Evidence, Interviews, and Process Reality
A manufacturing deviation is reported following an unexpected equipment stoppage. The batch record shows that the operation was completed, the equipment was restarted, and the required in-process checks were satisfactory.
At first, the event appears straightforward.
However, further review reveals that the equipment alarm history does not fully match the recorded sequence. An operator describes an interruption that was not captured in the batch record. Engineering identifies a condition that may have affected equipment performance.
The investigation now has several sources of information, each describing a different part of the event.
This is a common challenge in manufacturing investigations.
The written record provides essential evidence, but understanding what actually happened may also require electronic process data, equipment history, operator interviews, technical assessments, and knowledge of how the process operates under routine conditions.
A strong manufacturing investigation brings these sources together to establish a reliable explanation of the event and its potential consequences.
Manufacturing Investigations Begin with Process Understanding
Manufacturing deviations can arise during dispensing, material transfer, blending, granulation, compression, filling, packaging, cleaning, or other production activities.
Each operation has defined requirements, equipment conditions, operating parameters, and controls.
Before investigating why something went wrong, the investigator needs to understand how the process was intended to operate.
This includes establishing:
the approved manufacturing instructions and applicable process parameters
the manufacturing stage and equipment involved
what was actually observed or recorded
when and how the deviation was detected
what immediate actions were taken
which product, batches, materials, or processes may have been affected
For example, an equipment alarm during blending cannot be evaluated adequately without understanding the required mixing conditions, operating sequence, and significance of the interruption.
The investigation should establish the relevant process requirements before interpreting whether the observed event affected process control.
Within the broader Pharmaceutical Investigations & CAPA framework, this factual understanding provides the foundation for scope, root cause, impact assessment, and CAPA decisions.
Reconstruct What Actually Happened
Manufacturing investigations often involve reconstructing the sequence of events from multiple sources.
The batch record may identify the start and completion of an operation. Equipment data may reveal an interruption. A logbook may describe an intervention. An interview may clarify the sequence of operator actions.
These sources should be compared rather than reviewed independently. The investigator should establish the known sequence of events, identify any gaps, and determine which operations can be supported by objective records.
A useful reconstruction distinguishes between the time an event occurred and the time it was detected. For example, a process failure may have occurred during manufacturing but only been identified during later batch record review. This difference can affect the exposure period, the affected population, and the assessment of prevention or detection controls.
The investigation should also preserve uncertainty. If the exact occurrence time cannot be established, the record should explain the available evidence and any resulting limitations. For practical guidance on documenting these distinctions, see Deviation Writing: Examples & Templates.
Build the Evidence Picture
Manufacturing evidence is rarely contained in a single record.
Depending on the event, relevant sources may include:
Batch manufacturing records: Documented processing steps, times, equipment, material use, in-process checks, interventions, and personnel verification.
Equipment and automation data: Alarm histories, parameter trends, system events, audit trails, recipe information, and operating-status changes.
Maintenance and calibration records: Equipment condition, previous repairs, recurring failures, and relevant technical assessments.
Process monitoring data: Critical process parameters, in-process measurements, environmental conditions, and other monitoring information.
Historical quality records: Previous deviations, CAPAs, change control, equipment history, and comparable manufacturing events.
Personnel interviews and observations: Firsthand descriptions of the activity, unusual conditions, interventions, handoffs, and practical workflow.
Each source should answer a relevant investigation question.
The aim is to establish what the evidence demonstrates and how it relates to the event.
A statement such as “Equipment records reviewed; no issue identified” provides limited value unless the investigation explains which records were reviewed, what they showed, and why those findings are relevant.
Understand the Reality of Manufacturing Execution
Approved procedures define how manufacturing activities should be performed. Actual execution occurs within an operating environment that may include interruptions, equipment constraints, material variability, handoffs, and competing tasks.
These conditions can influence how a deviation develops.
For example, a verification step may be clearly written in an SOP but positioned awkwardly within the actual sequence of work.
An equipment interface may require several manual actions that increase the opportunity for selecting an incorrect setting.
A process may depend on operators recognizing subtle changes in equipment behavior before an alarm occurs.
These conditions should be investigated when they are relevant to the event.
Understanding process reality may require reviewing the operation at the point of execution, observing the normal workflow, examining equipment interfaces, or discussing how tasks are performed during routine and unusual conditions.
Where direct observation is not possible, the investigator should identify what other evidence can establish the relevant operating conditions.
A difference between documented instructions and actual practice should be evaluated carefully. It may indicate a procedural deviation, a design weakness, an informal workaround, or a misunderstanding of the approved process.
The investigation should determine which explanation the evidence supports.
Interview Operators and Subject Matter Experts
Operators and SMEs can provide information that is not fully visible in manufacturing records.
An operator may remember an unusual sound before an equipment stoppage. Engineering may identify a recurring equipment condition. A supervisor may explain how the shift handoff was conducted.
These details can help the investigator identify evidence sources and test possible failure mechanisms.
However, interviews should be conducted objectively.
Questions should encourage people to describe what happened rather than confirm a preferred explanation.
For example:
Leading question: “Did you forget to perform the verification because you were distracted?”
Neutral question: “Can you walk me through how the verification was performed during this operation?”
The second question allows the investigator to understand the activity before deciding whether distraction, procedure design, system conditions, or another factor contributed.
Interview statements should distinguish direct observations from opinions and assumptions.
Where possible, relevant information should be corroborated through records, system data, or other evidence.
The interview process is discussed in greater detail in Interviewing SMEs for GMP Investigations.
Reconcile Conflicting Evidence
Different evidence sources do not always agree.
A batch record may show that an operation was completed without interruption, while equipment data indicate an unexpected stoppage.
An operator may recall performing a check, while the corresponding entry is missing.
A maintenance assessment may identify an equipment fault that does not explain the entire event sequence.
These differences require examination.
The investigator should consider whether the apparent conflict results from different timestamps, incomplete records, system configurations, interpretation of the data, or genuinely inconsistent accounts.
The resolution should be documented.
If the discrepancy cannot be fully reconciled, the investigation should explain the limitation and assess whether it affects root cause, product impact, or the final quality decision.
Unresolved evidence should not be treated as confirmation of the preferred conclusion.
Practical Example: Unexpected Interruption During Blending
Consider a hypothetical deviation involving a blending operation.
The batch manufacturing record shows that blending began at 10:10 and was completed at 10:50.
During routine review, QA identifies an equipment alarm recorded at 10:35.
The investigation gathers the following evidence.
| Evidence Source | Documented Finding | Investigation Significance |
|---|---|---|
| Batch manufacturing record | Blending start and completion times are recorded; no interruption is noted | Establishes the documented sequence but leaves the interruption unexplained |
| Equipment event history | Motor-fault alarm and equipment stoppage recorded from 10:35 to 10:42 | Confirms an interruption requiring further assessment |
| Operator interview | Operator describes acknowledging the alarm and restarting the equipment | Helps reconstruct actions taken, subject to corroboration |
| Maintenance assessment | Motor overload condition identified | Identifies a possible technical contributor requiring causal evaluation |
| In-process testing | Recorded results met the applicable acceptance criteria | Contributes to impact assessment but does not independently resolve every process question |
Note: All findings are illustrative examples.
The investigation now needs to determine whether the equipment resumed the approved operation correctly, whether the required effective blending duration was achieved, and whether the interruption could have affected blend uniformity or another relevant quality attribute.
The investigator may need to examine the control-system logic, actual mixing-time calculation, validated operating conditions, available process data, and the significance of the in-process results.
Engineering must evaluate the overload condition and its relationship to the event.
The operator interview helps explain the sequence, but it does not independently establish the equipment failure mechanism.
If the investigation concludes that the batch was unaffected, it must demonstrate why that conclusion is supported.
Similarly, if the motor overload is identified as the root cause, the causal evidence should show how the overload occurred and whether the condition reflects an isolated fault or a broader equipment weakness.
This example illustrates why manufacturing investigations benefit from combining records, technical data, interviews, and process knowledge.
Connect Process Findings to Impact and CAPA
Manufacturing investigations should distinguish between understanding the equipment or process failure and determining its potential consequences.
A process interruption may have no demonstrated product impact while still revealing a meaningful control weakness.
Conversely, successful equipment recovery does not automatically establish that the affected batch met all relevant process requirements.
The impact assessment should consider the affected manufacturing stage, critical process parameters, product quality attributes, exposure period, and limitations of the available evidence.
Scope may need to extend to other batches, equipment, or processes when a common failure mechanism or shared control creates plausible broader exposure.
For significant events, the response may require additional technical oversight or escalation. The practical management of such cases is explored in Critical Deviations: How to Manage Them.
CAPA should then address the supported cause and relevant control weaknesses.
Replacing a failed component may resolve the immediate problem. If the failure resulted from a recurring design, maintenance, or operating-condition weakness, additional action may be needed.
The appropriate response depends on what the investigation establishes.
Regulatory Perspective
21 CFR 211.100 requires written production and process-control procedures to be followed, with deviations recorded and justified.
21 CFR 211.188 establishes requirements for batch production and control records, including documentation of significant manufacturing steps, equipment, materials, and in-process results.
21 CFR 211.192 requires thorough investigation of applicable unexplained discrepancies and failures, including other potentially associated batches, with written conclusions and follow-up.
FDA’s Process Validation: General Principles and Practices guidance also emphasizes understanding process variation and using manufacturing data to maintain a state of control.
These expectations support investigations that evaluate relevant manufacturing evidence and explain how the observed event affects process control and product quality.
QA Review Perspective
A QA reviewer should be able to follow the manufacturing event from the approved process requirement through actual execution, evidence evaluation, and final decision.
Useful review questions include:
Is the expected manufacturing process clearly established?
Has the actual event sequence been reconstructed from relevant evidence?
Were equipment data, records, and technical assessments evaluated where applicable?
Did operator and SME interviews contribute useful, corroborated information?
Were important differences between records and actual operating conditions addressed?
Does the investigation explain the potential product or process impact?
Were broader equipment, batch, or control exposures considered where relevant?
Are the root cause and CAPA decisions supported by the investigation findings?
Manufacturing investigations are strongest when they reflect how the process actually operated.
A complete batch record provides essential documentation, but its value becomes greater when it is evaluated alongside equipment behavior, process conditions, technical evidence, and firsthand observations.
The goal is a coherent, evidence-supported understanding of the event that allows QA to make defensible decisions and helps the organization improve the reliability of its manufacturing controls.
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.