APQR: Annual Product Quality Review (APR / PQR)
A practical guide to product-level quality reviews, source-specific expectations, traceable evidence and accountable decisions.

On this page · 12 sections
- 1What APQR means—and what the review decides
- 2APQR, APR, PQR, CPV and batch review compared
- 3Regulatory mapping by jurisdiction and product type
- 4US APR: the evaluation required by 21 CFR 211.180(e)
- 5EU PQR: objectives, timing and starting-material review
- 6ICH Q7: the API quality unit's review responsibility
- 7A review lifecycle from scope to controlled follow-up
- 8Build a traceable evidence model across source systems
- 9Three worked cases: rates, missing links and masked shifts
- 10Failure modes that make a polished APQR unreliable
- 11Conclusions, CAPA, MAH responsibilities and management review
- 12Implementation checklist and records that preserve the reasoning
How does APQR apply to your shop floor?
Pick your industry and scale — Ask V5 rewrites the definition in your context, gives a worked example, and shows what V5 does on day one.
01What APQR means—and what the review decides
APQR (Annual Product Quality Review) is a documented, product-level evaluation of manufacturing and quality information to assess consistency, identify trends and determine whether changes are needed. It is a practical pharmaceutical quality term, not a single statutory name or definition shared by every jurisdiction.
For a short terminology explanation, see the APR/PQR definition. This page focuses on regulatory distinctions and execution: framing the review, interpreting product information and preserving the reasoning behind its conclusions.
As an operating model, a useful APQR connects observations, interpretation, decisions and follow-up. Assembling tables establishes what happened; evaluating them explains what the findings mean and what response is justified. A conclusion should make that connection visible, including the rationale when no change is proposed. This is a practical evidence-to-decision structure, not an additional regulatory content list or a prescribed report format. The aim is a reasoned product-level assessment, rather than a collection of departmental summaries.
Compiling or approving an APQR is not a substitute for authorised product-disposition decisions, including batch release. For US drug products, 21 CFR 211.22(a) assigns the quality control unit responsibility and authority to approve or reject drug products. Keep the product-review conclusion distinct from that disposition authority.
02APQR, APR, PQR, CPV and batch review compared
These are functional comparisons of common quality activities, not statements that their legal scope, frequency or required records are interchangeable. The useful distinction is the question each activity answers and the level at which its conclusions apply.
| Term or activity | Main question | Typical output | Boundary |
|---|---|---|---|
| APQR / APR / PQR | What does the periodic product-level quality assessment indicate? | Product review conclusions and proposed actions | Terminology and obligations are source-specific; see the shorter APR / PQR overview. |
| Continued process verification (CPV) | What does continuing assessment show about process performance? | Process trends and signals for investigation | Process monitoring, not the whole product review |
| Batch record review | What do this individual batch’s records show? | Documented findings and resolution of record questions | One batch, not a periodic product assessment |
| PQS management review | How effectively is the pharmaceutical quality system working? | System-level conclusions and improvement priorities | Quality-system performance, not solely one product |
Process-monitoring outputs can inform the product review. Monitoring methods and detailed CPV interpretation belong on the linked CPV page; this interface does not create an additional mandatory APQR category.
The next sections distinguish the wording and scope of the US regulation, EU GMP guidance and ICH guidance, including their source-specific review obligations.
03Regulatory mapping by jurisdiction and product type
Start with the product and jurisdiction, not the APQR label. These three frameworks have distinct scopes and source statuses; keep their review provisions separate rather than combining them into one mandatory checklist.
| Jurisdiction or framework | Product focus | Source status | Relevant provision |
|---|---|---|---|
| United States | Drug products addressed by Part 211 | Binding regulation | 21 CFR 211.180(e) |
| EU GMP | Authorised medicinal products, including export-only products | Official GMP guidance | Chapter 1, 1.10–1.11 |
| ICH Q7 | Active pharmaceutical ingredients (APIs) | Harmonised guidance | 2.22(15); detailed review section at 2.5 |
The US provision focuses on the quality standards of each drug product; see the broader 21 CFR 211 framework for context. EU Chapter 1, 1.10 expressly includes export-only products. Q7 is API guidance: its harmonised status should not be read as independent legal force in every jurisdiction. Check its applicable local implementation.
For supplements, food, devices and non-API chemicals, check the applicable product and jurisdiction rules separately. This medicines-focused mapping is not a cross-sector requirement.
04US APR: the evaluation required by 21 CFR 211.180(e)
21 CFR 211.180(e) requires written records to support evaluation, at least annually, of the quality standards of each drug product. The decision objective is specific: determine whether drug product specifications, manufacturing procedures or control procedures need changes. Within 21 CFR Part 211, this is a product-level evaluation directed toward those decisions.
The enabling obligation concerns the underlying records, not simply preparation of an annual document. Written records required by Part 211 must be maintained so that their data can be used for the evaluation. A report is therefore not a substitute for maintaining usable records. The provision also requires written evaluation procedures to be established and followed: documenting a procedure and conducting the evaluation are both part of the obligation.
Practical conclusion design
As a practical reporting approach—not a statutory report template—organise the conclusion around three questions:
- What evidence was considered? Identify the records and analyses supporting the evaluation.
- Are changes indicated? Explain whether the findings support changes to specifications, manufacturing procedures or control procedures.
- Why is that conclusion justified? Connect the evidence to the proposed changes, or explain the basis for concluding that no changes are indicated.
This structure makes the reasoning reviewable rather than leaving the reader with a summary of results alone. It is a suggested way to express the evaluation’s outcome, not the US required-content list.
Check the current text of 21 CFR 211.180(e) for the complete provisions governing the evaluation.
05EU PQR: objectives, timing and starting-material review
EU GMP Chapter 1, 1.10 frames the PQR around four objectives: verifying the consistency of the existing process; checking the appropriateness of current specifications for both starting materials and finished product; highlighting trends; and identifying product and process improvements. The guidance therefore addresses both process performance and whether the specifications remain appropriate—not simply whether individual results met them.
The paragraph describes regular periodic or rolling reviews of all authorised medicinal products, including export-only products. It then states that reviews should normally be conducted and documented annually, taking previous reviews into account. These are distinct points: periodic or rolling describes the review approach; normally annually describes the expected review and documentation cadence. “Normally” should not be rewritten as an unconditional annual deadline.
Starting materials: the specific review element
Paragraph 1.10(i) calls for reviewing starting materials, including packaging materials used in the product, especially those from new sources. It particularly identifies review of active-substance supply-chain traceability. Packaging is therefore part of this review element, not a separate category to overlook when concentrating on the active substance.
The rest of the §1.10 minimum list covers results, failures and change. It includes "(ii) A review of critical in-process controls and finished product results"; "(iii) A review of all batches that failed to meet established specification(s) and their investigation"; and "(iv) A review of all significant deviations or non-conformances, their related investigations, and the effectiveness of resultant corrective and preventive actions taken." Note that (iv) asks about the effectiveness of the actions taken, not merely whether they were closed.
| §1.10 item | Review subject |
|---|---|
| (v) | All changes carried out to the processes or analytical methods |
| (vi) | Marketing Authorisation variations submitted, granted or refused, including third-country (export only) dossiers |
| (vii) | Results of the stability monitoring programme and any adverse trends |
| (viii) | All quality-related returns, complaints and recalls and the investigations performed at the time |
| (ix) | Adequacy of any other previous product process or equipment corrective actions |
| (x) | Post-marketing commitments, for new marketing authorisations and variations |
| (xi) | Qualification status of relevant equipment and utilities, e.g. HVAC, water, compressed gases |
| (xii) | Contractual arrangements as defined in Chapter 7, to ensure they are up to date |
Practical illustration—not an additional content requirement
Assume an active substance keeps the same internal material code and commercial supplier, but its manufacturing source changes during the review period. An illustrative review structure would connect the source-change record to the affected incoming material lots, then to the product batches that used those lots and the associated review observations.
Keep material identity, supplier identity and active-substance source as separate fields in that illustration. An unchanged supplier name alone would not distinguish the assumed source change. Use backward genealogy tracing to navigate from a product batch to its material lots, and connect the source information to supplier risk management. Record which observations relate to which source rather than presenting an undifferentiated material summary. This illustrates how to organise the review evidence; it does not expand the wording of 1.10(i).
06ICH Q7: the API quality unit's review responsibility
For API product quality reviews, ICH Q7, 2.22(15) includes “Performing product quality reviews (as defined in Section 2.5).” among the independent quality unit's main responsibilities. The governing wording in 2.22 is: “The main responsibilities of the independent quality unit(s) should not be delegated.” These are statements from ICH guidance, not a combined US–EU review checklist.
ICH Q7 §2.50 sets the API framework. Regular quality reviews of APIs "should be conducted with the objective of verifying the consistency of the process" and "should normally be conducted and documented annually". They should include at least: a review of critical in-process control and critical API test results; all batches that failed to meet established specification(s); all critical deviations or non-conformances and related investigations; any changes carried out to the processes or analytical methods; results of the stability monitoring program; all quality-related returns, complaints and recalls; and adequacy of corrective actions.
§2.51 then turns review into decision: "The results of this review should be evaluated and an assessment made of whether corrective action or any revalidation should be undertaken." Reasons for corrective action "should be documented", and agreed corrective actions "should be completed in a timely and effective manner." Compared with EU §1.10, the Q7 list is shorter—it has no starting-material supply-chain, marketing-authorisation variation, equipment-qualification or contractual-arrangement items—and it uses "critical" deviations where the EU text says "significant".
As an operating arrangement, production, laboratory and quality-system contributors can prepare source-linked summaries, explain discrepancies and identify unresolved questions. Where records span systems, MES–QMS integration is relevant to that preparation. The hand-off should give the responsible quality unit both the summaries and access to underlying records for evaluation. Preparing data supports the review; it does not transfer responsibility for performing it.
07A review lifecycle from scope to controlled follow-up
The following sequence is a practical implementation model, not a prescribed seven-step regulatory method. Adapt it to the product, applicable source requirements and the organisation’s review procedure.
Start with a compact review charter: product identity, site coverage, period boundaries, contributors, reviewer roles and planned evidence sources. These suggested planning fields are not a combined US/EU/ICH mandatory checklist and do not create additional approval roles.
- Define the product and period. Record the scope and explain any grouping so contributors work to the same boundaries.
- Identify source-specific obligations. Map applicable review topics separately rather than treating every framework’s content list as interchangeable.
- Establish the batch population and cut-off. State the inclusion rule, retain the batch list and document exclusions with reasons.
- Compile and reconcile evidence. Connect each included batch to its batch record and relevant review data; identify gaps before analysis.
- Analyse meaningful comparisons. Choose comparable populations, periods and denominators; document limitations that affect interpretation.
- Reach reviewed conclusions. Have designated reviewers assess whether the evidence supports each conclusion and record unresolved questions.
- Assign and follow up actions. Name owners, agreed target dates and closure evidence; route proposed changes through change control where appropriate.
Keep three date concepts distinct. The review period defines the interval being assessed. The extraction date records when data were retrieved. The cut-off convention defines which records and record versions enter the review dataset. Document how late-entered records and subsequent updates are assessed and incorporated or referenced. A consistent convention makes successive extracts explainable; it does not establish a universal grace period. Preserve the original dataset identity when issuing an updated analysis.
- Define scopeSet product, period, sites, responsibilities and applicable framework.
- Identify evidenceLocate source records and establish the review population.
- Reconcile inputsResolve missing links, conflicting versions and repeated transfers.
- Analyse patternsUse explicit denominators, units and meaningful comparison groups.
- Evaluate findingsSeparate supported conclusions from limitations and open questions.
- Record decisionsResponsible reviewers document conclusions and proposed responses.
- Follow actionsTrack ownership, assessment points and later review references.
08Build a traceable evidence model across source systems
An illustrative APQR evidence model links the review identifier to product and batch identities, source records, observations, conclusions and actions. Each observation should point back to its supporting records; each conclusion should identify the observations considered; each action should retain its decision reference. Keep source records distinct from derived tables and report summaries: tables contain selected or calculated data, while summaries communicate interpretation. Retain extraction criteria, calculation rules and source-version references so another reviewer can reproduce the analysis. This is a proposed structure, not a prescribed database schema.
Candidate evidence connections—not a regulatory minimum-content list—include:
- Production to laboratory: connect batch identifiers to sample identifiers, test results and methods.
- Production to materials: connect consumed material lots to receipt identifiers and the batches using them.
- Production to quality events: connect affected batches to deviations, investigations and action records; MES–QMS integration addresses this interface.
For every interface, define stable source identifiers, field mappings and the meaning of dates, units and record status. Distinguish manufacture dates from approval dates, preserve original units alongside conversions, and specify how draft, superseded or cancelled records are handled. These are practical data-integrity-by-design controls, not additional review categories.
Recover integration exceptions explicitly. Separate failed transfers from unmatched identifiers, repeated transfers and conflicting versions: each can distort the dataset differently. Preserve source references and exception details; correct the mapping or extraction, reconcile duplicates or versions against the source, and rerun affected comparisons. Record the resolution and whether it changes observations or conclusions. Do not silently discard unresolved exceptions or treat a missing transfer as evidence that no event occurred.
Generic software support can be configured for collection, reconciliation and presentation of information. Attribute interpretations and decisions to named reviewers; the human approval and release boundary described above remains unchanged.
Review
Defines the assessment boundary.
- review_id
- product_id
- period_start
- period_end
Batch
Identifies reviewed production.
- batch_id
- product_id
- site_id
- manufacture_date
Source Record
Preserves evidence identity.
- record_id
- batch_id
- source_system
- source_record_id
Observation
Carries interpretable results.
- observation_id
- record_id
- result_value
- result_unit
Conclusion
Records accountable interpretation.
- conclusion_id
- review_id
- rationale
- reviewer_id
Action
Connects decisions to follow-up.
- action_id
- conclusion_id
- owner_id
- status
09Three worked cases: rates, missing links and masked shifts
These hypothetical cases illustrate review reasoning, not regulatory thresholds. None creates an automatic CAPA or rejection rule; reviewers must interpret the findings within the stated population and assumptions.
Case 1—Hypothetical event-linked batch rates
Assume the same event definition and comparable product review populations across two periods. Each affected batch counts once, regardless of how many events it has. The denominators include all reviewed batches in each period: 40 in A and 60 in B.
Period A: 2 ÷ 40 × 100 = 5% affected batches. Period B: 6 ÷ 60 × 100 = 10%. The increase is 10 − 5 = 5 percentage points. Raw event counts answer how many occurrences were recorded; affected-batch rates answer what share of reviewed batches experienced the event. These numerators are affected batches, not necessarily total events. Neither counts nor the percentage-point change establishes a cause.
Review decision: investigate population comparability and recurrence using the relevant batch execution histories.
Case 2—Hypothetical incomplete batch join
Assume the production register defines the review population: 25 unique batches for one product and period. A laboratory extract contains 30 result rows linked to 24 unique batch identities within that population; batches can have multiple results.
Batch-link coverage is 24 ÷ 25 × 100 = 96%; 25 − 24 = 1 batch remains unmatched. Thirty result rows do not establish complete batch coverage. An unmatched link does not itself establish that testing was omitted: the demonstrated gap is in the dataset relationship, not yet in test execution.
Review decision: resolve the evidence link before interpreting completeness; MES-to-LIMS integration is the relevant system boundary.
Case 3—Hypothetical pooled mean hiding a shift
Assume eight batches of one product, one comparable assay observation per batch, measured in percent of label claim. The hypothetical product specification is 95.0–105.0%. The population comprises four pre-change and four post-change batches, with no other observations included.
Pre-change mean: (98 + 99 + 100 + 99) ÷ 4 = 99%. Post-change mean: (101 + 102 + 103 + 102) ÷ 4 = 102%. Pooled mean: (396 + 408) ÷ 8 = 100.5%. The subgroup means differ by 102 − 99 = 3 percentage points.
All observations meet the assumed specification, but the shift still merits interpretation. Eight observations do not establish causation.
Review decision: assess the segmented shift in context rather than relying on the pooled mean.
10Failure modes that make a polished APQR unreliable
Use this matrix as an analytical and governance check, not an additional statutory checklist. A polished report can still conceal weak comparisons or unsupported decisions.
| Failure mode | Why it misleads | Practical control |
|---|---|---|
| Changing denominators without explanation | Rates lose comparability | Display population definitions beside metrics |
| Pooling changed methods or specifications | Unlike results appear comparable | Segment or justify comparison |
| Averaging across materially different conditions | Local changes disappear | Show relevant subgroups using production context tagging |
| Treating every repeated observation as a new event | Counts inflate | Define event and batch counting rules |
| Unsupported no-change conclusion | The report outruns its evidence | Link conclusions to assessed observations |
| Ownership without follow-up | Actions become static entries | Define the next review point |
Distinguish three states when writing conclusions:
- Absent evidence: the relevant information is missing; state what cannot be assessed.
- Inconclusive evidence: information exists but does not resolve the question; state the uncertainty and what would resolve it.
- Evidence of deterioration: observations support an adverse change; describe that change and the assessment or action it warrants.
Each warrants a different conclusion; none should be silently rewritten as satisfactory performance. Refer to the evidence-model and system-integrations section for technical recovery and the worked cases for arithmetic. For action follow-up, distinguish the next review point from a CAPA effectiveness check: scheduling a review does not itself demonstrate that an action worked.
11Conclusions, CAPA, MAH responsibilities and management review
EU GMP guidance connects product-review findings to decisions under the Pharmaceutical Quality System (PQS). Under EU GMP Chapter 1, 1.11, the manufacturer and, where different, the marketing authorisation holder (MAH) should evaluate the review results and assess whether corrective and preventive action (CAPA) or revalidation should be undertaken. The decision is whether those responses are warranted; a trend is not, by itself, an instruction to initiate both.
Where the MAH is not the manufacturer, paragraph 1.11 calls for a technical agreement between the parties defining their respective responsibilities for producing the PQR. This responsibility allocation sits alongside the expectation that the manufacturer and the different MAH evaluate the results.
Manage follow-up without confusing effectiveness questions
Paragraph 1.11 also calls for management procedures for ongoing action management and review, with the effectiveness of those procedures verified during self-inspection. That is a procedure-level effectiveness question—not a finding that an individual CAPA has succeeded. Keep it distinct from a particular action’s CAPA effectiveness check.
Product-review evaluation is also distinct from periodic management review of the PQS. Chapter 1, 1.6 addresses review of the system’s operation with senior-management involvement. It should not be read as prescribing an APQR signature list or a fixed management-review calendar.
A practical conclusion structure
As an operating practice, write each conclusion so another reviewer can distinguish the finding from the proposed decision:
- Observation: identify the result and the records supporting it.
- Interpretation: explain its product-quality significance and the reasoning used.
- Proposed response: state the action proposed, or explain why no new action is proposed.
- Responsible person: name the person accountable for progressing the response.
- Next checkpoint: identify when or under what condition progress will be reviewed.
- Remaining uncertainty: state what is unresolved and what information would resolve it.
This structure supports explicit decisions without assigning approval authority or imposing a universal deadline.
12Implementation checklist and records that preserve the reasoning
Use this implementation checklist before final approval. It tests whether the APQR is understandable and actionable; it is not the required-content list for any jurisdiction.
- Applicability is identified, with the product and review period defined.
- The source population is reconciled, including references to relevant batch records.
- Relevant units of measure and versions of methods, specifications and datasets are preserved.
- Comparisons are explained and limitations remain visible.
- Conclusions link to evidence, and decision responsibilities are recorded.
- Follow-up actions and their owners are identifiable.
- The review version and supporting record references are retrievable.
Record form and the supporting evidence package
For records required under Part 211, 21 CFR 211.180(d) permits originals or true copies, with photocopies, microfilm, microfiche and other accurate reproductions as examples. This permission should not be read as prescribing a universal electronic format or a single APQR retention period; check the applicable retention provisions for each record category.
As an implementation practice, retain source references, extraction and transformation definitions, unresolved limitations, conclusion history and action links in a controlled arrangement appropriate to the organisation. Where relevant, include references to the audit trail that explains revisions. This recommended evidence package preserves the reasoning behind the review; it is distinct from the regulation’s wording on record form.
Final test: can another authorised reviewer understand what changed, what remains uncertain, what was decided and who will follow up?
Frequently asked questions
Q.What should an APQR document when no batches were manufactured during the review period?+
A practical approach is to record the review period, confirm that no batches were manufactured, and distinguish unavailable production observations from other review information. Assess relevant stability results, complaints, returns, recalls and corrective actions rather than treating zero production as zero activity. For APIs, these subjects appear in ICH Q7 §2.50. Explain which conclusions cannot be drawn without new manufacturing data, and identify any follow-up through the stability program or other applicable quality processes.
Q.How can reviewers avoid overinterpreting trends for a low-volume product?+
As a practical analytical approach, show the number of batches and observations behind each apparent trend, retain individual results, and explain gaps in production. Avoid treating a percentage change based on very few events as proof of process deterioration or improvement. Historical comparisons can add context, but distinguish genuinely comparable batches from those made under different conditions. Use a batch comparison overlay to support interpretation, not to substitute a visually convincing chart for a reasoned assessment of limited data.
Q.What belongs in the current review when a complaint arrives after the data cut-off?+
A practical cut-off procedure should distinguish the complaint’s receipt date, the affected batch and the period to which the underlying event relates. Record whether the complaint changes the current review’s conclusions; if it does, consider updating the review before approval or issuing a controlled follow-up afterward. For APIs, ICH Q7 §2.50 includes all quality-related complaints among review subjects. Keep complaint investigation and any recall assessment separate from the administrative decision about which review contains the information.
Q.When is grouping products in an EU PQR scientifically justified?+
EU GMP Chapter 1 §1.11 permits grouping by product type where scientifically justified, giving solid dosage forms, liquid dosage forms and sterile products as examples. A practical justification should explain why shared analysis can answer the review questions without concealing product-specific differences. Examine whether manufacturing processes, specifications and quality attributes support meaningful comparison; a shared dosage-form label alone is not the analysis. Preserve identifiable product-level findings, especially where a critical quality attribute behaves differently within the proposed group.
Q.How should a manufacturer and MAH address conflicting interpretations of review results?+
EU GMP Chapter 1 §1.11 says the manufacturer and, where different, MAH should evaluate review results; a technical agreement should define their respective responsibilities for producing the review. A practical response to disagreement is to document each interpretation, its supporting data and the unresolved decision. Use the agreed governance route to obtain a recorded resolution from responsible people, rather than blending incompatible conclusions into vague wording. Link resulting actions to management review where appropriate.
Q.How should results be compared when a specification or test method changes mid-period?+
For APIs, ICH Q7 §2.50 includes review of changes to processes or analytical methods. As a practical comparison method, label results with the applicable method and specification versions, show the change date, and separate pre-change and post-change populations initially. Explain whether a combined trend remains meaningful; do not assume identical units establish comparability. Reference the relevant change control, and distinguish a shift in measured results from a change in acceptance criteria when drawing conclusions.
Q.What should happen if a source record is corrected after APQR approval?+
A practical response is to assess the correction’s effect on the approved review, rather than silently replacing the underlying value. Identify affected tables, calculations, trends, conclusions and actions. If the correction changes the assessment, issue a controlled amendment or revised review with a clear reason and appropriate human approval. If it does not, record the impact assessment and its rationale. Maintain a traceable connection between the corrected source and the review through document control, preserving the history of the decision.
Q.Can rolling review dashboards replace a documented annual product quality review?+
EU GMP Chapter 1 §1.10 recognises periodic or rolling reviews and says reviews should normally be conducted and documented annually, taking previous reviews into account. A practical dashboard-based approach therefore needs more than continuously refreshed charts: retain an identifiable review period, the information assessed and the resulting conclusions. Dashboards can support preparation and ongoing analysis, while responsible people evaluate findings and actions. The shorter APR / PQR definition explains the terminology; the operational question is whether the review remains documented and interpretable.
Primary sources
Want to see how APQR could fit into your own records and workflows? Explore the related V5 pages or talk to our team about what applies to your operation.
