Independent and not affiliated with the FDA, MHRA, ISPE, PDA, or any agency. Get the appgoutham@madhadi.com
madhadi.comData Integrity & GxP Quality
Browse all topics → Articles Templates & Procedures Learning paths GlossaryScenariosToolsRegulatory ReferencesLearning PathsTopics About Start here
Checklist Plug-and-play starting point Quality Assurance

Checklist: Pre-Issue GxP Document Writing Quality Review

A structured review checklist an independent reviewer runs on a protocol, report, deviation, or investigation before it is routed for final approval, checking objectivity, traceability, root cause and CAPA language, and multi-author terminology consistency.

Document type: Checklist

Read and copy the template below into your own quality system. It is a generic starting point for your own internal use, provided as is, with no warranty; see the Terms and License. Adopting it does not by itself create compliance.

This is a working checklist. It runs on the written content of a protocol, report, deviation record, investigation narrative, or CAPA record, checking writing quality and defensibility rather than document-control mechanics. Mark each item Yes, No, or N/A, and record the evidence or location. Any No is a defect to fix before the document is routed to its final approver. A filled specimen follows. This checklist complements, and does not replace, checklist: SOP drafting and readiness, which checks SOP structure specifically.

Document control header

FieldEntry
Document under review<<FILL: document ID and title>>
Document typeProtocol / Report / Deviation / Investigation / CAPA record (circle one)
Draft version<<FILL>>
Author<<FILL: role and name>>
Checklist completed by<<FILL: role and name, independent of the author>>
Date<<FILL: date>>

How to use this checklist

Run this after the author’s own self-review and before the document reaches its final approver. The reviewer should not be the document’s author. Where the document has multiple authors or spans multiple sites, run section D in full; it is the section most often skipped and the one most likely to catch a real defect.

A. Objectivity and tone

#CheckYes / No / N/AEvidence or note
A1No minimizing word (minor, slight, just, only, isolated) appears ahead of the evidence that would support it.<<FILL>>
A2No conclusion, criticality rating, or impact statement (including “no impact”) appears before the evidence supporting it.<<FILL>>
A3No speculation is written as fact; any hypothesis not yet confirmed is labeled as a hypothesis.<<FILL>>
A4In narrative or event text, the actor is named (active voice); passive voice is used only where a procedure step’s actor is already defined by role.<<FILL>>
A5No vague, unfalsifiable phrase (“appropriate action was taken,” “the issue was addressed,” “as needed”) remains without the specific action, trigger, or frequency stated.<<FILL>>

B. Completeness and traceability

#CheckYes / No / N/AEvidence or note
B1Every field has an entry; N/A is used, initialed, and justified where a field genuinely does not apply.<<FILL>>
B2Every factual claim traces to a named source (a log, a system, a record, a person).<<FILL>>
B3Every acceptance result ties back to the specific pre-defined criterion it is judged against, not a restated summary of the criterion.<<FILL>>
B4Numbers, limits, durations, and identifiers are stated where the draft currently uses an adjective (“high,” “long,” “significant”).<<FILL>>

C. Root cause and CAPA language (where applicable)

#CheckYes / No / N/AEvidence or note
C1The root cause statement explains why the failure was possible, not only what the individual did.<<FILL>>
C2Every claim in the root cause statement traces to evidence already shown in the timeline or investigation section.<<FILL>>
C3The stated corrective action would prevent the same failure from recurring the same way (the recurrence test).<<FILL>>
C4At least one candidate cause considered and ruled out appears with its evidence, not only the accepted cause.<<FILL>>
C5The CAPA effectiveness check names a metric, a baseline, an observation window, and a reopening trigger, not a restated action.<<FILL>>

D. Multi-author, multi-site, and terminology consistency

#CheckYes / No / N/AEvidence or note
D1The same concept (system, role, event) is named the same way in every section, regardless of which author drafted it.<<FILL>>
D2Tense and voice are consistent with the document’s actual state (a report describing completed work stays in the past tense throughout).<<FILL>>
D3No reference to another document uses a nickname or an unversioned phrase (“the current SOP,” “the attached form”) instead of a specific number and version.<<FILL>>
D4No placeholder text, draft marking, or “TBD” remains in a document about to be issued as effective.<<FILL>>
D5Where the document will be translated or read across sites, sentences carry one idea each and every acronym is defined at first use.<<FILL>>

E. Conclusion integrity

#CheckYes / No / N/AEvidence or note
E1Every claim appearing in the conclusion has already appeared, with its supporting evidence, earlier in the document.<<FILL>>
E2The conclusion does not contradict data presented earlier in the same document.<<FILL>>
E3Reading from the conclusion backward through the evidence, every claim can be traced to a specific reference.<<FILL>>

Disposition

FieldEntry
Total Yes<<FILL>>
Open defects (any No)<<FILL: count, with item numbers>>
DecisionReady to route / Return to author (circle one)
Completed by (name, signature, date)<<FILL>>

Any No must be closed or formally dispositioned before the document is routed for final approval.


Filled specimen

The following shows the checklist completed for an example draft investigation narrative into a temperature excursion during intermediate storage. The company, IDs, and notes are illustrative. Only the items that drove a finding are shown filled; the rest were marked Yes.

FieldEntry
Document under reviewINV-2026-0317, Investigation into cold-room temperature excursion, Lot 26-0420
Document typeInvestigation
Draft version0.2 draft
AuthorInvestigation owner, T. Nolan
Checklist completed byQA Reviewer, S. Kim (independent)
Date25 August 2026
#CheckResultEvidence or note
A2No impact statement before evidenceNoDraft opened with “the excursion had no impact on the batch” in the deviation statement, before the investigation. Moved to the impact assessment section, after the stability data citation.
C4Ruled-out cause shownNoDraft showed only the accepted cause (compressor failure). Added the two ruled-out candidates (setpoint drift, sensor error) with their evidence.
C5Effectiveness check measurableNoDraft read “alarm escalation reviewed, issue resolved.” Rewritten with a metric (recurrence count), baseline, six-month window, and reopening trigger.
D3Versioned referencesNoDraft referenced “the cold room SOP” with no number. Corrected to SOP-QA-071 v3.0.

Disposition: 4 open defects at first pass, all closed in version 0.3 draft. Decision, ready to route. Completed by S. Kim, signed, 25 August 2026.

In this specimen the checklist caught an impact statement written before the evidence, a one-sided root cause presentation, an unmeasurable effectiveness check, and a dangling reference, the four defects most likely to become inspection findings in an investigation narrative.

Common inspection findings this checklist prevents

  • “No impact” or a criticality conclusion written into a record before the investigation supporting it exists.
  • An investigation that shows only the accepted cause, with no evidence that alternatives were considered.
  • A CAPA effectiveness check that restates the action rather than defining a measurable outcome.
  • The same system or event named differently across sections of the same document.
  • A dangling reference to “the current SOP” or “the attached form” with no document number or version.
  • Placeholder text or a future-tense description of completed work left in an issued document.

How to adapt this checklist

  1. Add or remove rows so section headings match the document types your quality system produces most often.
  2. Keep section E (conclusion integrity) as a mandatory final pass; reading from the conclusion backward is the single fastest way to catch an unsupported claim.
  3. For multi-site or multi-language documents, treat section D as a hard gate rather than an advisory check.
  4. Attach the completed checklist to the document’s review record so the readiness evidence travels with it through approval.
  5. Pair this checklist with your risk-based investigation depth and root cause method; it checks how the conclusion is written, not whether the underlying analysis was deep enough.
Use madhadi.com as an app Full screen, works offline, one tap from your home screen.