APQR and PQR. A governed product review, not a year-end spreadsheet project.

An Annual Product Quality Review, Annual Product Review, or Product Quality Review should evaluate whether a product and its manufacturing process remained consistent and appropriately controlled during a defined period. It should not begin with weeks of copying numbers into a presentation.

Seal assembles the review population continuously from governed manufacturing, laboratory, quality, supplier, stability, complaint, change, validation, and regulatory records. Reviewers approve the population and methods, investigate signals, record conclusions, and track actions to completion.

A review definition identifies product family, strengths and presentations, sites, markets, contract parties, period, batch populations, included data domains, methods, statistics, owners, due dates, approvers, and required sections.

Grouping across products or presentations retains rationale and boundaries. A product introduced, transferred, discontinued, or manufactured at multiple sites may require partial periods or separate populations. Seal makes inclusion logic visible so the finished report cannot conceal missing scope.

The APQR definition should state what arrives from each source and how completeness is proven. For manufacturing that may be every started order and final disposition. For LIMS it may be every required test, approved reportable result, invalid result history, and open investigation. For complaints it includes a denominator and market population, not only closed complaint counts.

The contract records source system, owner, identifiers, inclusion logic, status mapping, units, effective dates, refresh cadence, reconciliation method, and exception owner. Partner submissions include the expected period, format, approval, and revision history.

Seal uses these contracts to expose missing or incompatible evidence before narrative review begins. A report section cannot quietly declare “not applicable” because an interface returned no rows.

Before calculating trends, the owner reconciles all batches manufactured, packaged, tested, released, rejected, reworked, reprocessed, returned, or otherwise dispositioned during the period. ERP, MES, LIMS, QMS, and external partners may each contribute.

Seal compares expected and received records by identity and status. Missing batches, duplicate transfers, late partner data, changed dispositions, and unresolved interfaces remain exceptions. The reviewer approves the population and records any justified exclusions.

FDA 21 CFR 211.180(e) requires at least annual evaluation of drug-product records to determine the need for changes in specifications or manufacturing and control procedures. Seal supports the evidence workflow; the manufacturer defines the applicable regulatory scope and conclusion.

Batch outcomes include yields, reconciliations, cycle times, hold times, critical process parameters, critical quality attributes, alarms, interventions, deviations, rework, and disposition. Statistics remain tied to product, process version, site, scale, equipment class, material sources, and effective control strategy.

Comparability comes before aggregation. A process change or site transfer can create a new cohort instead of disappearing inside an annual average. The review can still evaluate the complete year while showing which populations are scientifically compatible.

The review includes significant raw materials, components, packaging, suppliers, manufacturer sites, incoming failures, changes, complaints, shortages, and qualification status. Actual use genealogy identifies which lots and suppliers contributed to the product population.

A recurring quality attribute can therefore be explored by material source, lot property, supplier change, or receipt condition. Supplier metrics remain linked to the specific events and affected batches rather than appearing as an unrelated scorecard.

Release, in-process, incoming, environmental, utility, and applicable microbiology results enter with sample, method, instrument, specification, units, calculation, OOS or OOT state, and review.

The system distinguishes true result trends from method changes, specification changes, reporting-unit changes, and different sample stages. Invalidated results remain part of the analytical history while only approved reportable results enter the configured product statistic.

Events are grouped by product, process stage, cause, equipment, material, shift, site, recurrence, severity, impact, and disposition. Open events, overdue investigations, repeated root causes, and ineffective actions remain visible.

The review evaluates whether individual investigations collectively indicate a systemic issue. CAPA status alone is insufficient; the reviewer sees the source events, implemented actions, effectiveness evidence, and recurrence after implementation.

Approved and implemented changes enter with rationale, affected product and process, implementation date, batches, validation, regulatory impact, training, and effectiveness plan. Cancelled and pending changes remain distinguishable.

The review compares the expected outcome with post-change evidence. A change implemented near period end may require follow-up in the next review, recorded as a controlled commitment rather than a note that disappears in the final document.

Protocols, batches, markets, storage conditions, pulls, tests, results, OOT events, excursions, commitments, and shelf-life decisions form the stability section. Missing pulls and pending investigations remain visible.

Seal separates timepoint trends, batch-to-batch variation, storage-condition comparisons, and method or specification changes. Conclusions reference the exact datasets and support any required action on shelf life, storage, packaging, or ongoing commitments.

Complaint categories, rates and denominators, seriousness, investigations, returned samples, defects, adverse events where applicable, recalls, field actions, and distribution patterns contribute to the review.

The product and batch genealogy links market evidence back to manufacturing and testing. A complaint trend can be compared with process, component, supplier, packaging, label, and stability populations without manually joining records.

Counts without exposure can mislead. Seal retains batches produced, units packaged, doses distributed, tests performed, complaints received, and other approved denominators alongside event counts.

Rates define the population, unit, inclusion rules, missing-data behavior, and calculation version. The report shows both numerator and denominator so a reviewer can reconstruct the metric.

The approved analysis plan defines variables, cohorts, exclusions, transformations, control or specification limits, minimum data, chart type, and interpretation. Cp, Cpk, Pp, Ppk, control charts, and trends are used only where appropriate to the data and purpose.

Seal preserves the dataset version and analysis method. A refreshed live dashboard cannot silently alter the evidence behind an approved annual conclusion. Reviewers can inspect every contributing record and justified exclusion.

Each section should resolve whether processes and controls remain consistent, specifications and methods remain appropriate, changes are needed, validation state remains acceptable, supplier controls are effective, stability supports claims, and prior actions achieved their purpose.

Conclusions carry evidence references, author, review, approval, confidence or limitation where appropriate, and resulting action. “No trend observed” is not accepted as an unsupported free-text closure when required data is missing.

An APQR action has owner, due date, priority, rationale, source conclusion, affected product or process, required evidence, status, escalation, and effectiveness requirement. It can instantiate change control, CAPA, validation, supplier, method, or regulatory work.

Closure returns evidence to the review. Open actions carry into the next period automatically, preserving whether they were late, changed, cancelled, or found ineffective.

The approved review retains scope, population, source-record versions, datasets, analyses, charts, narrative, conclusions, actions, and signatures. Later source corrections or disposition changes do not rewrite it.

Material post-approval information creates an amendment or documented assessment. Meanwhile the next review period continues accumulating current evidence. Seal therefore supports both an immutable approved report and a continuous product-quality view.

Data owners reconcile source populations. Process, laboratory, supplier, stability, complaint, validation, regulatory, and quality specialists evaluate their domains. The review owner resolves cross-domain conclusions and overdue evidence. Authorized quality roles approve the final product assessment and resulting commitments.

Seal records delegation, due dates, review status, comments, returned sections, changes after review, and signature meaning. A contributor can correct an underlying source through its governed workflow but cannot edit a frozen analysis to obtain a preferred conclusion.

The calendar begins before period close. Continuous readiness allows teams to resolve missing sources, definitions, and recurring signals during the year while preserving a formal cutoff, approved population, and accountable periodic decision.

The first implementation should reconcile one product with multiple strengths or presentations, all batch outcomes, laboratory results, OOS and OOT, deviations, CAPA, changes, stability, complaints, suppliers, validation, prior commitments, statistics, conclusions, and actions.

Include missing partner data, a late disposition change, a method revision, an excluded batch, an ineffective CAPA, and a post-approval correction. The system is ready when every chart is reproducible and every conclusion resolves to governed evidence.