ICH Q7 clause 6: Documentation and records

The 24 audit questions covering clause 6, each with the objective evidence to request, the nonconformities most often raised against it and what to sample. Part of the free ICH Q7 API GMP audit checklist, which holds 350 items across 18 clauses.

24 items in this clause 1 section 350 items in the full checklist ICH Q7 · updated 2026-06-22

All 24 questions for clause 6

Open any row for its objective evidence, common nonconformities and auditor tips. You can check items off as you go. This browser remembers your progress across all 18 clauses of this checklist.

§6 Documentation and records 24 items · ~120 min
6.10 Is there a documentation system covering the full lifecycle of GMP documents (creation, review, approval, issuance, revision, retirement)?
Objective evidence
  • Document control SOP covering creation, review, approval, distribution, and retirement
  • Master document list or document index showing current revision status
  • Revision history logs for key documents (SOPs, specifications, master batch records)
  • Evidence of periodic review cycle for GMP documents
  • Superseded document archive with controlled access
  • Electronic document management system (EDMS) validation records if applicable
  • Distribution records showing controlled copies issued to relevant personnel
  • Training records confirming document control SOP training for authors and approvers
Common nonconformities
  • No master document list — unable to determine which revision is current
  • Obsolete documents found at points of use alongside current versions
  • No revision history trail — documents appear to have been changed without tracking
  • Electronic documents stored on shared drives without access controls or audit trails
  • Retention periods not defined or not consistently applied across document types
  • Critical documents (batch records, specifications) missing approval signatures
Auditor tip

Auditors should verify the existence of a comprehensive document management system covering the full document lifecycle from creation through retirement. For API manufacturers, this system must handle both paper and electronic documentation with equivalent controls. Key areas to probe include whether the system tracks revision history with clear supersession trails, whether obsolete documents are promptly removed from points of use, and whether there is a master document list or equivalent index. The system should extend to all GMP-critical documents including specifications, SOPs, master batch records, and laboratory methods.

6.11 Are GMP documents prepared, reviewed, approved, and distributed according to written procedures, with defined retention?
Objective evidence
  • Document approval SOP defining roles (author, reviewer, approver)
  • Signed and dated approval pages on SOPs, specifications, and master batch records
  • Retention schedule document listing all record types and their retention periods
  • Archive inventory demonstrating records retained per policy
  • Batch record retrieval demonstration from archive within reasonable timeframe
  • Evidence of periodic archive condition monitoring (temperature, humidity if paper)
  • Destruction logs for records past their retention period
  • Backup procedures for electronic records with verified restoration capability
Common nonconformities
  • Batch records destroyed before expiry date + 1 year period elapsed
  • No formal retention schedule — records kept ad hoc or only until space runs out
  • Approval signatures missing or backdated on critical documents
  • Unable to retrieve batch records for recently distributed batches
  • Paper archives stored in uncontrolled conditions (moisture, pest exposure)
  • Electronic backup system untested — no verified restoration capability
Auditor tip

This clause specifically addresses the approval workflow and retention periods. Auditors should verify that every GMP document has evidence of preparation by a qualified author, review by a subject matter expert, and approval by an authorized person (typically QA). The retention periods are critical compliance points: 1 year post-expiry for APIs with expiry dates, or 3 years post-complete-distribution for APIs with retest dates. Whichever is longer should apply. Verify retention through sampling batch record archives.

6.12 Is the issuance and revision of documents controlled, with each revision carrying a documented reason and history?
Objective evidence
  • Revision history pages on key documents showing reason for change
  • Change request forms linking document revisions to change control records
  • Evidence of same-level approval for document revisions (QA approves QA documents)
  • Superseded document collection and destruction logs
  • Controlled copy distribution and recall records for superseded versions
  • Effective date management showing no gap between supersession and new version
  • Training records for personnel affected by significant document revisions
  • Periodic audit of points of use to verify only current versions are available
Common nonconformities
  • Revision history states only 'updated' without specific reason for change
  • Production personnel approving changes to QA-owned specifications
  • Multiple current versions of the same document in circulation
  • No evidence of superseded document collection after new version issued
  • Significant time gap between effective date of new version and removal of old
  • No formal change request linking document revision to its triggering event or CAPA
Auditor tip

Focus on the complete lifecycle trail for document revisions. Every revision must have a documented reason (not just 'updated' or 'revised'). The same organizational function that originally approved the document must approve changes — this prevents production from unilaterally changing QA-approved specifications. Auditors should sample several key documents and trace their revision history to verify consistent application.

6.13 Are there procedures defining the retention of all GMP-relevant record types with specific retention periods?
Objective evidence
  • Written retention procedure listing all document types and their retention periods
  • Retention schedule matrix covering production, laboratory, quality, and distribution records
  • Archive facility with organized storage and retrieval system
  • Periodic archive inventory or audit reports
  • Evidence of regulatory-driven retention extensions (e.g., ongoing investigations)
  • Destruction authorization and log for records past retention
  • Backup and disaster recovery procedures for electronic records
  • Cross-reference to contract storage facilities if off-site archival is used
Common nonconformities
  • No formal retention schedule — retention periods decided case by case
  • Retention periods shorter than regulatory minimum (1 yr post-expiry or 3 yr post-distribution)
  • Development history reports discarded after regulatory approval
  • No disaster recovery plan for electronic or paper archives
  • Off-site archive without quality agreement or periodic audit
  • No documented procedure for extending retention when regulatory investigations are ongoing
Auditor tip

This clause goes beyond simply keeping records — the retention procedure must cover ALL GMP-relevant document types with specific retention periods for each. The list in the guidance is illustrative, not exhaustive. Key considerations for API manufacturers include maintaining development history (for regulatory submissions), technology transfer records (for site changes), and distribution records (for recall capability). Auditors should request the retention schedule and verify it against actual practice.

6.14 Are records retained for at least one year past the batch expiry, or three years after distribution, whichever is longer?
Objective evidence
  • Retention schedule specifying the two retention rules and how they are applied
  • Batch record archive demonstrating compliance with retention periods
  • Sample batch records from oldest retained batches to verify legibility
  • Distribution completion dates recorded for retest-date APIs
  • Environmental monitoring of paper archive storage conditions
  • Electronic media migration records (e.g., tape-to-disk conversions for aging media)
  • Regulatory hold procedures for records under investigation
  • Destruction log showing no premature destruction of batch records
Common nonconformities
  • Batch records destroyed before the retention period elapsed
  • No tracking of distribution completion date — cannot calculate 3-year retention trigger
  • Paper records stored in damp or thermally unstable conditions becoming illegible
  • Electronic records on obsolete media (floppy disks, DAT tapes) with no migration plan
  • Retention policy only addresses one rule (expiry OR distribution) but not both
  • No regulatory hold capability — records destroyed during ongoing investigation
Auditor tip

This is one of the most frequently cited clauses in API inspections. The two retention rules are not alternatives — you must apply whichever results in a longer retention period. For APIs with expiry dates: all records retained for at least 1 year after the expiry date. For APIs with retest dates: at least 3 years after the batch is completely distributed. 'Completely distributed' means when the last portion of the batch leaves the manufacturer's control. Local regulations may require longer periods. Storage conditions must maintain legibility — this applies to thermal paper, ink, and electronic media degradation.

6.15 Are record entries made indelibly, legibly, contemporaneously, and signed and dated?
Objective evidence
  • Batch records with contemporaneous entries (ink, signed, dated at time of activity)
  • Single-line corrections with original entry readable, initialed, dated, and reason noted
  • SOP on good documentation practices (GDP) covering data integrity requirements
  • Training records on GDP for all personnel who make GMP entries
  • Blank line-out practices on partially completed forms
  • No evidence of white-out, overwriting, or pencil entries
  • Electronic audit trails showing who made entries and when (for electronic records)
  • Periodic GDP audit results demonstrating ongoing compliance
Common nonconformities
  • White-out or correction fluid found on batch records
  • Pencil entries on GMP documents
  • Entries clearly made retrospectively (times out of sequence, same pen for entire shift)
  • Corrections that obscure the original entry
  • No reason documented for corrections
  • Blank spaces on completed batch records not lined out
  • Entries without initials, signatures, or dates
Auditor tip

Data integrity fundamentals for GMP records. Every entry must be: (1) indelible (no pencil, no erasable ink), (2) made at the time of the activity (not recorded later from memory), (3) attributable to a specific person (signature or initials), and (4) dated. This is the ALCOA+ principle applied to API manufacturing. Blank spaces should be lined out to prevent retrospective entries. Auditors should examine actual batch records for evidence of contemporaneous recording, single-line corrections with initials/date/reason, and absence of white-out or overwriting.

6.16 Are corrections made so the original entry remains legible, with a reason, date, and signature?
Objective evidence
  • Batch records showing single-line corrections with all required elements
  • GDP SOP section specifically addressing correction procedures
  • Electronic audit trail examples showing before/after values with reason codes
  • Training materials covering proper correction technique
  • GDP audit findings showing correction compliance rate
  • Deviation reports referenced from corrections where applicable
  • No evidence of undocumented alterations in reviewed batch records
  • Correction reason categorization (transcription error, calculation error, etc.)
Common nonconformities
  • Corrections that completely obscure the original entry
  • Corrections without a documented reason or with only generic 'error' notation
  • Multiple corrections by the same person on the same record without explanation
  • Electronic records modified without audit trail capture
  • Backdated corrections discovered during batch record review
  • Electronic system allows deletion of original entries without preserving audit trail history
Auditor tip

Complements clause 6.15 with specific correction requirements. The four elements of a compliant correction are: (1) original entry remains readable (single-line strikethrough for paper), (2) corrected value clearly written, (3) date of correction, (4) signature/initials of person making correction, and (5) reason for correction. The reason must be specific — 'error' alone is insufficient; it should explain what was wrong (e.g., 'transcription error — correct value per equipment readout'). For electronic systems, the audit trail must capture all five elements automatically.

6.17 Are written specifications established for raw materials, intermediates, APIs, and packaging materials?
Objective evidence
  • Approved specifications for all raw materials with acceptance criteria
  • Intermediate specifications where applicable (e.g., isolated intermediates)
  • API finished product specification with identity, assay, impurities, and physical tests
  • Packaging material specifications (containers, closures, labels)
  • Process aid specifications (filter media, activated carbon, chromatography resins)
  • In-process control limits documented in master batch records
  • Specification change history linked to change control records
  • Cross-reference between specifications and analytical test methods
Common nonconformities
  • Materials in use without approved specifications
  • Specifications that lack quantitative acceptance criteria (e.g., 'meets requirements')
  • No specifications for process aids that contact the product stream
  • Specifications not updated to reflect process or regulatory changes
  • In-process controls without documented acceptance criteria
  • API specification missing impurity limits or residual solvent limits
Auditor tip

Specifications are the quality benchmarks against which materials and products are evaluated. Every material that contacts or could impact the API must have written specifications with defined acceptance criteria. This includes obvious items (raw materials, APIs) and less obvious ones (process aids, filter gaskets, lubricants). In-process controls need documented acceptance criteria even if they are not formal specifications. Auditors should verify that specifications are current, approved by QA, and referenced in batch records and test methods.

6.18 Where electronic signatures are used, are they validated and controlled to be equivalent to handwritten signatures?
Objective evidence
  • Electronic signature system validation documentation (IQ/OQ/PQ)
  • Written policy on electronic signature use and equivalence to handwritten
  • User access control procedures (unique IDs, password complexity, lockout)
  • Audit trail configuration showing capture of all record modifications
  • Periodic user access review and deactivation of departed personnel
  • System backup and disaster recovery procedures with verified restoration
  • Compliance assessment against 21 CFR Part 11 or EU Annex 11 as applicable
  • Training records for electronic signature system users
Common nonconformities
  • Electronic signatures used without system validation documentation
  • Shared login credentials — multiple users sharing one account
  • No audit trail enabled on electronic systems used for GMP records
  • Electronic records not protected against modification after signing
  • No 21 CFR Part 11 or EU Annex 11 compliance assessment for e-signature systems
  • Departed employee accounts still active in the system
Auditor tip

If an API manufacturer uses electronic signatures (e-signatures), the system must be validated per applicable regulations (FDA 21 CFR Part 11, EU Annex 11). The e-signature must be legally equivalent to a wet signature, which requires: unique to the signer, under sole control of the signer, linked to the record such that any change invalidates it. The underlying electronic records must have audit trails, be tamper-resistant, and remain retrievable for the full retention period. System validation should cover IQ/OQ/PQ with emphasis on security controls.

6.20 Are equipment cleaning and use logs maintained showing date, time, product, batch, and the person responsible?
Objective evidence
  • Equipment logbooks with chronological entries (product, batch, dates, operator)
  • Clean/dirty status tags or labels on equipment
  • Equipment cleaning SOP referenced in logbook entries
  • Cleaning verification records (visual inspection, analytical testing)
  • Maintenance records integrated or cross-referenced in equipment log
  • Electronic equipment management system records if applicable
  • Equipment use records for dedicated equipment (in batch record or separate)
  • Campaign length tracking for dedicated equipment
Common nonconformities
  • Equipment logbooks with missing entries or gaps in chronological sequence
  • No clean status indication before a new product campaign begins
  • Cleaning not recorded between different products on shared equipment
  • Logbook entries that cannot be attributed to a specific person
  • No cross-reference between equipment log and batch record
  • Equipment cleaning status not verified before starting a new batch on shared equipment
Auditor tip

Equipment logbooks (or electronic equivalents) must maintain a complete chronological record of what was processed, when it was cleaned, and by whom. This supports product-to-product cross-contamination risk assessment and cleaning validation. For non-dedicated (shared) equipment, the logbook is critical for traceability. For dedicated equipment, the logbook may be integrated into the batch record. Auditors should review equipment logbooks for completeness and verify that cleaning status is clearly indicated before each use.

6.30 Are records maintained for the receipt and disposition of all raw materials?
Objective evidence
  • Incoming material receiving records with all required data fields
  • Internal lot numbering system description and assignment procedure
  • CoA from supplier cross-referenced to receiving record
  • Quarantine and release labels with internal lot numbers
  • Material traceability from receiving record to batch record usage
  • Disposition records (accepted, rejected, returned) with QA approval
  • Supplier lot number recorded and linked to internal tracking
  • Electronic inventory management records if applicable
Common nonconformities
  • Raw materials in production area without receiving records or internal lot numbers
  • Supplier lot numbers not recorded — traceability to supplier broken
  • No disposition status recorded — unclear whether material was approved for use
  • Multiple internal lot numbers assigned to same supplier shipment without justification
  • Missing CoA for received materials that require certificate
  • Receiving records lacking container number or quantity verification against shipping documents
Auditor tip

Complete receiving records establish the traceability chain from supplier through to finished API. Each receipt must create a documented record with sufficient detail to trace back to the supplier's manufacturing batch. The assigned internal lot number is the link between the supplier material and internal production records. Auditors should trace a raw material from receiving record through to its use in a specific batch to verify end-to-end traceability.

6.31 Do raw-material records link supplier, receipt, test results, and onward use for full traceability?
Objective evidence
  • Test results for incoming materials linked to receiving record lot number
  • Supplier CoAs filed with or cross-referenced to batch production records
  • Re-evaluation test results for materials held beyond original evaluation period
  • Material usage records showing which batches consumed which material lots
  • Packaging material test or inspection records
  • Label material records including supplier, lot, and inspection results
  • Traceability matrix showing material lot to batch number linkage
  • QA review and disposition documentation for each material lot
Common nonconformities
  • Supplier CoAs accepted without being linked to the batch record
  • No re-evaluation testing performed for materials held beyond their evaluation period
  • Test results for incoming materials cannot be traced to specific supplier lots
  • Packaging or labelling materials used without inspection or testing records
  • Gap in traceability chain — material lot to finished batch linkage missing
  • Re-evaluation periods not defined for stored materials — expired materials used without retesting
Auditor tip

Extends 6.30 by requiring that test results and CoAs are formally linked to the receiving record and ultimately to the batch record that consumed the material. When supplier CoAs are accepted in lieu of full testing (per clause 7.31), those CoAs must be filed with or cross-referenced to the batch record — not simply kept in a separate supplier file. Re-evaluation records for materials held beyond their re-evaluation date must also be maintained. This ensures a complete quality dossier for each batch.

6.40 Are master production instructions prepared, independently checked, dated, and signed?
Objective evidence
  • Master Production Instructions with dual signatures (author + QA reviewer)
  • Complete raw material list with quantities or ratios and units
  • Equipment identification (specific equipment or equipment type/size)
  • Step-by-step process instructions with critical parameters specified
  • In-process control points identified with acceptance criteria
  • Yield calculation instructions with expected range
  • Storage and handling instructions for intermediates
  • MPI version history linked to change control records
Common nonconformities
  • Master Production Instructions without QA independent review signature
  • MPI with vague instructions (e.g., 'add solvent as needed' without quantity range)
  • Raw material quantities expressed without units or tolerances
  • No identification of critical process parameters or in-process controls
  • MPI modified without change control documentation
  • Operators consulting handwritten notes instead of the MPI during production
Auditor tip

Master Production Instructions (MPIs), also called Master Batch Records, are the template from which all individual batch records are generated. They must be prepared by a qualified person (typically production) and independently verified by QA. 'Independently checked' means a separate person reviews the MPI for accuracy and completeness — not the same person who wrote it. The MPI must be sufficiently detailed that a trained operator can execute the process without needing to consult additional undocumented instructions. Every change to an MPI must go through change control.

6.41 Do master production instructions contain the required content (product, formula, steps, controls, expected yields)?
Objective evidence
  • MPI containing all 7 required content elements
  • Expected yield range with defined investigation trigger (e.g., <90% or >110%)
  • Intermediate storage instructions specifying conditions, container type, and maximum hold time
  • Sampling points clearly identified in process flow with reference to test methods
  • Parameter ranges for critical steps (temperature, pH, time, agitation speed)
  • Cross-references to relevant SOPs for ancillary procedures
  • Precautions for hazardous materials or operations noted at relevant steps
  • Reprocessing instructions included where reprocessing has been validated
Common nonconformities
  • MPI missing expected yield range — no basis for yield deviation investigation
  • No storage instructions for intermediates held between process steps
  • Critical process parameters specified as exact values without allowable ranges
  • Sampling points not identified — operators decide when to sample
  • No precautions noted for exothermic reactions or hazardous reagent additions
  • Reprocessing instructions referenced but no validated reprocessing procedure available
Auditor tip

This clause specifies the minimum content requirements for MPIs. Auditors should use this as a checklist when reviewing actual MPIs. The seven required elements ensure that the MPI is a complete, self-contained instruction set. Expected yield ranges are particularly important — they serve as the basis for yield deviation investigations per clause 8.12. Storage instructions for intermediates prevent degradation during hold periods. Reprocessing instructions, where included, must be validated per Section 14.

6.50 Is a batch production record prepared for each batch of intermediate and API from the approved master instruction?
Objective evidence
  • Batch production records with reference to specific MPI version number
  • Pre-issuance verification sign-off on batch record (QA or document control)
  • Legibility check documentation (no missing pages, all fields present)
  • Cross-reference between batch record header and current MPI revision
  • Batch numbering system procedure and unique batch number assignment
  • Electronic batch record system validation if automated generation is used
  • Reconciliation of issued batch records (controlled issuance and return)
  • Blank batch record template showing all data entry fields
Common nonconformities
  • Batch record generated from superseded MPI version
  • No pre-issuance verification — batch records issued without checking correctness
  • Missing pages discovered during batch record review after production
  • Batch records with pre-filled data (entries made before production activities)
  • No unique batch number assignment system — duplicate numbers possible
  • Batch records issued to production floor without document control verification of current MPI version
Auditor tip

Each batch must have its own batch production record (BPR) generated from the current, approved Master Production Instruction. Before issuance to production, someone (typically QA or document control) must verify that: (1) the BPR was generated from the correct MPI version, (2) the reproduction is legible and accurate (no missing pages, no cut-off text), and (3) all blank fields are present for completion during production. This pre-issuance check prevents use of obsolete instructions. For electronic batch records, system controls should enforce version currency automatically.

6.51 Does each batch production record carry a unique batch number and capture the required production data?
Objective evidence
  • Completed batch records with unique batch numbers per numbering procedure
  • Date AND time entries at each significant production step
  • Equipment identification numbers recorded at each relevant step
  • Raw material lot numbers and actual quantities used recorded per step
  • In-process control results recorded at designated sampling points
  • Actual yield calculations at intermediate isolation and final API
  • Operator and verifier signatures at critical steps
  • Sequential step completion demonstrating process flow adherence
Common nonconformities
  • Batch number not unique — same number used for different batches or products
  • Time entries missing from significant steps (only dates recorded)
  • Equipment not identified by specific unit number
  • Raw material lot numbers not recorded at point of use
  • Yield calculations missing at intermediate stages
  • Only operator signature without verifier at critical steps
Auditor tip

This clause defines what data must be captured in every BPR during production execution. The batch number must be unique across the site to prevent confusion. Every significant step must have both a date and time recorded (not just date). Major equipment must be identified by its specific unit number (not just type). Raw material lot numbers establish forward/backward traceability. Actual yields documented at appropriate phases enable comparison to expected ranges. Signatures at critical steps confirm both the operator and supervisor/verifier witnessed the operation.

6.52 Are batch record entries made contemporaneously, at the time each activity is performed?
Objective evidence
  • Batch records with entries demonstrating contemporaneous recording (logical time sequence)
  • No evidence of transcription from informal notes (no scratch paper culture)
  • Deviation entries on batch record with cross-reference to deviation report numbers
  • QA investigation and approval documentation for all batch record deviations
  • SOP prohibiting transcription from unofficial notes to batch records
  • GDP training records emphasizing contemporaneous recording requirements
  • Batch record review checklist verifying logical time sequence of entries
  • Electronic systems with time-stamped entries preventing backdating
Common nonconformities
  • Entries in batch record all in same handwriting and ink as if written in one session
  • Time entries out of logical sequence (later step recorded before earlier step)
  • Scratch paper, sticky notes, or unofficial notebooks found in production area
  • Deviations not documented in batch record — discovered only during QA review
  • Batch released before deviation investigation was complete
  • Identical handwriting for entries spanning multiple shifts
Auditor tip

This is the contemporaneous recording requirement — entries must be made 'at the time the specific activity is performed.' Copying data from sticky notes, personal notebooks, or scratch paper to the batch record after the fact is a significant data integrity violation. Deviations from the MPI must be captured in real-time on the batch record (not hidden or documented separately without cross-reference). QA must investigate and approve all deviations before the batch can be released. This clause is among the most frequently cited in FDA warning letters to API manufacturers.

6.53 Are critical deviations investigated, with the investigation extended to other potentially affected batches?
Objective evidence
  • Deviation reports with documented investigation findings and root cause
  • Impact assessment extending to associated batches (same equipment, same material lot)
  • CAPA records linked to deviation investigations
  • Batch record entries cross-referencing deviation report numbers
  • QA review and approval of deviation investigations before batch release
  • Deviation trending reports identifying recurring issues
  • Timeline evidence showing investigation completed before batch disposition
  • Associated batch review documentation (other batches evaluated and cleared)
Common nonconformities
  • Deviations closed without root cause investigation
  • No impact assessment on associated batches
  • Batch released before deviation investigation was completed
  • Recurring deviations without effective CAPA implementation
  • Deviation reports without documented conclusions
  • No deviation trending — individual events investigated in isolation
Auditor tip

Goes beyond simply documenting deviations — requires investigation with documented conclusions, root cause analysis, and evaluation of impact on associated batches. 'Associated batches' is a critical concept: if a deviation involves contaminated equipment or a faulty instrument, all batches processed on that equipment or measured by that instrument must be evaluated. CAPA must be implemented to prevent recurrence. Auditors should verify that deviation investigations are thorough, timely, and that associated batch impact assessments are documented.

6.60 Do laboratory control records capture the complete data derived from all testing, including raw data?
Objective evidence
  • Laboratory notebooks or worksheets with complete raw data for each test
  • Chromatograms, spectra, and instrument printouts attached to or filed with records
  • Calculations documented with formulas, input values, and results
  • Results compared against specification acceptance criteria with pass/fail determination
  • OOS investigation triggered for any result outside acceptance criteria
  • Laboratory records included in batch record review package
  • Analyst signature and date on all test records
  • Second-person review of calculations and data transcription
Common nonconformities
  • Only final results reported — raw data not retained or not traceable
  • OOS results not investigated — retested without formal investigation
  • Calculations without documented formulas or input values
  • Lab records not included in batch record review
  • Analyst entries not signed or dated
  • Raw data (chromatograms, printouts) not attached or cross-referenced to lab record
Auditor tip

Laboratory records must capture the complete analytical dataset — not just pass/fail results. This includes raw data (chromatograms, spectra, weighing printouts), calculations, and final results compared against specifications. Lab records are integral to the batch record review process (clause 6.71). Any result outside acceptance criteria must trigger a formal OOS investigation per a written procedure (clause 11.15). Auditors should review lab records for completeness of raw data, proper calculation documentation, and timely OOS trigger recognition.

6.61 Do laboratory control records contain the required content (sample description, method, results, who, when)?
Objective evidence
  • Lab records containing all 8 required content elements
  • Sample description with batch number, date of sampling, and test request reference
  • Test method reference (SOP number and version) for each test performed
  • Sample weight/volume recorded with balance/pipette ID for traceability
  • Reference standard preparation logs with purity, expiry, and lot number
  • Raw data (chromatograms, spectra) attached or cross-referenced by run ID
  • Calculation worksheets showing formulas, conversion factors, and stepwise calculations
  • Dual signatures (analyst + reviewer) with dates on each test record
Common nonconformities
  • Lab records missing one or more of the 8 required elements
  • No second-person review signature on analytical data
  • Reference standard information not recorded (purity, lot, source)
  • Raw instrument data not retained or not linked to lab record
  • Sample weights not recorded — unable to verify calculation inputs
  • Test method reference missing or referencing an obsolete version
Auditor tip

This clause provides the definitive checklist for laboratory record content — 8 mandatory elements. Auditors should use this as a point-by-point checklist when reviewing lab records. Element (5) — complete record of all raw data — is the most frequently deficient, especially when instrument data is stored electronically but not linked to the lab record. Element (8) requires two signatures: the analyst who performed the test AND a second person who reviewed the data. This dual-signature requirement is a key data integrity control.

6.70 Are there written procedures for the review and approval of batch production and laboratory control records before release?
Objective evidence
  • Batch record review SOP defining scope, responsibilities, and checklist
  • QA batch record review checklist used for each batch
  • Evidence of QA signature on batch record before release
  • Review timeline showing review completed before distribution date
  • Laboratory record review integrated into batch review process
  • Packaging and labelling records included in review scope
  • Review findings documented and resolved before release
  • Training records for batch record reviewers
Common nonconformities
  • No written batch record review procedure
  • Batches distributed before QA review was completed
  • Review checklist not used — review is informal or undocumented
  • Laboratory records not included in batch record review
  • Packaging and labelling records excluded from review scope
  • Review findings not documented or tracked to resolution
Auditor tip

ICH Q7 §6.70 verbatim establishes two load-bearing obligations: (1) written procedures for review and approval of batch production AND laboratory control records (including packaging/labelling) determining compliance with specifications before release/distribution, and (2) QA review and approval of critical-step records before API batch release. The standard does not prescribe a review checklist, reviewer-qualification framework, or specific approval mechanism — those are at the organization's discretion. The 'critical process steps' qualifier means the QA review obligation specifically attaches to critical steps, not necessarily all steps.

6.71 Are batch records reviewed and approved by the quality unit before an API batch is released?
Objective evidence
  • QA signature on batch record confirming review completion before release date
  • All in-process and release test results reviewed and meeting specifications
  • Confirmation that current MPI version was used (version number verified)
  • All deviations documented, investigated, and dispositioned before release
  • Release certificate or disposition form signed by authorized QA person
  • Review of environmental monitoring data if applicable
  • Verification that all critical step signatures are present
  • Reconciliation of materials and labels
Common nonconformities
  • Batch released with incomplete testing (pending results at time of release)
  • Deviation investigation not completed before batch was distributed
  • No QA release signature on the batch record
  • MPI version used in production does not match current approved version
  • Missing critical step verification signatures discovered during review
  • In-process control results not evaluated against acceptance criteria during batch record review
Auditor tip

The review must confirm three things: (1) all required testing is complete and results meet specifications, (2) the batch was manufactured according to the current MPI, and (3) all deviations have been investigated and resolved (or a justified rationale exists for releasing with known deviations). 'Resolved' does not necessarily mean the CAPA is closed — it means the impact assessment is complete and the batch has been evaluated as safe for distribution. No partial releases without full review of completed steps.

6.72 Are deviations investigated and resolved before the affected batch is released?
Objective evidence
  • Deviation investigation reports completed before batch release dates
  • Impact assessment documenting evaluation of deviation on batch quality
  • Corrective actions implemented for critical deviations before release
  • Batch record entries cross-referencing deviation investigation numbers
  • QA disposition decision documented on batch record
  • CAPA tracking for preventive actions that extend beyond batch release
  • Timeline evidence: investigation completion date before distribution date
  • Risk assessment for non-critical deviations with justified release decision
Common nonconformities
  • Batch released with open deviation investigations
  • Critical deviations released without corrective action implementation
  • No impact assessment documented for deviations
  • Batch record missing deviation cross-references
  • Disposition decision not documented by QA
  • Minor deviations released without documented risk assessment justifying no corrective action
Auditor tip

This clause makes clear that deviation investigation is a prerequisite for batch release — not something that can be completed retroactively. For critical deviations, both the investigation AND the corrective actions must be implemented before release. For minor deviations, the investigation must be complete but CAPA may be tracked separately as long as the impact assessment demonstrates no risk to the batch. The batch record must contain the complete audit trail: what deviated, the investigation reference number, and the final disposition decision.

6.73 When a batch fails, is the investigation extended to identify other potentially affected batches?
Objective evidence
  • Failed batch investigation report with extended scope to associated batches
  • Associated batch identification methodology documented
  • Equipment use logs reviewed to identify other batches on same equipment
  • Raw material lot traceability used to identify other batches using same material
  • Disposition decisions for all associated batches (released, quarantined, rejected)
  • Root cause analysis addressing systemic factors, not just the specific batch
  • Cross-product impact assessment for shared equipment
  • Retrospective review of recent batches on same production line
Common nonconformities
  • Failed batch investigated in isolation — no associated batch review performed
  • Associated batches already distributed without evaluation
  • Investigation limited to the failing test only — other quality attributes not evaluated
  • No cross-product impact assessment despite shared equipment
  • Root cause not identified but batch failure closed anyway
  • Operator and environmental factors not evaluated during associated batch impact assessment
Auditor tip

When a batch fails, the investigation scope must extend beyond the failed batch to identify all potentially affected batches. 'Associated' means batches that share common factors with the failure: same raw material lot, same equipment, same operator, same environmental conditions, same time period. The investigation must explicitly evaluate whether other products manufactured on shared equipment could be affected. This clause drives thorough root cause analysis and prevents the release of compromised batches that happened to pass specification by chance.

Each item shows its evidence, common nonconformities and auditor tips. The clause index has the PDF of all 350 items, formatted for a clipboard.