentation review, with an assignment of actual or probable cause; the results of a review for prior occurrence; and a description of the corrective actions taken. Everything in this module is how that documentation actually gets produced.
Module C1 — the full-scale investigation
Five questions.
MHRA's Phase III: disposition synthesis, formalized
Arc A flagged that a genuine, citable third phase exists in the public domain, even though FDA's guidance and Barr Laboratories describe only two. This module builds it out in full.
C2.1 What MHRA's Phase III actually reviews
Requirement The UK's Medicines and Healthcare products Regulatory Agency has published its own OOS/OOT investigation guidance describing four stages: Phase Ia and Phase Ib (together covering the ground Arc A called Phase I), Phase II, and Phase III. MHRA's Phase III is triggered by the completion of the laboratory and manufacturing investigations — it reviews their combined results together, rather than opening a new line of inquiry of its own.
C2.2 What Phase III determines
Requirement Per MHRA's guidance, Phase III determines whether other batches or products are affected, identifies the corrective and preventive actions the findings call for, and reaches the batch's disposition decision. It also documents the investigation's conclusions regarding impact on ongoing stability studies and validated processes — results already in hand that a confirmed cause might now call into question.
C2.3 Why this doesn't contradict FDA's two-phase model
Practice MHRA's Phase III is not a rejection of FDA's framework, and it does not conflict with Barr Laboratories. It formalizes, as its own labeled stage, a synthesis-and-disposition step that FDA's guidance already requires — cause assignment, prior-occurrence review, and corrective action — without separating it out under its own phase number. Two regulators describing the same underlying expectation in a different number of stages is not the same as two regulators disagreeing about what an investigation has to accomplish.
MHRA's Phase III is not the same thing as the R&D-escalation practice some firms also call "Phase III," even though the two share a name. Module C3 puts them side by side.
Module C2 — MHRA's Phase III
Six questions.
When Phase II doesn't find a cause: the R&D-escalation practice
Sometimes a genuine, well-run Phase II investigation — laboratory and manufacturing sides both covered — still comes up empty. This module covers what many firms do next.
C3.1 A different kind of escalation
Practice Without a regulatory citation behind it, many firms' internal SOPs define a further stage — also frequently labeled "Phase III" — triggered specifically when a genuine Phase II investigation cannot identify a manufacturing-related root cause. At that point the matter moves to Research and Development for deeper technical characterization: identifying an unknown chromatographic peak through techniques such as LC-MS or NMR, for example, or manufacturing pilot-scale batches under deliberately varied, controlled conditions to reproduce the failure and isolate which variable actually produces it.
Practice This work is real, common, and often slow — structural characterization and pilot-batch studies can take substantially longer than either Phase I or Phase II. It is a genuine extension of the investigation, not a way of abandoning it once the easy questions run out.
C3.2 Two "Phase III" concepts, compared directly
| MHRA's Phase III | Firm-practice "Phase III" | |
|---|---|---|
| Trigger | Completion of the lab and manufacturing investigations | Phase II cannot identify a manufacturing-related root cause |
| Conducted by | Quality unit, synthesizing existing findings | R&D, generating new technical findings |
| Typical activities | Reviewing combined findings; scoping other batches/products; identifying CAPA | Unknown-peak characterization; pilot-scale batch manufacture |
| Regulatory citation | MHRA OOS/OOT investigation guidance | None — a firm-level SOP practice |
| Outcome | Disposition decision and CAPA scope | A technical root cause (or documented confirmation that none exists), feeding into disposition and CAPA |
C3.3 Why the destination matters, not just the cause
Practice An issue that survives R&D characterization without ever turning up a manufacturing-related cause is telling you something specific: the problem may be intrinsic to the drug substance, the formulation, or the specification itself, rather than a one-off process deviation. That distinction matters for what comes next — a process deviation calls for a process fix, but a cause rooted in the material or the specification calls for a different kind of corrective action entirely, a point Arc D returns to directly.
A shared label is not proof of a shared definition. Before assuming a "Phase III" in one document means the same thing as a "Phase III" in another, read what each one actually says — the same caution Arc A raised about the word "atypical."
Module C3 — the R&D-escalation practice
Six questions.
The disposition decision
Every thread in this arc — the manufacturing investigation, MHRA's Phase III, the R&D escalation — ultimately feeds one decision: what happens to the batch.
C4.1 Release, reject, or reprocess
Requirement A completed OOS investigation supports one of three dispositions: release, rejection, or reprocessing. None of the three is automatic. Each has to be justified by what the investigation actually found, documented as part of the same written record 211.192 requires.
C4.2 Hypothesis-testing results and batch release
Practice Module C1 introduced Phase II's reliance on written and approved hypothesis testing. The purpose of that testing is to prove or disprove a specific theory of cause — not to generate the number a disposition decision will ultimately rely on. Industry practice generally does not specify, in the hypothesis-testing protocol itself, that a passing result may be used to release the batch under investigation; the protocol exists to answer the scientific question it was designed to answer, not to substitute for a release decision.
Practice Where a hypothesis is confirmed, or a cause is narrowed down, the result a firm actually relies on for disposition typically comes from a separate, formal confirmatory retest: performed on a fresh sample, in duplicate or triplicate, by the original analyst or a new one, under a protocol with its own pre-defined acceptance criteria — the same kind of pre-established, protocol-driven testing Arc B described. The hypothesis-testing results themselves are not discarded once that confirmatory retest is run. They are retained and included in the investigation package, where they document the reasoning behind the investigation and support regulatory review if the investigation is ever examined.
Hypothesis testing and the confirmatory retest that may follow it answer two different questions. Hypothesis testing asks whether a specific theory of cause is correct. The confirmatory retest asks whether the batch itself meets specification, under criteria fixed before the retest is run — the same standard that makes any retest legitimate rather than testing into compliance, per Arc B. Keeping the two separate, rather than asking hypothesis-testing data to serve as both the investigation and the release justification, is what keeps the investigation package clean and defensible on review.
C4.3 Reprocessing has its own gate
Requirement Where reprocessing is the chosen disposition, 21 CFR 211.115 sets its own separate requirement:
“Written procedures shall be established and followed prescribing a system for reprocessing batches that do not conform to standards or specifications and the steps to be taken to insure that the reprocessed batches will conform with all established standards, specifications, and characteristics.” “Reprocessing shall not be performed without the review and approval of the quality control unit.”21 CFR § 211.115(a)–(b), Reprocessing.
Practice In other words, a disposition decision to reprocess does not close the file — it opens a second, separately governed procedure, with its own quality control unit sign-off, before the batch can move forward.
C4.4 Disposition is not just about the one batch
Requirement Because 211.192 requires the investigation itself to extend to other batches and products that may share the same cause, the disposition decision inherits that same scope. A confirmed cause found in one batch is not filed away as a single-batch problem; it has to be chased into every other batch or product the investigation identifies as potentially affected, with its own disposition decided for each.
A documented disposition decision resolves the batch in front of you. It does not, by itself, prevent the same failure from happening again — that is the CAPA's job. Arc D is where this course finally answers the questions it was built around: what a conclusive root cause actually requires, what a properly scoped impact assessment has to cover, and whether every CAPA needs a monitoring and confirmation component.
Module C4 — the disposition decision
Seven questions.