Written and maintained by CASRAI Editorial Board
Last updated
An Annual Product Quality Review (APQR) is the yearly, written evaluation required by 21 CFR 211.180(e) for every drug product manufactured for the U.S. market — and it is not the same document as a data dump of the year’s testing results. The regulation requires that a representative sample of batches, whether approved or rejected, and the complaints, recalls, returns, and investigations tied to that product all be pulled together and evaluated to determine whether specifications or manufacturing/control procedures need to change. A binder of charts with no conclusion attached does not satisfy that requirement, no matter how complete the underlying data is.
This guide covers what an APQR actually has to draw on, how the U.S. requirement differs in name and scope from the EU’s Product Quality Review (PQR), and how to structure the trending and conclusions sections that turn a compliance obligation into a document an inspector reads with confidence rather than skepticism.
APQR vs. PQR: same idea, two regulatory names
“APQR” and “PQR” get used almost interchangeably in industry conversation, but they trace to two different rules, and a site that supplies both U.S. and EU markets needs to satisfy both:
- United States — Annual Product Review, under 21 CFR 211.180(e). The regulation’s own text requires that written records “be maintained so that data therein can be used for evaluating, at least annually, the quality standards of each drug product to determine the need for changes in drug product specifications or manufacturing or control procedures.” It then requires written procedures covering, at minimum, a review of a representative number of batches (approved or rejected) and a review of complaints, recalls, returned or salvaged product, and investigations conducted under 21 CFR 211.192 for that product. “APQR” is industry shorthand — the regulation itself doesn’t use that acronym — but it’s the term almost every U.S. quality system uses for this review.
- European Union — Product Quality Review (PQR), under EU GMP Chapter 1 (Pharmaceutical Quality System), section 1.10, EudraLex Volume 4. The EU version is framed more broadly than 211.180(e): it’s explicitly a periodic or rolling quality review conducted for all authorised products, including export-only products, with the stated purpose of verifying the consistency of the existing process, the appropriateness of current specifications for both starting materials and finished product, and highlighting any trends. It’s conducted with defined manufacturer responsibility and, where the manufacturer and marketing authorisation holder are different parties, coordination between them — a distinction that doesn’t exist in the FDA framework, which addresses the manufacturer directly.
In practice, a single combined APQR/PQR document that satisfies both requirements is the norm for globally-registered products — building two separate reviews from the same underlying data is rarely worth the duplication. The data inputs below are written to satisfy 211.180(e) at minimum; where EU distribution applies, the same document typically needs to add sections on starting/packaging material sourcing and process consistency to close the PQR gap.
The data-source inputs an APQR has to draw on
211.180(e) names two review categories explicitly — batch records and the complaints/recalls/returns/investigations set — but a defensible APQR in practice assembles data from five source systems. Treat each as a distinct pull, not a single export from one system, because they typically live in different places (LIMS, the deviation/CAPA system, the complaint-handling system, and the document/change-control system) and get merged into the review manually or via a validated data aggregation tool.
1. Batch records
A representative sample of batches manufactured in the review period — both those that were released and, notably, any that were rejected. Pulling only released batches is a common shortcut that misses exactly the signal an APQR exists to catch: a batch that failed doesn’t get to disappear from the annual picture just because it never shipped. For each batch reviewed, the relevant data is yield, in-process control results, and any batch-specific deviations already on file. See Electronic Batch Record (EBR) for what a batch record itself has to contain.
2. OOS results and deviations
Every out-of-specification result and every deviation logged against the product in the review period, not just the ones that were ultimately confirmed as investigation-worthy. The APQR’s job is to look at the aggregate pattern — is a particular test, stage, or piece of equipment generating a disproportionate share of the deviations? — which is a different question than each individual deviation investigation already answered on its own. See OOS Investigation: The FDA Two-Phase Process and Deviation Management in a Regulated Lab for how each individual event gets classified and investigated before it ever reaches the annual roll-up.
3. Stability data
211.180(e) doesn’t name stability data explicitly, but ongoing stability results are standard APQR content in practice, and they’re the direct link to ICH Q1 stability testing commitments and to FDA’s process-validation Stage 3 concept — Continued Process Verification (CPV), the ongoing statistical trending of process and product data that keeps a validated process in a demonstrated state of control. Stability data belongs in the APQR because a shelf-life or storage-condition issue emerging mid-year is exactly the kind of signal the annual review exists to surface before it becomes a field complaint.
4. Complaints, recalls, returns, and salvaged product
Named directly in 211.180(e). This pull comes from the complaint-handling system rather than the quality/LIMS side of the operation, which is why it’s easy to under-scope an APQR by treating it as a lab-data exercise only. Complaint volume and category by product, any recalls issued, and any product returned or salvaged all need to be reviewed and trended, not just logged as a count.
5. Change control actioned against the product
Not named explicitly in 211.180(e)’s text, but essential context: any specification, process, or procedure change implemented against the product during the review period explains discontinuities in the trend data that would otherwise look unexplained. A shift in yield or an OOS rate that changes mid-year against a change-control record showing a raw-material supplier change is a very different finding than the same shift with no explanation on file. See Change Control in Pharma for how a change gets classified and what triggers a regulatory filing.
Trending, not just tabulating
The distinction between a compliant APQR and a weak one usually isn’t which data got pulled — it’s whether the data was actually trended. A table listing twelve months of a given test result is a data dump; the same twelve numbers plotted against their specification limits, with a note on direction (improving, stable, or drifting toward a limit) and a stated statistical basis for what counts as a meaningful shift, is a trend analysis. FDA’s process-validation guidance frames this as the same discipline that underlies Stage 3 CPV: an ongoing program of stability and process data review intended to catch drift before it produces an out-of-specification result, not after.
Where a metric shows a real trend — rising deviation rate on a specific test, a complaint category increasing quarter over quarter, a stability parameter trending toward its limit before expiry — the APQR needs to say so explicitly and assign it an owner, rather than leaving the reader to notice the slope in a chart. This is also where the batch-record, OOS/deviation, and change-control pulls connect: a trend that coincides with a specific change-control record is a finding with an explanation; the same trend with no corresponding change on file is a finding that needs its own investigation.
The conclusions and actions section
21 CFR 211.180(e)’s actual regulatory purpose is “to determine the need for changes in drug product specifications or manufacturing or control procedures” — which means the review is legally incomplete without an explicit conclusion answering that question for each data source reviewed. A defensible APQR conclusions section states, for every category above, one of three outcomes: no action needed (with the basis stated, not just asserted), a recommended change with an owner and target date, or an open item carried forward with its status. Any recommended change identified in the APQR then routes through the site’s actual change-control process — the APQR itself doesn’t implement a specification change, it triggers the record that does.
Two structural mistakes account for most of the weak APQRs that draw inspector attention: closing the document without a stated conclusion for a section that clearly needed one (a rising trend acknowledged in the data but not addressed in the narrative), and treating “no action needed” as a default rather than a finding that itself needs a stated basis. Both are documentation failures more than data failures — the underlying pulls were usually fine.
Frequency and responsibility
211.180(e) sets the U.S. cadence at least annually, per product, under written procedures the site defines itself — the regulation doesn’t prescribe the review’s internal structure beyond the two named data categories, which is why APQR templates vary meaningfully between companies even though the underlying obligation is identical. EU GMP Chapter 1.10 frames the EU PQR the same way: a regular, at minimum annual, review, with the manufacturer responsible for conducting it and, where a separate marketing authorisation holder exists, coordinating with them on any resulting action. In both frameworks, quality unit sign-off on the finished review is standard practice, consistent with the quality unit’s general responsibility for GMP disposition decisions.
Frequently asked questions
Is APQR the same document as a Product Quality Review (PQR)?
They serve the same purpose and, in practice, are usually built as one combined document for products sold in both markets. Formally they trace to different rules — APQR to 21 CFR 211.180(e) in the U.S., PQR to EU GMP Chapter 1.10 — and the EU version has a somewhat broader stated scope (process consistency, specification appropriateness for starting materials as well as finished product), so a combined document needs to cover the EU-specific sections explicitly rather than assuming the U.S. content already satisfies them.
Does the APQR have to include stability data?
21 CFR 211.180(e) itself names batch records and the complaints/recalls/returns/investigations set explicitly; it doesn’t name stability data in that specific clause. In practice, ongoing stability trending is standard APQR content and is the natural link to a site’s Continued Process Verification program — omitting it is a common way an otherwise-compliant APQR still misses a real quality signal.
What happens if the APQR finds a real trend?
The finding gets a stated conclusion and an owner in the APQR itself, and any resulting specification, process, or procedure change routes through the site’s change-control process rather than being implemented directly from the review. The APQR identifies the need for a change; it isn’t the vehicle that makes the change.
How does an APQR differ from a CAPA report?
A CAPA report responds to a specific event that already happened and documents its correction. An APQR is a periodic, product-level roll-up across an entire year that can surface the need for a new CAPA, or reference existing CAPAs as part of its data set, but it is not itself a corrective-action record — it’s the review that decides whether one is warranted.
