Preliminary QIG Considerations regarding Pharmaceutical Process Models
6
- 6
Preliminary QIG Considerations regarding Pharmaceutical
Process Models7
Background
8
This Quality Innovation Group (QIG) document follows on from the first QIG Listen & Learn Focus
9
Group (LLFG) on Continuous manufacturing and the second QIG LLFG on Digital novel technologies,
5
10
held on 13 March 2023 and 12-13 October 2023 respectively.
- 12
It is recognised that regulatory expectations for process models in pharmaceutical manufacturing are
13
evolving; the intent of this document is to share QIG?s current thinking with stakeholders and seek
14
their comments.
- This, in turn, supports adoption of advanced process
25
control strategies, continuous process verification, real-time process monitoring and optimisation, and
26
automated or even autonomous operation and management of manufacturing processes.
- Process
27
models play an increasingly important role in process design and validation, in control strategies and
28
during manufacturing process lifecycle.
- The expected outcome from the use of process models is
29
enhanced process understanding, (multivariate) monitoring and control, robustness, performance and
30
adaptability.
- 31
A model (in the context of pharmaceutical manufacturing) is a mathematical representation of a
32
physical or biological process or system.
- Empirical models (e.g., multivariate models used for Statistical Process Control, regression models
39
derived from data collected from Design of Experiments), and
40
3.
- 45
Scope
46
This document addresses preliminary considerations (general principles) for process models, reflecting
47
the use of performance-based approaches in pharmaceutical manufacturing processes.
- 48
The scope of this document is limited to process models such as first-principle models, regression
49
models, system models, multivariate statistical process control models, and Machine Learning models
50
(ML).
- Complex datasets need not be
Preliminary QIG Considerations regarding Pharmaceutical Process Models
EMA/90634/2024Page 3/7
113
submitted.
- 137
Preliminary QIG Considerations regarding Pharmaceutical Process Models
EMA/90634/2024Page 4/7
151
Of note, these are just examples.
- Process validation for finished products ? information and data to be provided in regulatory submissions
(EMA/CHMP/CVMP/QWP/BWP/70278/2012-Rev1,Corr.1)
Process validation for the manufacture of biotechnology-derived active substances and data to be provided in the regulatory
submission (EMA/CHMP/BWP/187338/2014)
4
5Preliminary QIG Considerations regarding Pharmaceutical Process Models
EMA/90634/2024Page 5/7
163
Some models may have a dual purpose e.g., used for process development and used as part of the
164
control strategy e.g., to set control limits.
- 200
Process validation (as described in the process validation guidelines) has an overarching role to ensure
201
that the process consistently delivers material of the intended quality.
- Depending on the model risk, a model verification protocol may be requested,
Preliminary QIG Considerations regarding Pharmaceutical Process Models
EMA/90634/2024Page 6/7
209
including the model performance metrics and the manufacturing process IPC and CQAs that should be
210
followed, the respective acceptance criteria, the number of additional data (independent) that would be
211
used, and the monitoring period (parallel testing).
- Preliminary QIG Considerations regarding Pharmaceutical Process Models
EMA/90634/2024Page 7/7