Pioneering a Data Strategy to Support Pharmaceutical Operations and Technology

By: 

• Erika L. Cerwin - Co-Lead Digital Evolution Initiative - Lead of Data Program
• Robert D. Guenard - Senior Director, Product and Technology Development
• Timothy B. Alosi - Head of PO&T IT, Global Product & Technology Development, Data & Analytics - Biogen

Executive Summary

Regardless of industry, maximizing the value of data has become an important objective in the digital age. As more pharmaceutical operations find themselves tasked with building long-term digital strategies, the most fundamental goal is to maximize the value of data for patients. Developing a comprehensive strategy for how a business manages its data is not easy, especially when it involves changes to the workforce, existing business processes, and technology choices. It requires executive sponsorship and building a case for change. Once committed to designing and implementing a data strategy, an organization can start to treat data, information, and knowledge as assets; moreover, business critical assets ARE guided data principles, governed by policies, managed as capabilities, and built up by frameworks. In order to prioritize future investment decisions, assessing data maturity and measuring progress towards a future state will lead to a sustainable data program. This article was adapted from a presentation given during the Information Architecture session at the IFPAC (International Forum for Process Analytics & Control) Conference in February, 2021. It focuses on why to build a data strategy and steps to build one.

Why Build a Data Strategy?

Organizations are rapidly shifting towards core competencies shaped by an integrated foundation of data, analytics, and software.1 In order to infuse operational agility to accelerate leadership decision-making, leaders are building a single data strategy, a clear architecture, and new data capabilities.1 In addition, megatrends in digitalization, globalization, and environmental awareness require constant focus and attention.2 Linking data from silos across the organization to a digital value chain with common language and clear data standards through a clear information architecture will help to bring the answers to important business questions that are hidden in the data. At the intersection of all of this, organizations will elevate the primary importance of data as a key asset for success.

In pharmaceutical operations, a data strategy closely links to digitally-enabled manufacturing processes. Business and society hold high expectations for technology and science-based developments, such as the Internet of Things (iOT) and Industry 4.0.4 As pharmaceutical operations continue to experience accelerated growth in data connected to business processes, data capabilities will become strategic and represent the underground plumbing required to manage data in a reliable and consistent manner across an organization as seen in Figure 1.

Leaders have expressed that the intent of investment in digital capabilities is to gain insight into the execution of their strategies.5 If using a home-building analogy, think of a business strategy as the blueprint for the house and the digital strategy as appliances like a dishwasher which enables activities like cleaning dishes to make life easier. Data is the water – linked through an underground network of pipelines not always visible to the end consumer – but critically important to cleaning dishes. The rooms in the house are delivering functional activities like discovering, developing, producing, or delivering products to patients.

Leaders agree that data capabilities must be closely linked to digital efforts and be the pipeline for rapidly evolving digital use cases.2 Analytics, like data, is also a key capability a digital strategy. Analytics pushes data into intelligence, predictions, and choices that guides operational workflows and enhances innovation.1 Digital and analytics are both being driven by data, but like a home’s plumbing, data is oftentimes taken for granted until it breaks. IT infrastructure is similarly at risk for a “run to obsolescence” operating model. Without data quality, it is always a simple formula: garbage in, garbage out. Once the data ecosystem is harmonized, there is significant opportunity. It is estimated that digitally transformed organizations in the US healthcare system, that also have a data strategy and an information architecture built to inform decision making, will derive over $100 billion in value annually.3 

With a clear value proposition, leaders from business and information technology should cooperatively work to evolve data management practices in order to fully realize data’s impact on the organization’s success. In the past several years, many pharmaceutical IT and business functions spanning regulatory, development, manufacturing, and supply chain have consolidated data practices into a centralized model. This underscores the importance of a growing trend to invest in data programs. One of the most critical areas of focus is to have a vision on how to include and manage data as an asset with a “business-led” and “technology-enabled” approach to harmonize business and IT perspectives.1 As organizations work harder to protect their data across the entire value chain to the patient, many digital use cases are emerging from digital initiatives, requiring both the business subject-matter-experts and functional IT working in full partnership.

Steps to Build A Data Strategy

Regardless of an organization’s approach to designing a strategy, an organization can start with the basics: set a clear vision, tie objectives to business strategy, address current pain points, and build a roadmap with the necessary steps to realize a future state. The following three steps are a path to get there:

Current State Assessment – Evaluate what data is important to drive the organization and what practices are being done today. Conduct broad sweeping interviews across all levels of the organization. Identify pain points and key challenges. Move from disparate thinking to common themes and topics to create a case for change with a clear problem statement. Know why the organization has been operating the way it has for many years. Discover how well data flows through business processes and workflows.

Future State Proposal – Brainstorm with leadership key opportunity areas. Define and prioritize data through a system assessment and landscape, understand high priority work linked to data, and conduct a high-value/low-effort approach to measure potential value to the business. Identify key opportunities and translate into a selection of capabilities from a framework.

Strategy and Roadmap – Build a 3-5 year plan. Use a simple way to communicate with a list of steps sequenced on a timeline. By determining a path forward, an organization is making a set of choices in the context of a future state. Keep it simple and always try to identify quick wins where the business can see the value the data program is bringing. Focus first on a minimum viable model required to advance people, process, and technology.

Current State Assessment

The first step to build a data strategy is understanding an organization’s data and business needs. Even if an organization has tried and failed to establish a data strategy in the past, it is important the organization does not shy away from this activity. Any previous work has value and relevancy in identifying and understanding organizational roadblocks that might undermine. Evaluating feedback and data needs from a wide swath of talent from analysts to leaders will expose the existing pain points and business challenges and start to characterize the current state of people, process and technology to define enablers. 

A balanced top-down/bottom-up approach to prioritize all the diverse data needs across these functional areas is important; it’s essential to prioritize in a systematic way.6 Conducting qualitative interviews to identify pain points requires a cross-section of perspectives on challenges. How are people managing or mismanaging data today for their functional area? How is the data sourced then mapped to processes? What technology debt hinders functional long-term objectives? What are common data domains? How are business rules stored and shared? Once the interviews are completed, the themes can be written out in a problem statement to support a case for change. Figure 2 provides an illustration of common pain points for organizations across a range of external leaders, external partners, quality operations, enterprise supply chain, and system owners as an organization started out on a journey to build a data strategy.

Leaders are increasingly finding new ways to link data across the digital value chain to patients and digitally-enabled processes are linking data sources across the chain with advanced technology, such as iOT, bots, and cloud, as examples.5 While sensors and edge computing

have been around for a while, the new applications of robotic process automation software for instance can even capture data management business rules for a master data team to remove low-value work to focus on new projects and address more complex business questions. The data is complex and differs by functional area. Here are a few examples of how it is different across the value chain:

1. development data in discovery and through the research-to-development handoff,
2. manufacturing data across drug substance and drug product is measured and monitored by complex processes,
3. supply chain data from raw material to finished goods being shipped from global vendors and partners,
4. regulatory submission data across a large portfolio of R&D medicines, and
5. many other internal and external pharmaceutical manufacturing and development activities generating data.

The power of data is enhanced by collecting and correlating information across key systems like actual productivity, batch release lead times, or maintenance status.2 Pharmaceutical operations are increasingly turning to cloud-based architectures to store data for a solution that can remain cost-effective while handling data complexity. The ability to tie this information together through reference data and master data allows an organization to make linkages across different systems. By building comprehensive end-to-end data models with a common set of entities and attributes, an organization can build repeatable datasets for analytics teams so less time is spent aggregating; as a result, more time is spent. 

After an organization conducts its qualitative analysis, it can build a quantitative assessment to track data maturity over time using a maturity model. Formulating data maturity models in highly automated and highly regulated industries like biopharmaceuticals is challenging.5 Similarly, the use of an industry-compatible data maturity model is complex; however, it can allow for necessary benchmarking against other organizations with similar data management challenges to help determine where to begin; advancing data capabilities, such as ingesting, prepping, or sharing data as seen in Figure 3.

Future State Proposal

Data has always had its own lifecycle of continual evolution and change.6 Published on Scientific Data in 2016 by a consortium of scientists and organizations, FAIR data principles are defined as Findable (metadata easy for both humans and computers to find), Accessible (users know how to access data), Interoperable (ease to integrate), and Repeatable (optimize reuse).7 In order to stop the proliferation of data across siloed functions with separate-but-not-equal data management practices, it is imperative that leaders advocate for designing data capabilities by stating how data needs to be collected and managed and then designing a system upfront to meet those future state needs.2 This is called a solution-based architecture and means to build with the end in mind. The data strategy implementation should have a charter to define the project team objectives, a resourcing plan, and a roadmap for milestones and timeline. With an end goal to increase speed to insights through analytics, a future state pharmaceutical organization is shifting to more high-value work to optimize business processes and workflows where regulatory compliance and data integrity are critical, as seen in Figure 4.

Data Strategy and Roadmap

Once future data opportunities are identified and articulated to leadership, organizations must prioritize them in a three- to five-year roadmap, which is simply a prioritized set of choices or steps over time. An investment level and spend profile will be needed to support the roadmap to ensure the effort is properly resourced. This level of investment is significant and should be considered a necessary infrastructure initiative. With a common vision defined, an organization is able to take the current and future state assessments to articulate to leadership where to begin. To build more impetus across the organization to begin on its journey, managers should consider categorizing the necessary steps by where the change will occur across the most common dimensions: people, process, and technology as seen in Figure 5.

Conclusion

Building a data strategy is as important as the plumbing of a house. It must be linked to business goals to maximize value. Data is used to run, improve, and transform business operations. The combination of data and analytics are common competencies expected of all organizations. By building an effective data strategy, in close partnership with business and IT stakeholders, an organization can design a clear architecture closely linked to solutions across the digital value-chain to patient success. Building a roadmap for how to get there will ensure a shared vision on how to start on the journey and develop best practices of managing data to benefit the patient.

Acknowledgments

Thanks to Roland Zhou for his foundational data capabilities and data maturity continuum graphics.

References

1. Competing in the Age of AI. Marco Inansiti and Karim R Lakhanl. Accessed May 15, 2021. Available at: https://hbr.org/2020/01/competing-in-the-age-of-ai
2. Managing data as an asset: An interview with Informatica CEO. May 1, 2019 interview. Accessed May 15, 2021. Available at: https://www.mckinsey.com/business-functions/mckinsey-digital/our-insights/managing-data-as-an-asset-an-interview-with-the-ceo- of-informatica#
3. McKinsey Center for Government: How Big Data Can Revolutionize Pharmaceutical R&D. McKinsey information page. Accessed May 15, 2021. Available at: How Big Data Can Revolutionize Pharmaceutical R&D.
4. Networks, ecosystems, fi elds, market systems? Making sense of the business environment. Published online 2020 Aug 24. Doi. Accessed May 15, 2021. Available at: 10.1016/j. indmarman.2020.07.013
5. Maturity Assessment: The First Step of the Pharma 4.0 Journey. Accessed May 15, 2021. Available at: http://www.qgdigitalpublishing.com/publication/?i=559340&article_ id=3282331&view=articleBrowser
6. RAND Data Strategies for Policymaking: Identifying International Good Practice. RAND information page. Accessed May 15, 2021. Available at: https://www.rand.org/content/dam/rand/pubs/technical_reports/2011/RAND_TR769.pdf.
7. FAIR Defi nition. Published online. Accessed May 19, 2021. Available at: https://en.wikipedia.org/wiki/FAIR_data

Author Biographies

Erika L. Cerwin, MSA, is the Digital Transformation Lead at Biogen and co-leading the Digital Evolution Initiative and leading the Data Program for Pharmaceutical Operations & Technology. Her area of expertise includes advanced analytics & optimization, management consulting, and technology innovation within the healthcare industry. She received her Master of Science in Analytics from North Carolina State’s Institute for Advanced Analytics. She earned her BSBA from the University of North Carolina at Chapel Hill Kenan-Flagler Business School where she was a Morehead-Cain Scholar.

To correspond with the author, please email: [email protected]. 

Robert D. Guenard has worked in multiple capacities in the chemical and pharmaceutical industries for more than 25 years. Rob currently leads the Digital Development and Analytics team at Biogen where he has been since 2015. He is currently leading the Biogen Digital Evolution Program for the Pharm Operations and Technology division. Prior to his current role, Rob spent 12 years at Merck (MSD) in various leadership capacities working on strategic initiatives, process analytics & control and knowledge management. Rob started his career and spent 7 years at the Dow Chemical Company in the Central R&D division working in process analytics and molecular spectroscopy. Rob received a B.S. in Chemistry from the University of Massachusetts - Lowell and a Ph.D. in Analytical Chemistry from the University of Florida. He is engaged in industry activities serving on the Pharmaceutical Process Analytics Roundtable (PPAR) steering committee, Co-Chair of IFPAC 2020 & 2021 and a member of the IAAE Life Sciences Advisory Board.

Timothy B. Alosi has over 25 years of process automation, operations management, and data analytics experience in the Biologics industry with such leading companies as Biogen, Sanofi , Genzyme, and Emerson Process Management. Currently, Tim is the head of the Global Data & Analytics team for Biogen’s Pharmaceutical Operations & Technology IT. The Global Data & Analytics team provides analytics solutions that support Biogen’s Product Technical Development, Manufacturing, and Quality operations, and includes Biogen’s global Digital and Data strategy development and execution. These solutions are utilized across Biogen’s global manufacturing network and supply chain. Prior to his work at Biogen, Tim has held leadership positions at Sanofi , Genzyme, and New England Controls, the local Emerson Process Management representative. Tim has a degree in Chemical Engineering from Massachusetts

Subscribe to our e-Newsletters
Stay up to date with the latest news, articles, and events. Plus, get special offers
from American Pharmaceutical Review – all delivered right to your inbox! Sign up now!