When Root Cause Analysis Goes Wrong
Root Cause Analysis (RCA) is one of the most important parts of a GMP investigation. It is also one of the easiest parts to weaken.
The problem is usually not that the investigation team ignores RCA. Most records include a root cause section, a selected tool, a conclusion, and a CAPA. The problem is that the reasoning may not actually support the conclusion.
A root cause analysis can look complete on paper while still failing to explain what happened, why it happened, why existing controls did not prevent or detect it, and what needs to change to reduce recurrence risk.
As discussed in Pharmaceutical Investigations & CAPA, the purpose of an investigation is not only to close the deviation. It is to reach a supported conclusion that can guide an appropriate response. When RCA goes wrong, the CAPA often becomes misaligned, the effectiveness check becomes weak, and closure becomes harder to defend.
The Root Cause Is Selected Too Early
One common problem is selecting the root cause before the evidence is complete.
The investigation may begin with a likely explanation, such as operator error, analyst error, equipment issue, procedure not followed, or isolated documentation mistake. These may be valid hypotheses. They become a problem when the investigation treats the early explanation as the answer.
Once the conclusion is selected too early, the rest of the investigation may become a search for supporting evidence. Conflicting information may receive less attention. Alternative causes may be listed only briefly. The final report may appear organized, but the logic is already narrowed.
This connects closely to Cognitive Bias in GMP Investigations. Early assumptions, prior experience, and familiar explanations can influence what evidence is gathered and how it is interpreted.
A stronger RCA keeps the early explanation as a hyopothesis until the evidence supports it.
The Investigation Stops at the Symptom
Another common weakness is treating the visible problem as the root cause.
Examples include:
“Procedure not followed.”
“Operator missed a step.”
“Analyst error.”
“Equipment malfunction.”
“Incorrect entry.”
“Sample mix-up.”
“Line clearance failure.”
These statements may describe what happened, but they often do not explain why it happened.
If an operator missed a step, the investigation should ask why the step was missed. Was the instruction unclear? Was the step poorly positioned in the batch record? Was there an interruption? Was the task memory-dependent? Was the verification step ineffective? Was the process designed in a way that made the error likely?
If equipment malfunctioned, the investigation should ask why. Was maintenance overdue? Was the failure mode known? Were alarms adequate? Was the equipment operated within expected conditions? Was the problem detectable earlier?
Stopping at the symptom may lead to a CAPA that addresses the event without addressing the condition that allowed it.
The Causal Chain Is Not Clear
A defensible RCA should show the connection between the event, evidence, cause, and conclusion.
The reviewer should be able to follow the logic:
What was expected?
What actually happened?
What evidence was reviewed?
What possible causes were considered?
What causes were ruled out?
What cause remains supported?
How does that cause explain the event?
How does that cause connect to impact and CAPA?
When the causal chain is missing, the root cause may feel like a conclusion inserted into the record rather than a conclusion reached through investigation.
As explained in Writing Defensible Investigation Reports, strong reports show how conclusions were reached. They do not rely only on final statements.
Alternative Causes Are Not Meaningfully Ruled Out
Many investigations include a list of possible causes. That is useful only if the list is evaluated.
A weak RCA may state that equipment, material, method, personnel, environment, or procedure were considered, but provide little explanation for why each was accepted or ruled out. In some cases, alternate causes are dismissed because there is “no evidence”, without explaining what evidence would have been expected or what review was performed.
Ruling out a cause requires documented logic. The investigation should show what was checked, what was found, and why the finding supports exclusion.
This does not mean every possible cause needs a long discussion. The depth should match the risk, complexity, and available evidence. But when an alternate cause is plausible, it should not be dismissed without rationale.
Human Error Becomes the Default Answer
Human error is one of the most common places where RCA goes wrong.
A person may have made the final action that triggered or failed to prevent the deviation. That does not mean the person is the root cause. Human performance is influenced by procedure design, training, competency, workload, supervision, tools, equipment, handoffs, environment, and controls.
Investigations should avoid stopping at who made the mistake. They should evaluate what allowed the mistake to occur and whether the system adequately supported correct performance.
If the RCA ends at “operator error” and the CAPA is retraining, the record may be difficult to defend unless the investigation shows a specific training, knowledge, skill, or competency gap.
The Root Cause Does Not Match the Evidence
Sometimes the investigation includes evidence that does not clearly support the selected root cause.
For example, the root cause may be assigned to training even though the employee was trained, qualified, experienced, and following the same process used successfully before. Or the cause may be assigned to equipment even though the equipment review found no abnormal condition. Or the event may be called isolated even though similar events are documented.
When the evidence and conclusion do not match, the issue should be reconciled in the investigation. The record should explain why the conclusion is still supported or whether the root cause needs to be revised.
Unreconciled conflict is a major review concern because it shows that the final conclusion may not reflect the full evidence set.
The Root Cause Is Too Vague to Support CAPA
A root cause must be specific enough to guide action.
Statements such as “lack of attention”, “procedure not followed”, “communication failure”, “inadequate process”, or “human error” may be too broad unless the investigation explains the specific condition behind them.
A vague root cause creates vague CAPA.
If the root cause is “communication failure”, the CAPA may become “communicate requirement to team”. But what failed? Was the handoff undefined? Was the escalation path unclear? Was information missing from the batch record? Was responsibility unclear between departments? Was the system notification not visible?
CAPA depends on the quality of the causal conclusion. The action should address the supported cause and recurrence risk. When the root cause is vague, CAPA alignment becomes difficult.
The RCA Tool Is Used Mechanically
RCA tools can help structure thinking, but they do not guarantee a good investigation.
A 5-Why can become a chain of unsupported assumptions. A Fishbone diagram can list many possibilities without evidence. A Fault Tree can look sophisticated while still depending on weak inputs.
The tool is only useful if it improves the reasoning.
A strong RCA tool should help the team ask better questions, test possible causes, organize evidence, and explain the conclusion. It should not become a formality added to the record after the answer is already selected.
This is why method selection matters. Simple deviations may not need a complex tool. Complex or repeat deviations may need more structured analysis. The method should match the event, risk, and evidence.
The RCA Does Not Explain Control Failure
A good RCA should explain not only why the event occurred, but also why existing controls did not prevent, detect, or limit it.
If a second-person check existed, why did it not catch the issue? If an alarm existed, why was it not effective? If QA review was expected to identify the problem, why was it missed? If a procedure controlled the process, why did the procedure not support correct execution?
This does not mean every control failed. Some controls may have worked as intended. But where a relevant control did not prevent or detect the event, the investigation should address that logic.
If control failure is not evaluated, CAPA may target the immediate cause while leaving the weak control in place.
QA Review Perspective
When root cause analysis goes wrong, the weakness usually appears as a broken connection.
The event does not connect to the evidence.
The evidence does not connect to the root cause.
The root cause does not connect to CAPA.
The CAPA does not connect to effectiveness.
The closure rationale does not connect to remaining risk.
QA reviewers should look for those connections.
Useful review questions include:
Does the root cause explain the actual event?
Is the conclusion supported by documented evidence?
Were plausible alternate causes evaluated?
Were conflicting facts reconciled?
Does the timeline support the selected cause?
Does the RCA explain why existing controls were ineffective?
Is the root cause specific enough to support CAPA?
Does the CAPA address the supported cause?
Does closure explain why the investigation is complete?
A strong RCA shows how the investigation moved from facts to conclusion. When that logic is clear, the CAPA is easier to justify, the effectiveness check is easier to design, and the final investigation is easier to defend.
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.
