What is in Arc H
- Content uniformity and weight variation — USP ⟨905⟩ and the Acceptance Value, worked as two real batches
- Particle size methods — laser diffraction, sieve analysis, microscopy, and why particle size drives both dissolution and uniformity
- Case study — Cadila Healthcare and two Glenmark Pharmaceuticals letters
Each module ends with a knowledge check. A cumulative assessment covering Arcs A through H is issued separately.
Content uniformity and weight variation
Two tablets from the same batch, both weighed or assayed individually rather than pooled — because a batch's average can look perfect while individual units vary wildly around it.
Requirement USP General Chapter ⟨905⟩, “Uniformity of Dosage Units,” defines two methods. Content Uniformity directly assays individual dosage units against label claim, and is always permitted, for any dosage form. Weight Variation estimates content from unit weight instead of a direct assay, and is permitted only under specific conditions — per USP's own public FAQ, generally uncoated or film-coated tablets containing at least 25 mg of drug substance that also makes up at least 25% of the tablet's weight, with analogous conditions for hard capsules and certain weight-filled solutions or suspensions. A firm runs one method or the other, not both, and the choice is not arbitrary — it depends on whether the product actually meets Weight Variation's narrower eligibility conditions.
H1.1 The Acceptance Value, worked as two real batches
Requirement Uniformity is judged using the Acceptance Value (AV), which combines how far the sample mean sits from a reference value with how spread out the individual results are:
AV = |M − X̄| + k·s
where X̄ is the sample mean (% of label claim), s is the sample standard deviation, k is an acceptability constant (2.4 for the 10-unit Stage 1 sample), and M is a reference value: the sample mean itself if it falls inside a 98.5–101.5% “indifference zone,” or the nearer edge of that zone (98.5% or 101.5%) if the mean falls outside it. Requirement Stage 1 requires AV ≤ 15.0 across 10 units; if Stage 1 doesn't pass, Stage 2 tests 20 more units (30 total) against the same AV ≤ 15.0 limit, with individual units additionally bounded to 75–125% of M. These figures are drawn directly from USP's own public FAQ page.
Practice This course does not reproduce paywalled USP-NF chapter text. The formula, the k=2.4 constant, and the 15.0 acceptance limit above are corroborated directly from USP's own public FAQ and an independently checked secondary source, and are used here with that level of confidence. Some secondary summaries describe Stage 2 nuances slightly differently depending on chapter revision or harmonization status — confirm the precise current wording against a live USP-NF subscription before relying on it in a real validation protocol.
Two ten-unit batches, worked with the same formula:
Batch 1's individual assay results cluster tightly around label claim (mean 100.3%, s = 1.22), giving AV = 2.92 — comfortably under the 15.0 limit. Batch 2, same ten-unit sample size and the identical formula, scatters from about 85% to 122% of label claim (mean 101.9%, s = 13.31), giving AV = 32.29 — more than double the limit.
Practice Notice what the formula actually rewards and penalizes: it isn't simply a "how close is the average to 100%" check. A batch could have a mean sitting exactly on label claim and still fail on the k·s term alone if individual units are wildly inconsistent, and a batch with a mean slightly off-center can still pass if its spread is tight enough. Both bias and variability are captured in one number, which is exactly why AV, not a raw mean, is the pass/fail criterion.
Module H1 — content uniformity and weight variation
Six questions.
Particle size methods
Particle size sits upstream of two very different-looking failures — a dissolution OOS and a content uniformity OOS — which is exactly why it shares a module with Module H1 rather than standing alone.
H2.1 Three ways to characterize a particle population
Requirement Laser diffraction (USP ⟨429⟩) is the workhorse for particle size distribution — inverting a laser diffraction pattern, via Mie or Fraunhofer theory, into a volume-based distribution, typically reported as D10/D50/D90. It's fast and statistically robust, sampling many thousands of particles per run, and is harmonized with ISO 13320.
Requirement Sieve analysis (USP ⟨786⟩) mechanically separates material through graduated mesh sieves, used for coarser material (roughly above 75 µm) where diffraction is less suitable, and is harmonized internationally with the European Pharmacopoeia. Microscopy provides direct visualization and sizing — useful for morphology, and as a cross-check when diffraction results look anomalous, such as for needle-shaped or agglomerated particles that violate diffraction's underlying spherical-equivalence assumption.
H2.2 Why particle size drives both dissolution and content uniformity
Practice For poorly soluble (BCS Class II/IV) drugs, dissolution is often surface-area-limited — smaller particles mean more total surface area for a given mass, and faster dissolution. An unplanned particle-size shift (a milling process drifting, a polymorphic change, agglomeration during storage) is a classic root cause of a dissolution OOS result, a connection Module H3's case study returns to directly.
Practice For low-dose, high-potency drugs, the same physical property threatens a different test: if API particle size is too large, too few particles end up in each individual dosage unit. Simple sampling statistics then means unit-to-unit content variability increases — directly threatening the Module H1 Acceptance Value. One upstream physical property, two very different downstream test failures.
Module H2 — particle size methods
Six questions.
Case study
An investigation that never found a root cause, and a firm cited twice at different facilities for variations on the same underlying gap: an OOS explained once, without checking whether the explanation applied anywhere else.
H3.1 Cadila Healthcare Limited — a narrow-therapeutic-index drug, no root cause
“…the following OOS investigation reports associated with potency and content uniformity specifications for warfarin sodium, a narrow therapeutic index drug, failed to identify a root cause.”Cadila Healthcare Limited — Warning Letter, December 23, 2015. fda.gov
Practice Warfarin sodium's narrow therapeutic index raises the stakes on exactly the Module H1 concern: a content uniformity failure for a drug where the difference between a therapeutic and a harmful dose is small is not an abstract statistical concern. An OOS investigation that closes without a root cause leaves that risk unaddressed for every subsequent batch made the same way.
H3.2 Glenmark Pharmaceuticals Limited — two facilities, two related findings
“Although you attributed the content uniformity failure to the lack of defined compression parameters for desmopressin acetate 0.1mg batch 20210776, you failed to test other batches or drug products that used the same (b)(4) process and compression equipment.”Glenmark Pharmaceuticals Limited — Warning Letter, November 22, 2022. fda.gov (Colvale, Goa facility; 21 CFR 211.192)
Fourteen batches of desmopressin acetate had been rejected for OOS content uniformity between 2018 and 2021 without the investigation ever being extended to check whether the same compression-parameter gap affected other products on the same equipment. Practice This is the exact failure mode Arc D4's Stason case study named for dissolution, now showing up for content uniformity: an investigation that correctly identifies a proximate cause for one batch, but never asks the obvious next question — does this same cause affect anything else made the same way?
Cites a dissolution OOS investigation for potassium chloride extended-release capsules attributed to API particle size distribution, where FDA found the root-cause analysis and supporting morphology evidence inadequate.Glenmark Pharmaceuticals Limited — Warning Letter, July 11, 2025. fda.gov (Pithampur, Indore facility; a different facility from the 2022 letter above, which this same 2025 letter cross-references as a recurring pattern at the firm)
Practice This second Glenmark letter is Module H2's connection made concrete and, at the same time, a caution about it: the firm attributed a dissolution OOS to API particle size distribution — the exact causal link this arc's H2.2 describes — but FDA found the supporting root-cause analysis and morphology evidence inadequate. Naming the right physical mechanism is not the same as demonstrating it actually happened in this batch, with real data. Both steps are required.
Module H3 — case study
Six questions.