Knowledge Management (KM) for a Faster Time to Market by Irwin Hirsh

Forging the empire-building discipline

• 03.02.2024

In this first of five articles two knowledge management (KM) systems will be presented through the lens of CMC (Chemistry, Manufacturing & Controls) process development. Common challenges to implementation and fundamental structures of the systems will be discussed. Additionally, the (ROI) return on investment in the tools and changes to your way of working will be presented as the business case.

Although the examples and context for discussion come from CMC Process Development the approach and tools presented for knowledge management are agnostic and will add value anywhere in your value stream.

What is Knowledge Management

Knowledge management in the context of CMC development is the systematic organisation and sharing of critical knowledge related to the process of development so that it is available for use when needed and continues to be built upon as new knowledge enters the organisation and or knowledge gaps are identified.

The Costs of Poor Knowledge Management

In new drug development CMC failures in knowledge management can significantly extend the time required for regulatory approval and thus market access. Failure to supply clinical trials is a particularly egregious example, other more subtle, yet more commonly experienced examples include:

• 1.A poorly understood Manufacturing Process with high levels of variability. In the best case, such a process will reduce productivity and increase costs of non-quality for products that make it to the market. In the worst case, the inconsistency of batch performance also results in failed process validations and launch delays.
• 2.Lack of robust analytical methods result in the inability to accurately assess variability in process parameters and their impact on product quality attributes. This can lead to a rejected dossier and request for additional data and experiments. On market it increases risks for scrap and even recalls.
• 3.Inadequate documentation (knowledge sharing) commonly occurs when communicating about process capabilities, for example at changes of scale and location or between upstream and downstream processes after changes are made. Even well-designed processes may fail to win regulatory approval or suffer costly delays if development claims cannot be supported with documented evidence.

KM Systems

The two most important knowledge management systems for CMC process development are:

• 1.Quality risk management (QRM), for the systematic management of the knowledge of quality risks (risk to the patient) across the product life cycle including and starting with development.
• 2.Business process knowledge management. The know-how for performance of the tasks that are required in process development: from controlling raw materials and CDMOs to COGS modeling and process characterisation studies.

The Business Case for KM

A Business should be eager to develop knowledge management, not because it increases compliance, but because of the return on investment (ROI) it delivers to the business. Some ROI for CMC programs is highlighted below:

Time to competency, for those new to the organisation and or project is reduced by good knowledge sharing. When critical knowledge about the project and training in performance of the critical processes is available in a timely and systematic way an individuals' movement in and out of a project causes minimal delays. Additionally, when the knowledge is explicit, for example, available as a digital resource, team members have minimal disturbances to their ongoing work due to on boarding.

Knowledge sharing kills ambiguity, reduces complexity, and increases productivity. Knowledge management achieves these benefits by creating clarity on roles and responsibilities. Task specific process aides utilise critical knowledge collected so individuals know who must perform what standard work when and how. In such an environment specialists can turn their attention away from training and fire fighting and look towards innovation.

A systematic application of QRM across the entire life cycle of development provides clarity about the gaps between product requirements and the manufacturing process capabilities. This knowledge enables limited resources to be focused on the most important needs of development.

Timely knowledge sharing about these quality gaps maximize the dynamic response of the process to the product demands and enables controlled changes in response to new clinical or business requirements. Additionally, should the entity be sold at any point in its development, a fair price can be negotiated based upon a clear and substantiated risk profile.

Challenges to Implementation

Compliance with QRM

Although QRM could be considered a business process it is singled out for its importance to process development and because since 2005 and the publication of ICH Q9 it has been a regulatory expectation to which our industry has continuously faced challenges. Systematic application of QRM throughout the product life cycle, in the experience of the author, has been tepid at best.

Long (7+ years) development cycles, multiple departments with overlapping responsibilities, but significantly different skills and focus areas, churn in staff and under allocation of resources until clinical phase 3 often result in QRM being reduced to a one-time effort in late staged development instead of a value adding life cycle tool.

Understanding QRM Knowledge Flows

This 7-minute video presents an overview of how QRM information flows and knowledge is created during development to align process capabilities with product requirements. Understanding this conceptual framework is the essential first step to the successful systematic deployment of QRM.

Subsequent articles will provide insights into the use of digital knowledge management tools to overcome the implimentation challenges mentioned.

QRM as a Compliance Exercise

Too often the first attempt for a systematic use of QRM begins in late-stage development, after the process is largely fixed. The focus of the QRM exercise is usually a process FMEA performed before PV to de-risk failure of consistent batches and provide the information required to fill out the relevant sections of 3.S.2 in the CTD.

However significant this work is, when QRM starts in late phase development it often contains significant information gaps on criticality for many of the process parameters. This results in gaps in the cause-and-effect knowledge about process parameters and their impact on product quality attributes. Additionally, such a late implementation of QRM is highly indicative that the knowledge captured will fail to become living knowledge passed on to the final site of manufacture to support changes planned or otherwise.

Similarly, any process that is not understood or fully implemented can at best deliver limited value. Many CMC processes suffer from a lack a common single source of knowledge. Multiple versions of the same process exist so that the same work is performed differently at different sites, departments or even within the same project. In short, poor governance.

To derive business benefit from the knowledge generated by our processes we must first understand how the processes work, what data are generated and how that data can be used as knowledge to improve processes to meet business goals.

Lack of Process Orientation

The first step in a digital journey, wherein we intend to manage our process knowledge on a shared digital platform should be the maturation of one's processes. The infographic in Figure 1 below shows why processes are broken processes, and this presents a systematic approach to their repair. Once chaotic processes are described and standardised, they can become well understood and consistently executed. In the most mature stage, they become automated and provide predictive insights.

Any process that is poorly described, understood, implemented and maintained when mapped onto a digital platform can only result in faster failures and the elimination of the analogue workarounds established to 'get the job done' within the broken system of processes.

Figure 2 below depicts process maturation as a progression through five (5) levels. This model was inspired by the work of Phillip Crosby, a late 20th-century quality guru. The maturity grid was created to support clients of the author who worked in a GMP environment and were kicking off a formal process governance system for their department.

For six months prior to this kick-off, to support the launch and ensure the process owners understood what they were working on, the fundamental work of simply cataloguing and mapping the departmental processes was performed by the line managers and specialists.

Management Participation

Direct participation by senior management is essential for success knowledge management initiative as with any significant change management initiative. Future discussions in these articles on the business case and ROI can be adopted and contextualised by businesses to fit their specific needs, budgets, and culture. This I believe applies to development as much as any other business unit.

Once management participation is firmly established and process maturation has moved forward, perhaps somewhere between levels 1 &2, a concurrent project to rethink the digital knowledge management strategy of the organization is warranted.

Outdated Document Orientation for Data

Digitisation of our processes and the data they create must change paradigms from document centric to a data centric. At present we often create documents that serve a purpose at a distinct point in time and for compliance purposes we must retain that document. The knowledge of that document is then frozen as a PDF file. Accessible and catalogued, but not likely to be widely known of or ever accessed. This is commonly called, 'paper to glass' and this knowledge for all practical purposes is lost.

Most manufacturing organizations begin their digital journey focused on mining manufacturing batch run data. However, development expects manufacturing methods to evolve, and the data mining gold is the experimental data that elucidates the cause-and-effect relationships between the process parameters and the product attributes.

Once development can prove, with data, that the process consistently produces product with the predefined quality attributes optimisation for business purposes can take precedence, features to delight the customer, and win market share come into being. Discussions around innovation and identification of IP can get the attention they deserve. Alternatively, development resources can simply be deployed to new entities.

KM Implementation Guidance

In the next article, guidance on the pre-digitisation work that must be done for QRM processes to avoid the 'Garbage In / Garbage Out' algorithm will serve as the model for process maturation.

In the subsequent articles you will be presented with examples of how digital tools for managing the processes support process owners to continuously improve processes and thus the performance of those that use those processes in the development of new medical entity manufacturing processes.

The tools presented will be off the shelf software solutions designed for specific purposes. A discussion of how to select a database for the raw data for example from a pH probe or other sensor is beyond the scope of this series. However, a quick internet search will provide you with a host of vendors and plethora of literature on this topic.