Rethinking Knowledge Management & Data-Driven Risk Management For Quality By Design
By Peiyi Ko, Ph.D. and Peter Calcott, Ph.D.
Successful implementation of quality by design (QbD) can lead to significant revenue growth and margins from the shortened development cycle.1 However, in our recent reports on the state of ICH Q8-11 guideline adoption in the industry, we confirmed that, despite the sound rationale of the new operating paradigms, industry uptake has been slow since the publication of the guidances nearly a decade ago.2-4 In addition to factors pertaining to regulatory agencies, the key challenges to a company’s QbD adoption include lack of technology and internal and third parties’ misalignment.5 For example, historically, departmental silos and the gap between R&D processes and commercialization prevented coordinated and transparent collaboration among departments with often-different objectives. That is, the knowledge/data commonly resided in numerous silos that are difficult, if not impossible, to utilize.6 This can be due to the complexity of the processes and the large amount of data generated at each stage of the product development life cycle. Furthermore, as the reliance on contract manufacturing organizations (CMOs) and suppliers across the globe increases, it becomes even more challenging to have universal visibility and control across the whole supply chain.
In this two-part article, we start by reviewing knowledge management (KM) and risk management (RM) approaches in the context of QbD and quality management systems (QMSs) in Part 1, and then expand to relevant data management and leadership principles in Part 2. The hope is to address the internal challenges companies face while attempting initiatives such as achieving QbD by leveraging prior knowledge through data-driven quality risk management.
Guidances As Paradigms And Methodologies For Operational Excellence
The main hope of ICH Q8-Q10 pertaining to operational excellence is to enable (bio)pharmaceutical companies to achieve product realization. The goal is to do so by establishing and maintaining a state of control and facilitating continual improvement while responding to pressures for efficiency and profitability improvements.
ICH Q8 R2 “Pharmaceutical Development” can be distilled to the concept of QbD. That is, using the understanding of the critical quality attributes to design critical process parameters for seamless coordination between upstream and downstream product development. In contrast to the traditional or minimal quality testing approaches, this calls for systematic pharmaceutical development that utilizes process analytical technology (PAT) tools to relate mechanistic understanding of material attributes and process parameters to critical quality attributes (CQAs) of drug products. This serves as the foundation of a risk-based control strategy for well-understood product and process improvement. Both the regulatory agencies and the industry proposed that a business transformation focusing on cross-functional collaboration whereby product knowledge can be uniformly leveraged would result in both productivity and revenue gains. These approaches improve reliable delivery of quality and cost-effective drugs to patients as quickly as possible.
ICH Q9 outlined the framework and principles for pharmaceutical quality risk management. It is essentially a decision-tree process that leads to a clear definition of the triggers for risk mitigation activities based on the assessment, which should be based on scientific knowledge for the protection of the patient. The level of effort, formality, and documentation of the quality risk management process should depend on the level of risk.
However, risk management is open to individual interpretations and at times has varied and been time-consuming and not-informative for prioritization, re-inventing risk assessments and leading to wasted resources and unsatisfactory outcomes. Therefore, it is proposed to use a quantitative approach with probabilistic calculations and monetized harm to account for occurrence and severity, respectively.7 Specifically, failure modes and effects analysis (FMEA) is a classic tool for summarizing the modes of failure, factors causing these failures, and the likely effects of theses failures to reduce process complexity for management. It generates a risk priority score for a failure mode by multiplying the ratings for severity, occurrence, and detection.
After incorporating probability of failure modes against the impact of the consequences, FMEA becomes failure modes, effects, and criticality analysis (FMECA) and is most useful for guiding resource deployment. How to incorporate a mixed stream of data from historical data, theoretical analysis, informed opinions, concerns of stakeholders, etc. to determine gaps in knowledge, gaps in pharmaceutical science and process understanding, sources of harm, and probability of detection of problems as well as innate variability then becomes the main task to achieving data-driven risk management.
ICH Q10 considers risk management and knowledge management the key enablers of a mature pharmaceutical quality system. To truly reap the benefits of this quantitative approach to risk assessment, it is important to have relevant information and more knowledge on source of uncertainty. This is where knowledge management comes into play: By establishing a “systematic approach to acquiring, analyzing, storing, and disseminating information related to products, manufacturing processes, and components” (ICH Q10), companies can then better leverage prior knowledge for data-driven quality risk management. This is the key to QbD.
Strategies For Knowledge Management
Knowledge management is arguably one of the most important systems for any (bio)pharmaceutical company. From the operations perspective, knowledge management is a vital connection between other management subsystems for learning and development, information management, project management, and organized literature. Employee learning and development, teamwork, knowledge-friendly culture, and infrastructures (organizational technology, information systems) are all critical factors for successful implementation of KM.8
However, at present the barriers for effective implementation are both technical and cultural. The technical challenges are the storage, organization, and access issues. This could be addressed by the codification strategy to knowledge management, which emphasizes building a comprehensive repository such as an electronic document management system (EDMS) of codified knowledge that can be made available for reuse and could connect the knowledge database to system users through a web-based interface. As the amount of data grows, this strategy requires integration across disparate systems with good master data management, well-designed information architecture, effective indexing with metadata, application of user-friendly analytical tools, and social-type communication systems with proper setup of access privilege and training on confidentiality.
Though tools such as groupware may help dissolve corporate hierarchy by making it easy to share information, cultural challenges such as solo/territorial and gatekeeping mentalities remain challenges and will require education and buy-in from staff led by progressive management that fosters a culture of collaboration through empowerment and rewards. When it comes to information-sharing with third parties such as suppliers and CMOs, monitoring through better visibility and collaboration through connectivity are main objectives. As cloud-based services mature and are more readily available, balancing potential benefits of the network effect (e.g., for sales and operations planning or transportation management) with sensitivity of data sharing (e.g., for planning and product life cycle management) can help prioritize implementation.9
While information technology can help to manage knowledge by storing and distributing explicit data or facts, it still requires people in the organization to interpret the meanings of this information. Tacit knowledge is where competitive advantage lies, as it is hard to imitate. This is especially true for manufacturing and development skills. Knowledge often comes from understanding cause-and-effect relationships and builds over time in the heads of employees in the form of past decisions, processes in the organization, characteristics of products, interests of customers, and similar experiences. It cannot be used elsewhere if it is not documented and shared.
In the context of QbD, an example of this knowledge is the proven acceptable range (or design space, DS) of an operation that is validated through scientific studies. A knowledge gap can be created by poor documentation and further aggravated by employee turnover. An example is prior developmental data that could help solve a deviation had to be recreated because the scientist who completed the work left the company without having properly documented it. The problem is likely to be prevented with an electronic notebook system, training on the best practices for using the tool and why, and management support of the implementation. This marks the importance of the personalization knowledge management strategy, which aims to excel in person-to-person communication and contributing knowledge through collaboration in teams.10 This strategy can benefit from systems for electronic coordination across time and space (e.g., group decision support systems, electronic meeting rooms, videoconferencing, chat rooms, groupware email) and groupware to search and retrieve organizational knowledge.
Leverage Prior Knowledge Toward Data-Driven Risk Management
In pharmaceutical production environments, extreme swings in variability are common, sometimes even after applying lean techniques for improvement(Auschitzky, Hammer, & Rajagopaul, 2016). There are many problems — including waste, deviations, and failures — that can be interpreted as risks. And risk assessments and control measures updates need to be made as changes are introduced. Some risks are direct and easy to quantify, such as losses from failed products or leftover material. Some incur indirect costs associated with failures, investigations, inventory holdings, and delays for released products. Some are due to unpredictability in demands, which may lead to designing a plant bigger than needed. Some issues may even have longer-term consequences of orders lost and damaged reputation. These issues are challenging to quantify and prioritize. Given the sheer number and complexity of production activities that influence yield in this industry, manufacturers need to use advanced analytics as a more granular approach to diagnosing and correcting process flaws.
Many tools can be applied to quality risk management. Basic risk management and facilitation methods include flowcharts, check sheets, process mapping, and Ishikawa/cause-and-effect diagrams. Quantitative statistical tools such as control charts, design of experiments (DOE), histograms, Pareto charts, and process capability analysis can facilitate more reliable decision making in the activities leading to QbD. The key challenge is developing the contextual link between routine manufacturing data sets and the vast array of experimental data obtained during product and process development and pilot life cycle stages to truly understand the risks. Under the QbD framework, a development organization can define the platform in the DS as the basis for continuous improvement and operational excellence by linking critical process parameters (CPPs) in the development phase to CQAs and understanding variability. Establishing a DS also provides manufacturing and regulatory flexibility, as movements within the space would not be considered changes. This helps reduce the need for performing risk assessment through the iterative process.
Since this set of guidances indicates product and process knowledge must be managed from development through the commercial life until the end of life (i.e., discontinuation) of the product, a well-designed QMS will demand effective knowledge management as described in the previous section. Examples of sources of knowledge, whether internally documented or in the public domain, are prior knowledge on other similar molecules, pharmaceutical development studies on the precise molecule, technology transfer activities, process validation studies over the product life cycle, manufacturing experience, innovation, continual improvement, and change management activities. And how does manufacturing experience add to the overall knowledge for success in development? After aggregation of previously isolated data such as historical process data, sophisticated statistical assessments are applied to identify patterns and relationships among discrete process steps and inputs, and then the factors that prove to have the greatest effect on yield are optimized.11 The knowledge developed through this approach could also help with investigation of deviations, development of control strategy to improve product quality, and better management of risks.
Conclusions
In this article, we discussed how implementing these guidances can help achieve operational excellence, barriers, and the importance of data-driven approaches for improving knowledge management and risk management to enable QbD implementation. In Part 2, we will dive deeper into data management strategies and leadership principles that address fundamental challenges, such as coping with complexity while attempting these initiatives.
References:
1. McKinsey & Company. (2009). Challenges to Quality by Design. Retrieved from http://www.ipqpubs.com/wp-content/uploads/2012/01/McKinsey_report_QbD.pdf
2. Calcott, P., & Ko, P. (2017). Measuring The Impact Of Recent Regulatory Guidances On Pharma Quality Systems. Retrieved from https://www.pharmaceuticalonline.com/doc/measuring-the-impact-of-recent-regulatory-guidances-on-pharma-quality-systems-0001
3. Calcott, P., & Ko, P. (2017). Survey Says: Pharma Perspectives On Implementation & Impact Of Recent Regulatory Guidances. Retrieved from https://www.pharmaceuticalonline.com/doc/survey-says-pharma-perspectives-on-implementation-impact-of-recent-regulatory-guidances-0001
4. Calnan, N. O. (2013). Enabling ICH Q10 Implementation--Part 1. Striving for Excellence by Embracing ICH Q8 and ICH Q9. PDA Journal of Pharmaceutical Science and Technology, 67(6), 581-600.
5. Nasr, M. M., & Winkle, H. N. (2011). FDA Perspectives: Understanding Challenges to Quality by Design. Pharmaceutical Technology, 35(9). Retrieved from Pharmaceutical Technology: http://www.pharmtech.com/fda-perspectives-understanding-challenges-quality-design
6. Helfrich, J. P. (2013). QbD: The Devil is in the Data. Retrieved from http://www.pharmamanufacturing.com/articles/2013/1307-qbdthedevilinthedata/
7. Strong, J. C., McDermott, T. S., Heinzerling, O., & Claycamp, H. G. (2017). The Risk of Trusting Risk Priority Numbers. AAPS Magazine, pp. 14-19.
8. Jain, T., & Pandey, B. (2012). Knowledge Management Implementation. Retrieved from BioPharm International: http://www.biopharminternational.com/knowledge-management-implementation?id=&sk=&date=&pageID=3
9. Courtin, G. (2013). Supply Chain and The Future of Applications. Retrieved from http://www.mypurchasingcenter.com/files/5013/9992/0399/SCMWorld-Supply-Chain-and-the-future-of-Applications.pdf
10. Hansen, M. T., Nohria, N., & Tierney, T. (1999). What’s Your Strategy for Managing Knowledge? Havard Business Review.
11. Auschitzky, E., Hammer, M., & Rajagopaul, A. (2016). How Big Data Can Improve Manufacturing. Retrieved from McKinsey & Company: http://www.mckinsey.com/businessfunctions/operations/ourinsights/howbigdatacanimprovemanufacturing 2/7
About The Authors:
Peiyi Ko, Ph.D., founder of KoCreation Design, creates opportunities for positive changes and innovation through human-system interaction research and human-centered design. She has guest lectured at universities and led workshops. She is also a Certified Professional Ergonomist and collaborates with the Interdisciplinary Center for Healthy Workplaces at UC Berkeley as a consulting expert. She has provided human factors/ergonomics consulting as well as software usability analyses and design recommendations for operational improvements at the Lawrence Berkeley National Laboratory and at BSI EHS Services and Solutions. She obtained her Ph.D. from UC Berkeley in 2012 with additional training from two interdisciplinary certificate programs there: Engineering, Business & Sustainability and Management of Technology.
Peter H. Calcott, Ph.D., is president and CEO of Calcott Consulting LLC, which is focused on delivering solutions to pharmaceutical and biotechnology companies in the areas of corporate strategy, supply chain, quality, clinical development, regulatory affairs, corporate compliance, and enterprise e-solutions. He is also an academic program developer for the University of California, Berkeley in biotechnology and pharmaceutics postgraduate programs. Previously, he was executive VP at PDL BioPharma, where he was responsible for development and implementation of quality and compliance strategy across the corporation. He has held numerous positions in quality and compliance, research and development, regulatory affairs, process development, and manufacturing at pharmaceutical companies including Chiron, Immunex, SmithKline Beecham, and Bayer. He has successfully licensed products in the biologics, drugs, and device sectors on all six continents. Dr. Calcott holds a doctorate in microbial physiology and biochemistry from the University of Sussex in England. You can reach him at peterc@calcott-consulting.com.