Continued ProcessVerification
Quality can not be
Tested into
‘Product’
Prathamesh P. Kulkarni,
Quality Assurance
It has to ‘Built- in’/ by QbD
1
Process Validation Life Cycle Approach- Enhanced Approach
2
Guidance's for CPV
3
Continued Vs Continuous Process Verification
Continued Process Continuous Process PART 820 : QUALITY SYSTEM REGULATION
Verification Verification SUBPART O : Statistical Techniques
(US FDA 2011) (ICH Q8 R2)
Section 820.250
Based on Knowledge
& confidence of Alternative approach
Design and to Process Validation (a) Where appropriate, each manufacturer shall establish and
development stage
maintain procedures for identifying valid statistical techniques
required for establishing, controlling, and verifying the
Replacement of
Can be used in acceptability of process capability and product characteristics
conjugation of
periodic
Process Validation (b) Sampling plans, when used, shall be written and based on a
‘Revalidation’
Traditional Approach
valid statistical rationale
Impact of any change
(Change control/ Increase in sampling
CAPA) on the and analytical load
parameter
Product in the ‘State
Of Control’
4
Continued ProcessVerification
This is 3rd Step in Product Life Knowledge gained from
Cycle Approach and Validation development is the foundation
efforts starts from Design stage for process Validation and CPV
Continued Process
Verification
21 CFR 211.180(e)
An approach to Process
validation that includes the Use of Statistical tool with more
continuous monitoring and numbers of Populations
evaluation of manufacturing
process performance
5
Steps in Continued ProcessVerification
Identify
Statistical
Collection of Evaluation of Analyze the process
Process
data data data stability and
Control
capability
6
Statistical Tools
Statistical Distribution Statistical Process Control
Characterize the ‘distribution’ of process Determine whether the parameters/ results are in
parameters closure time, temperature and ‘statistical control’ using control charts (with tests for
pressure using a Histogram - Normal overlay special cause 1, 2, 3)
7
Statistical Tools
Process Capability Design Of Experiments
Determine whether the process parameters are in Determine the parameters which affect seal strength
‘statistical control’, whether ‘normal distribution’ and their optimal values using a two level full factorial
models the data well and whether these process design of experiments.
parameters are ‘capable of consistently’ meeting
specifications.
8
Statistical Tools
Measurement Systems Analysis Sampling Plans
Determine whether measurements are accurate The Quality team want to inspect for the correct weight
and precise and if measurements can be trusted – of bags and to ensure the seal on the bag is closed.
are repeatable and reproducible. (Repeatability)
It is not always possible, efficient or cost effective to
inspect the entire lot. To determine an appropriate sample
size the team use international standard sampling
procedures…
1. Variables Method ISO 3951 (AS2490)
2. Attribute Method ISO 2859-1 (AS1199.1)
(Zero Acceptance Plans)
9
Continued ProcessVerification
When ? How? Why?
• Data collection • Statistical trend • To understand
starts from 1st by ‘Minitab the variability in
Commercial software’ the process
scale batch • Trained • To detect the
• Control limits to personnel on source of
be finalised after Statistical trend variation.
20 batches
10
Continued Process Verification : When and How?
Verification of
Commercial
After 20 batches batch within
Control limit Control limit
Finalization for
Data Collection Process
up to 20 batches Parameters.
within Validated
BMR Revision parameters.
as per PV Any Atypical OOT SOP shall be
recommendati Observation shall followed for
on be checked for Critical Quality
impact Attributes
Process
Validation assessment and
(03 batches) OOT procedure
recommend to be followed
11
ation
Benefits- Continued Process Verification : Why?
Real Time Assurance of Product Quality
Problem identification and Resolution
Enhanced Regulatory Compliance
Improved Operational (Manufacturing and Testing) Efficiency
Identify the source of Variation
Systematic improvement possibilities
Any Shift/ Drift in the parameter and its impact before batch release
Higher Statistical confidence for ongoing monitoring & verification
Enables to identify the ‘Control Variables’ 12
Continued Process Verification : Important Aspects
‘SEQUENTIAL TRENDING: For
correct identification of change
and adequate evaluation
‘Golden Batch Concept’- Batch
without any QMS event, no outlier
results and results close to ‘Target
Values’
‘VOICE OF BATCH’ Correct value
should be entered for trending
13
CPV : Case Study -Sequential Manner
Cpk with Correct Sequence of data Cpk with Change in Sequence of data
14
CPV : Case Study –Acceptance Criteria
Cpk for Related Substances Test Cpk for Assay Test
15
CPV : Case Study –Observed Values
I Chart for Critical Process Parameter with stringent I Chart for Critical Process Parameter with wider values
values
16
Continued Process Verification : Statistical Process Control
I Chart for Critical Process Parameter with Moving range (Variation
between the batches/ Lots)
17
CPV : Case Study –Control Limits
I Chart for Assay test Control limits within I Chart for Assay test Control limits are out of Specification
Specification limit limit
18
Continued Process Verification : What’s Next?
‘ARTIFICAL INTELLIEGENCE’/ INDUSTRY 4.0:
No human intervention and sound prediction, Simplification
and conclusion of data
19
20
