Data Criticality Assessment
PIC/S Position
2020 PDA Asia Pacific Virtual Conference
Pharmaceutical Manufacturing & Quality
22 September 2020, Singapore
Dr Dinesh Khokal
Director, Quality External Affairs
Amgen Biotechnology Singapore
2
Good Understanding Of The Use Of
Data, Its Risk And Lifecycle Is Important
Data
Data used
Effective for decision
onsite/ remote making on
inspection compliance
GMP/ GDP Environment
PIC/S Guidance: Good Practices For Data Management and Integrity In Regulated GMP/GDP Environments, PI 041-1
COPYRIGHT © PDA 2018
(Draft 3) 30 November 2018
3
Data Creates The Trust In Our Medicines
Data
management as
per GMP/GDPs
• Records, paper or electronic data are the
foundational evidence that our medicines
Digitalization Clinical trials & are safe and effective
and AI data
Data
• Data integrity controls should be from both
Integrity the behavioral and technical perspective
• Inspection confirms compliance
• When a firm fails to protect its data, it
Globalization of Risk based
Supply Chain control strategy cannot serve patients
COPYRIGHT © PDA 2018
4
Basic Data Integrity Principles Are GMP/GDP Requirements
With Harmonized Interpretation Among Agencies
• The data should be: Data Integrity is
- Attributable defined as “the extent
to which all data are
- Legible complete, consistent
- Contemporaneously recorded and accurate,
- Original or a true copy, and throughout the data
lifecycle”.
- Accurate
• Complete, Consistent, Enduring and Available
(PIC/S)
COPYRIGHT © PDA 2018
5
Where Are The Data Integrity Risks?
• Not Aware
PERSONNEL • Not Trained
• Culture
• Insufficient A robust QRM
DATA REVIEW • Not done
• Not in SOP
approach to
data
• Not Validating for governance is
RISK PROCESSES
intended use needed to
address the
• QC Testing risks in a
OUTSOURCING
• Manufacturing
systematic way
MANAGEMENT • Poor leadership
& QUALITY • Competition
CULTURE • Network
COPYRIGHT © PDA 2018
6
Balanced Risk Management Approach
Needed For Data Governance
Risk to Product
“…the protection
Resources Quality of the patient by
managing the risk
Other Quality
Efforts demand to quality should
be considered of
prime importance”.
Data ICH Q9
Governance
COPYRIGHT © PDA 2018
7
Relationship Between Data Risk And Data
Criticality Must Be Understood
Risk Quality Process Parameter
Attribute Criticality
• Severity of harm • Linked to the parameter’s effect
• Probability of occurrence Criticality on any critical quality attribute
• Detectability • Based upon severity of • Based on probability of
• Level of risk can change as harm occurrence & detectability
a result of risk • Level of risk does not • Level of risk can change as a
management change as a result of risk result of risk management
management
• Control strategy helps
COPYRIGHT © PDA 2018 ICH Q-IWG, Points to consider, 2011
8
Data Risk Assessment is crucial in
GMP/GDP Environment
Focus on
processes
Vulnerability
of Data e.g.
Production
Degree of QC
Deletion of Voluntary automation
action alteration of / human Evaluate
history data interaction data flow
Data Data Risk
Assessment
Risk Ability to
Deliberate
Deletion assure Method of
Falsification
of data data quality control and
of data
and review
Method of procedure
Loss or integrity generating
recreation processing
of data and storing
data
Conducted Proactively and Is An Integral Part of Decision Making Exercise
COPYRIGHT © PDA 2018
9
Data Needed To Ensure / Confirm Product
Quality and Patient Safety Is To Be Determined
Risk Data Criticality Data Risk
Management Assessment Considerations
• Determine • Impact to • Opportunity
importance decision for data
of data making alteration and
• Determine • Impact to deletion Protect
importance of product • Likelihood of
Process step quality detection Patients
COPYRIGHT © PDA 2018
10
Criticality Of The Data For Product Quality &
Patient Safety May Vary and Needs Assessment
Data •All data related to
Integrity manufacturing is
critical
Type of
Data data
Reliability •The criticality level
collected may differ due to the
data’s use and the
Data control under which it
Criticality is generated
Assessment
COPYRIGHT © PDA 2018
11
Classification of Data Criticality
Intended use the data directly impacts products quality & safety
High • Data that is related to product quality or supports a disposition decision and patient safety
• Critical Quality Attributes (CQAs) and Critical Process Parameters (CPPs)
Intended use of data relates to attributes that are not CQAs and CPPs
Medium • Data that is related to process robustness and consistency
• Need to be tightly controlled to assure process consistency
Intended use of data is to provide evidence of GMP compliance related to
monitoring and control of processes that do not fall in to high or medium category
Low •Data that is related to neither of the above would be considered general GMP
PDA Technical Report No. 84. Integrating Data Integrity Requirements Into Manufacturing and Packaging Operations,
COPYRIGHT © PDA 2018 2020.
12
The Data Criticality Assessment needs
consideration of several factors
Data
Criticality
Assessment
What is the
impact of the Which decision Impact of the
data to product does the data data on a
quality or influence? decision
safety?
COPYRIGHT © PDA 2018
13
The Impact To Quality Defines The Level Of Attention In
The Release Decisions – Example of role of models in development
• Models typically used to support product and/or process development
e.g., formulation optimization
Low
• Models can be useful in assuring quality of the product but are not the sole
indicators of product quality
Medium e.g., most design space models, many in-process controls
• Prediction from the model is a significant indicator of quality of the product
e.g., a chemometric model for product assay, a surrogate model for dissolution
High
The level of oversight should be commensurate with the level of risk
associated with the use of the specific model.
ICH Q-IWG Points to consider, 2011
14
Data And Its Assessment Are Crucial For A Release
Decision
Elements necessary
in the Batch Release process
1. Regulatory
Control Startegy & GMP elements
compliance data
Manufacturing
2. System related data for the Batch
current batch manufactured Release decision
Batch
according to regional procedures
3. Product related data from
manufacturing process
record
CoA
4. Product related data from
quality control
Batch
Elements defined regarding outsourced not released
activities, if applicable
ICH Q-IWG, Points to consider, 2011
15
Different data may have different impact to
product quality and patient safety
Pragmatic application of a risk-based and
confidence-based approach to data criticality
assessment saves efforts and resources to
both industry and health authority while
ensuring product quality and patient safety
COPYRIGHT © PDA 2018
16
Acknowledgement
•Stephan Roenninger, Amgen
•Patrick Swann, Amgen
COPYRIGHT © PDA 2018
17
Thank you for your attention
Email: dkhokal@[Link]
