APR / PQRAnnual Product Review / Product Quality Review

21 CFR 211.180(e) requires every drug product manufacturer in the US to evaluate, at least annually, the quality standards of each drug product to determine the need for changes in specifications, manufacturing or control procedures. EU GMP Chapter 1 §1.10 requires the same review — branded as the Product Quality Review (PQR) — covering an explicit list of inputs (starting materials, in-process controls, finished-product results, deviations, OOS, changes, complaints, recalls, returns, stability, contractual-arrangement review). The two regimes are operationally equivalent; companies that ship globally write one document that satisfies both.

• One review per product (the same active in different presentations / strengths may be grouped if the process is essentially the same).
• Period is normally the calendar year or the product anniversary; whichever is chosen must be defined in a written SOP and held constant year-over-year.
• Owned by Quality (the QA head signs). Production, supply chain, regulatory and the QP all contribute — but QA owns the conclusion and any CAPA that flows from it.
• Must be completed within 60-90 days of the period close — MHRA inspectors call out 'PQR not yet completed' as a common Major finding when the review sits open six months after period end.

1. Review of starting materials and packaging materials (especially from new sources) — incoming results, supplier-quality events and CoA-vs-actual deltas.
2. Review of critical in-process controls and finished-product results — every batch, all critical attributes, trended against spec and against the validated mean.
3. Review of all batches that failed to meet established specifications, plus the investigations.
4. Review of all significant deviations, non-conformities, their investigations and the CAPA effectiveness.
5. Review of all approved changes (formulation, process, equipment, supplier, analytical method) and their post-implementation performance.
6. Review of dossier variations / regulatory commitments submitted, granted or refused in the period — including those for export markets.
7. Review of the stability programme — every time-point, any adverse trend, OOS results.
8. Review of all quality-related returns, complaints, recalls and the investigations.
9. Review of adequacy of any previous corrective actions related to the product process or equipment.
10. Review of post-marketing commitments for new marketing authorisations and variations.
11. Review of qualification status of relevant equipment and utilities (e.g. HVAC, water, gases).
12. Review of contractual arrangements (Chapter 7) to ensure they are up to date.

21 CFR 211.180(e) is shorter on paper — three bullets — but FDA inspectors expect the same depth via the underlying §211.192 (production-record review) and §211.166 (stability) requirements. In practice the EU §1.10 list is the global minimum.

A list of batches is not a trend. The APR/PQR must trend each critical attribute over the period — typically as a run chart with the validated mean, ±3σ control limits, the registered specification and the OOT signals flagged. Common trended attributes: assay, content uniformity, dissolution time-points, impurity profile, particulate counts for sterile products, microbial limits, hardness / friability for tablets, fill weight for capsules / vials. The trends are compared to the prior year's APR/PQR — Stage-3 CPV trending feeds this directly when the process is run-rate enough.

Every deviation, OOS, OOT, complaint and recall opened in the period is listed with the root cause, the CAPA and the CAPA effectiveness verdict. The reviewer looks for repeat causes — three deviations all citing 'operator error' in different wording is a training-system finding. Three OOS results on the same impurity peak is a method-or-process finding. Inspectors compare the APR's repeat-cause analysis to the actual deviation register; mismatch is a Major.

Every approved change to the formulation, process, equipment, supplier or analytical method is listed with the change-control number, the regulatory classification (variation / notification / no-impact), the implementation date and the post-implementation performance. The reviewer answers 'did the change deliver the expected benefit, and did the process remain in control after it?' — this is where comparability data, Stage-3 CPV trending and complaint trends feed in. A change implemented in the period whose post-implementation performance is not yet assessed is named as such and carried to the next year's review.

• Is the process in a state of control? (Yes / No / Trending — with the supporting trends cited.)
• Are the registered specifications still appropriate? (If you've trended for 12 months ±0.3σ inside a ±10% spec, the spec is loose — the review is the place to flag a tightening.)
• Are the validated ranges still appropriate? (If Stage-3 CPV shows the process drifting toward the edge of the design space, the review proposes a re-PPQ or a design-space update.)
• Are there CAPAs from prior reviews still open or ineffective?
• What new CAPAs does this review open?

The conclusion is signed by QA. Any CAPA opened by the review enters the CAPA system with the APR/PQR as the trigger record — it is not a parallel side-channel.

1. APR not completed within the SOP-defined window (most-cited).
2. Review is a list, not a trend — no run charts, no statistical conclusion.
3. Repeat deviation causes not identified across the period.
4. Post-change performance not assessed for changes approved in the period.
5. Stability data summarised but adverse trends not investigated.
6. Contract-manufacturer batches treated separately from in-house batches instead of being rolled into the same product review.
7. Conclusion section omitted or signed without supporting data.
8. Prior-year CAPAs from the APR not tracked to closure.

• Per-product APR workspace auto-pulls every batch in the period from the BMR, every release decision from QA, every deviation / OOS / OOT / complaint / recall from the quality systems, every approved change from change control, every stability time-point from the stability programme — no spreadsheets, no CSV exports.
• Trending dashboards render run charts with validated mean, ±3σ control limits, registered specification and OOT flags per critical attribute — the same charts that drive Stage-3 CPV, so the APR and the CPV agree by construction.
• Repeat-cause analysis groups deviations and OOS by root cause across the period and flags clusters — the reviewer no longer eyeballs a 200-row register looking for 'operator error' six times.
• Post-change performance section auto-links each change-control number to the batches that ran under the change, so 'did the change hold?' is a query, not a fishing expedition.
• Sign-off uses the two-person e-signature contract per Part 11 §11.200 — the report is hashed and archived as an immutable PDF/A. CAPAs opened by the review are linked back to the report's section that triggered them.
• Multi-site APRs: when the same product runs at two or three plants, V5 produces one consolidated review covering all sites, with per-site sections and a cross-site trend — the layout EU §1.10 expects.