Deviation Writing: Examples & Templates
A GMP deviation record should allow someone unfamiliar with the event to understand what happened, what was affected, what was investigated, and how the final decision was reached.
That sounds straightforward. Yet many deviation records contain all the expected sections while leaving important questions unanswered.
A description may say that an operator made an error without explaining what was observed. An investigation may list records reviewed without documenting what those records showed. A conclusion may state that there was no product impact without explaining why.
These weaknesses often begin with how information is written.
Good deviation writing makes the event understandable, the evidence traceable, and the decisions defensible. The following examples and templates show how to achieve this without unnecessary complexity.
What Good Deviation Writing Should Accomplish
Deviation documentation develops throughout the investigation. The initial report captures what is known when the event is detected. Later entries add evidence, analysis, decisions, and conclusions.
Each stage should distinguish between three things:
Facts: What was observed, recorded, measured, or otherwise established.
Interpretations: What the available evidence suggests or supports.
Decisions: What action or conclusion was reached, and why.
As discussed in Pharmaceutical Investigations & CAPA, a deviation record connects event recognition with investigation, risk assessment, root cause determination, CAPA, and closure.
Good writing makes these connections visible.
Example 1: Writing a Clear Event Description
Consider a batch record discrepancy detected during QA review.
Weak description:
“Operator failed to follow the batch record and missed a verification step.”
This statement immediately assigns responsibility. It does not explain which requirement was involved, what was observed, or whether the required verification was actually omitted or simply not documented.
Stronger description:
“During QA review of Batch B-104, the second-person verification field for Step 8 was found blank. The approved batch record requires independent verification before proceeding to Step 9. The manufacturing sequence had been completed when the missing entry was identified. The batch was placed on hold pending investigation of whether the verification was performed and the potential impact of the missing documented verfication.”
This example establishes the requirement, observed discrepancy, detection point, affected batch, and initial action.
Importantly, it does not assume that the verification was never performed. That question remains open until the investigation establishes what occurred.
The same distinction applies to missing signatures, undocumented checks, incomplete logbooks, and electronic-record discrepancies.
A missing record is evidence of a documentation gap. It does not automatically establish that the underlying activity was not performed.
Example 2: Writing a Manufacturing Deviation
Consider an unexpected equipment stoppage during a manufacturing operation.
Weak description:
“Equipment malfunction occurred during manufacturing. Maintenance fixed the problem and production continued.”
This leaves the reader uncertain about the failure, affected process, product status, and response.
Stronger description:
“At approximately 10:35, Mixer M-12 stopped unexpectedly during the blending operation for Batch B-205. The equipment display showed an unexpected motor-fault alarm. The operation was stopped, the batch remained in the equipment, and Manufacturing notified Engineering and QA. The batch was placed on hold pending assessment of the interruption, actual blending duration, equipment condition, and potential process impact.”
This statement does not claim that the batch was unaffected merely because the equipment was restarted.
It establishes the questions the investigation must answer.
The later investigation can evaluate alarm history, blending parameters, equipment logs, maintenance records, and the interruption’s significance.
Example 3: Writing an Environmental Excursion
Environmental deviations need the same factual discipline.
Weak description:
“Room pressure excursion occurred due to frequent door opening. No product impact.”
Stronger description:
“During routine monitoring, the differential pressure between Rooms 201 and 202 was recorded below the approved operating limit for approximately six minutes. Manufacturing activities were in progress in Room 201 during the recorded interval. Operations notified QA and initiated the defined response procedure. Potential exposure, room activity, pressure-monitoring records, and relevant facility conditions were identified for investigation.”
The stronger version reports the excursion without prematurely assigning cause or dismissing impact.
Door activity may be a possible contributor, but its causal significance requires supporting evidence.
Template 1: Initial GMP Deviation Description
The following template can be adapted to manufacturing, laboratory, facility, documentation, or system-related deviations.
Deviation Title: [Short factual description of the event]
Date and Time: [Occurrence and/or detection time, distinguishing them where necessary]
Process / Location: [Operation, room, line, laboratory, system, or equipment]
Affected Item: [Product, batch, material, sample, record, equipment, or other relevant population]
Expected Condition: [Applicable approved requirement, specification, instruction, or control]
Actual Condition: [What was observed or recorded]
Detection Method: [How and when the discrepancy was identified]
Immediate Actions: [Containment, correction, notification, hold, shutdown, segregation, or other response]
Initial Status: [Status of affected product, process, equipment, or system]
Initial Impact Question: [What potential quality or compliance impact requires evaluation]
This is a framework, not a mandatory form. Information should be proportional to the event and aligned with the site’s approved deviation procedure.
Writing Evidence Reviews That Support Conclusions
Evidence sections often become lists of documents examined.
For example:
“Batch records, equipment logs, SOPs, and training records were reviewed. No abnormalities were found.”
This gives little insight into what each review established.
A stronger evidence section identifies the source, observation, and meaning.
| Evidence Reviewed | Documented Observation | Investigation Interpretation |
|---|---|---|
| Equipment alarm history | Motor-fault alarm recorded at 10:35 | Confirms timing of the stoppage |
| Batch manufacturing record | Blending operation began at 10:10 | Establishes the recorded process sequence |
| Maintenance assessment | Motor overload condition identified | Supports further evaluation of equipment contribution |
| Previous deviation history | Two related stoppages identified within the review period | Raises a recurrence question requiring assessment |
Note: All findings are illustrative examples.
The interpretation column is important. It explains why a piece of evidence matters.
Where evidence is incomplete, unavailable, conflicting, or inconclusive, the record should preserve that limitation.
Template 2: Evidence and Investigation Finding
For each material evidence item, investigators can use the following structure:
Evidence Source: [Record, system data, test result, observation, interview, or assessment]
Review Performed: [What was examined and the period/population reviewed]
Finding: [Specific documented result]
Interpretation: [How the finding supports, contradicts, or limits a hypothesis or conclusion]
Remaining Uncertainty: [What the evidence does not establish]
This format encourages investigators to explain the value of the evidence rather than simply confirm that a review took place.
Writing Impact Assessments Carefully
Statements such as “No impact”, “Batch acceptable”, or “No risk identified” need appropriate context.
A useful impact assessment explains which product, process, or system characteristics were considered and what evidence supports the decision.
Weak statement:
“No product impact because all final testing passed.”
Stronger approach:
“The impact assessment considered the affected process stage, relevant critical process parameters, in-process results, final testing, and the ability to establish that the process remained within the approved operating conditions. The conclusion is based on the documented results of these reviews, including any limitations identified.”
The actual conclusion must then state what the evidence demonstrated.
Passing final testing alone may not establish that every potentially affected quality attribute or process condition was adequately controlled.
Template 3: Writing the Investigation Conclusion
A useful final conclusion can follow this structure:
Event and Scope:
“The investigation evaluated [event] affecting [defined population and scope].
Evidence Summary:
“The material evidence included [key evidence sources and findings].
Root Cause:
“The investigation concluded that [supported cause], based on [specific evidence and causal rationale].”
Impact:
“The impact assessment determined [conclusion] because [supporting evidence and assessment logic].”
CAPA Decision:
“[Actions] were selected to address [supported cause/control weakness], or no formal CAPA was required because [documented rationale].”
Remaining Actions/Closure:
“[Open actions, effectiveness expectations, limitations, or follow-up controls] are documented and managed through [appropriate process].”
The language should be adapted where the root cause remains inconclusive or impact cannot yet be determined. Investigators should never force certainty into a conclusion that the available evidence cannot support.
Common Writing Mistakes
Several recurring habits weaken otherwise useful deviation records.
Assigning blame in the event description.
Write what occurred before stating why it occurred.
Using general statements without supporting detail.
“All records reviewed” should identify the relevant records and meaningful results.
Confusing absence of documentation with absence of activity.
Use “not documented” when the submitted record does not establish whether an activity occurred.
Overstating conclusions.
“All risk eliminated” is generally stronger than most investigations can demonstrate. Use language proportionate to the evidence.
Repeating facts without interpreting them.
Evidence should support a specific question, hypothesis, assessment, or decision.
Treating corrective actions as investigation conclusions.
The record should explain the cause before explaining the action selected to address it.
The distinction between containment, correction, corrective action, and preventive action is discussed further in CAPA Action Types Explained.
Writing Should Evolve with the Investigation
An initial deviation report may reasonably contain unknowns.
For example, the affected product population may initially be uncertain. The equipment failure mechanism may remain under assessment. The historical review may be incomplete.
The record should identify those uncertainties and update the investigation as evidence becomes available.
If the investigation reveals that additional batches, systems, or process conditions may be affected, its scope and impact assessment should reflect the new information.
As explained in Defining Investigation Scope in GMP Deviations, the investigation boundary should follow the event and plausible exposure, with meaningful inclusions and exclusions explained.
A well-maintained deviation record allows the reviewer to distinguish initial observations from subsequent findings without losing the history of how the investigation developed.
Regulatory Perspective
Under 21 CFR 211.192, relevant unexplained discrepancies and failures must be thoroughly investigated, including potentially associated batches, and the investigation must have a written record containing conclusions and follow-up.
The regulation does not prescribe a single deviation-writing template. The record must demonstrate an adequate investigation and provide a traceable basis for the resulting decisions.
Clear writing supports that expectation by showing which facts were established, what evidence was evaluated, how conclusions were reached, and what follow-up was required.
QA Review Perspective
A QA reviewer should be able to read a deviation record without needing the investigator to explain what each statement was intended to mean.
Before approval, useful questions include:
Is the event described factually and specifically?
Are the expected and actual conditions distinguishable?
Are the affected population and immediate response clear?
Does the evidence section explain what each important review established?
Are assumptions and uncertainties distinguishable from confirmed findings?
Are impact and root cause conclusions supported?
Does the CAPA or no-CAPA decision follow logically from the investigation?
Clear writing improves both investigation quality and review efficiency. It also exposes weaknesses while there is still time to address them.
The strongest deviation records are often concise. Their value comes from the accuracy of the facts, relevance of the evidence, clarity of the reasoning, and the strength of the resulting decisions.
Explore more on Investigations & CAPA Excellence
Browse VerethiQ resources on deviation handling, root cause analysis, investigation quality, CAPA design, effectiveness checks, recurrence prevention, and investigation governance.
