Summary
- The problem
- Annual reviews are compiled from separate manufacturing, laboratory and quality sources, and the population is easy to get wrong. A draft built from released batches alone leaves out rejected batches and those still awaiting disposition.
- Seal’s approach
- The review starts from an agreed population of product, period, sites and source domains, reconciled before anything is analysed. Batch outcomes, stability, complaints, changes and supplier history are evaluated in the same product context.
- What changes
- Each batch in scope appears with its state at the data cut. Conclusions carry into actions with owners and evidence of completion, and a later trend can be traced to the review that preceded it.
- Where to start
- One product review and the source populations your team currently has to reconcile. Book a demo.
Coordinate the review across every evidence domain.
Coordinate the annual product quality review (APQR) across manufacturing, laboratory, stability, suppliers and field experience. Carry the scientific conclusions into work with owners and follow-up.
The next review evaluates the effect of earlier actions.
Agree what the review includes before counting it.
Define products, presentations, sites, review period and any grouping rationale. Assign a source owner and completeness check to each evidence domain. Keep rejected, pending, excluded and late-arriving records distinguishable from missing data.
Reconcile expected records with the selected population before interpreting rates, trends or apparent absence of events.
Review scope
- Coverage
- Products, presentations, sites and period
- Population
- Inclusion, exclusions and data cut
- Coordination
- Source owners and section reviewers
Bring the whole product history into view.
Connect yields, batch outcomes, laboratory and stability results with methods and specifications. Include supplier changes, incoming quality, deviations, OOS/OOT, CAPA, complaints, returns and recalls within the agreed scope. Retain differences between sites and presentations.
Check completeness across domains, including external partners. An empty export is not proof that no relevant event occurred.
Product evidence set
- Process and quality
- Batch outcomes, results and stability
- Materials and changes
- Sources, genealogy and implemented changes
- Quality and field
- Events, CAPA, complaints and recalls
Keep the basis behind each chart and conclusion.
Review trends, recurrence, stability and the observed effect of changes. Retain cohorts, denominators, exclusions and analysis versions so charts remain reproducible. Ask Neil to prepare a referenced narrative; reviewers determine what the evidence supports.
Freeze the reviewed evidence cut with the issued report. Later corrections or information need a traceable update.
Product review report
- Analysis
- Population, calculation and source versions
- Interpretation
- Findings, limitations and product impact
- Approval
- Section review and final conclusion
Turn each finding into assigned work.
Connect conclusions to CAPA, process or method change, supplier follow-up, validation work and other agreed actions. Assign an owner, due date and evidence requirement. Keep prior commitments visible in the next review.
Distinguish administrative closure from an effective response. Recurrence remains connected to the earlier finding and decision.
Review follow-up
- Origin
- Finding and approved conclusion
- Work
- Owner, due date and linked action
- Effect
- Completion evidence and effectiveness review
A draft built from released batches omits two.
The draft used only released batches. The agreed manufacturing-period population also contains a rejected batch and a batch awaiting disposition.
Product P-022 / annual review population
12 batches manufactured in scope
Awaiting disposition
PQR022-OPEN- Batch
- P-022-12
- Disposition
- Open
- Data cut
- 11 September 2026
Batch P-022-12 was manufactured in the review period but has no final disposition at the data cut. Keep it in scope with its state explicit.
What follows
Record the open state and define how a later disposition will be handled in the review or an update.
The draft’s ten released batches omit two batches required by this review definition. I would restore the rejected and pending-disposition batches, preserving their distinct states.
P-022 / annual review population
DraftReview all twelve manufactured batches in scope. Release status is a subgroup, not the population-selection rule.
Review coordinator
Reconcile the ten released batches against the agreed manufacturing-period population.
Quality lead
Bring the rejected batch’s investigation and actions into the review.
Batch disposition owner
Confirm the pending state and the plan for any post-cut update.
What a complete review keeps.
Why teams choose Seal for product quality review
Annual product reviews are traditionally compiled once a year from exports of batch, laboratory, complaint and change records, and a large part of the effort goes into reconciling the population. Seal starts the review from an agreed population drawn from records that already share one platform, reconciled before anything is analysed.
Conclusions become owned actions on the same records, so a trend in next year’s review can be traced to the decision that preceded it.
A complete review connects different kinds of evidence.
Bring batch performance, methods, materials, stability, deviations, complaints and changes into the same product context.
Give each source domain an owner, period and completeness check. Separate missing data from a true absence of events. Keep product grouping and exclusions explicit rather than flattening different presentations or sites into one average.
The action survives the report.
Link each follow-up to the finding, owner, due date and evidence needed to resolve it.
Review whether prior actions were effective, not only administratively closed. A later trend or recurrence should be traceable to the earlier product review and the decision made then.
Connected to the surrounding work.
Keep specialised systems where they belong. Agree the source of each record, the permitted actions and the reconciliation at each handoff.
Explore integrationsAQuestions and answers
Are APR, PQR and APQR identical?
Terminology and expectations vary by jurisdiction and company procedure. Configure the product review against your applicable requirements; do not assume one fixed template covers every use.
Can multiple products or sites be grouped?
Where justified, retain the grouping rationale and preserve meaningful differences. The review population and analyses should remain traceable to the underlying products and sites.
What if a batch is not yet disposed?
Use the agreed inclusion rule and record the state at the data cut. Define how later information is reviewed; do not silently omit the batch or rewrite the issued report.