To lay down a procedure for failure investigation.
This procedure is applicable for investigation of process or system failure having impact on product quality, efficacy and patient safety.
All concerned Department, Quality Assurance
4.1 Failure investigation shall be performed for confirmed OOS, complaints resulting in recall, critical deviation associated with product or process or equipment, repeated similar deviations, ineffective CAPAs.
4.2 Failure investigation and root cause analysis shall be comprised of following steps:
– Description of failure
– Identification of probable root cause
– Evaluation of root cause
– Confirmation of true root cause
– CAPA plan
4.3 Description of failure:
4.3.1 Head of concerned department along with QA head shall initiate the failure investigation, however other department heads like engineering, QC etc. can be part of investigation depending upon nature and type of failure.
4.3.2 Following shall be taken care while defining problems:
– Ensure every concern person/department agrees on material/result/part/object and measurable, observable and specific deviation.
– The team shall be cross functional who shall not get caught in content.
Description of problem shall indicate:
* What is the problem?
* Where did it happen?
* When did it happen?
* Extent of the problem.
4.3.3 The team can use Is/Is Not approach to precise the problem. This approach describes other probable deviations but Is Not cause of failure. This will help in problem description and will reduce scope of investigation. It will help to collect facts, observations that are closely related to occurred deviation.
4.3.4 Below table provides guideline for Is/Is Not overview of the problem
|What||– What is the object, product?
– What is the deviation on the
|– What similar object could have
the deviation, but has not?
– What deviation could be on the
object, but is not?
|Where||– Where is the object Located?
– Where is the deviation on the
|– Where could the object be
located, but is not?
– Where could be the
deviation on the object?
|When||– When was the deviation first
– When was the deviation first
– In which Process stage it
|– When could the deviation have
been seen first?
– When could the deviation have
been seen since?
– In which Process stage could it
be, but is not?
|Extent||– How many objects have the
– What is the size of the single
– How many deviations are on
– What is the trend?
|– How many objects could have
– What could the size be but is
– How many deviations could
– What could the trend be?
4.3.5 Based on above, the team shall define the problem statement.
4.4 Identification of root cause:
Following tools can be used for identification of root cause (root cause analysis):
4.4.1 Five Why analysis:
220.127.116.11 Its question asking method to identify root cause.
18.104.22.168 Ask question 5 times (even more if necessary) to identify the root cause.
22.214.171.124 Stop asking, why? If, it does not further influence the result.
126.96.36.199 Refer below table for example:
|Why?||Answer and observation|
|Why was the pH found out of limit?||Due to pH meter error.|
|Why error found in pH meter?||pH meter was not calibrated.|
|Why pH meter was not calibrated?||Concern chemist forgot pH meter calibration.|
|Why concern chemist forgot pH meter calibration?||Due to oversight|
|As per above study concern chemist is the root cause.|
4.4.2 Fishbone / Ishikawa Diagram:
188.8.131.52 It helps to identify and categorize the root cause.
184.108.40.206 It shall be categorised as Man, Machine, Material, Method, Measurement and Environment.
220.127.116.11 Start from failure (Fish Head) and work backwards, start at high level, then increase degree of detail, collect possible causes.
4.5 Team shall identify possible causes based on above root cause analysis.
4.6 Evaluation of root cause:
4.6.1 The team shall review each possible cause and shall compare with actual fact and shall eliminate the cause by Cause/Fact matrix (Contradiction matrix) as below:
X … Fact contradicts cause ✔ Fact supports cause
The above chart indicates cause 2 and 5 are most probable causes.
4.6.2 Number of cause and facts are non-exhaustive therefore rows and columns can be added if required.
4.6.3 From above exercise the team shall arrive to the actual cause.
4.6.4 Wherever appropriate the actual cause shall be demonstrated and confirmed by following:
– Ask: what can be done to verify any assumptions made?
– How can this cause be observed in action?
– How can we demonstrate the cause- effect relationship?
– When corrective actions are implemented, how will the results be checked?
4.7 Confirmation of true root cause:
4.7.1 Actual root cause shall be confirmed as following:
– Check: Check assumptions made in contradiction matrix, against factual data.
– Observe: Obtain new measurements / observations that are conclusive about how the cause generates the deviation.
– Experiment: Make small scale experiments (e.g. in the lab) to prove cause.
– Change and monitor: Implement the solution and monitor the result to demonstrate success.
4.7.2 After above evaluation team shall state the true root cause.
4.8 Accordingly team shall plan corrective and preventive actions.
4.9 Effectiveness check of CAPA shall be done to ensure that failure investigation and subsequent root cause analysis has been successful and root cause has been fixed.
CAPA: Corrective and preventive actions
OOS: Out Of Specifications
QA: Quality assurance
QC: Quality control
Copyright Notice: This Article/SOP/Compilation/Published Content is protected by Indian & International Copyright Laws. Reproduction and Distribution of the same without written permission is prohibited. Mail us at: firstname.lastname@example.org
For any Feedback or suggestion mail at: email@example.com
Weblink – https://pharmastate.com/