The business value amplification of the data steward role - Insights shared by leaders at Takeda and Fujifilm Diosynth

This article summarizes an interview with industry leaders Daniel Carraher at Takeda , and Daniel Hill at FUJIFILM Diosynth Biotechnologies and covers:

1. how data stewardship is essential across all areas of pharmaceutical manufacturing in supporting regulatory compliance, product quality, and patient outcomes.
2. the operational value AND strategic value that Data Steward activities produce,
3. balancing formal and informal approaches to data stewardship across data domains,
4. examples of implementing data stewardship practices including what leaders expect to see from performers and go-to-experts, and
5. advice on how leaders can develop data steward skills in those starting out in their career.

In his role as Director of Products for Digital and Data Science at Takeda, Daniel Carraher leads cross-functional teams to deliver digital products across Data Management, Data Engineering, and Manufacturing Intelligence. He has also worked in roles at Genzyme and Pfizer.

Dan Hill recently joined FujiFilm Diosynth Biotechnologies (FDB) as Director of Process Analytics having worked at Thermo Fisher and Biogen. At Biogen he had responsiblities for holistic data and digital strategy development, and program execution for six CMC functions spanning three therapeutic modalities.

This discussion builds upon superb work by Robert Dimitri, who previously represented Thermo Fisher on the Life Sciences Advisory Board and is currently Director Digital and Analytics at DynamixE, Inc. . Rob authored the first draft of the Data Steward profile by drawing upon his experiences at Takeda that included expected knowledge, sources of information, and expected behaviors of the role of data steward at four distinct stages of their career, including new-in-role, contributor, performer, and go-to-expert.

The role and value of Data Stewards in pharmaceutical manufacturing

Malcolm Jeffers: Daniel, will you share where you have encountered the data steward role in your organization at Takeda?

Daniel Carraher: "I've come across it more formally in the clinical studies setting, and I think that's because of the scrutiny and criticality of managing and handling data in clinical studies. It could make or break the drug product approval, right? It also has direct consequences for patients potentially.

In manufacturing, we are working with data we consider important, often containing proprietary processes and industry secrets that require careful handling. While it may not contain patient information, it's crucial to be thorough and comprehensive due to regulatory scrutiny and its impact on product quality and patient outcomes.

So, data stewardship is essential across all areas, aligning with our enterprise-wide data governance strategies."

Malcolm Jeffers: Let's discuss the value that data stewards or data stewardship activities bring to an organization or function. Dan, could you start us off?

Dan Hill: "I think you can look at value in a couple of different ways.

There's operational value in running the business, especially in manufacturing spaces. Data stewardship is directly tied to Continued Process Verification (CPV) and annual product reviews, ensuring compliance and improving productivity. A data steward has a strong role there, just within the operations side.

Strategically, if a data steward understands where the business is going, the types of work the business would like to do, and the overall transformation it wants to take in terms of using analytics and artificial intelligence, they play a critical role. They help align data across the business, supporting where the business wants to go strategically. There's also the aspect of being aware of future compliance needs and being more predictive on compliance issues."

Operational Value that Data Steward activities bring:

• Ensuring compliance in areas like Continued Process Verification (CPV) and annual product reviews
• Improving productivity by supporting data users and scientists
• Reducing time spent on data management by scientists (which can take 30-40% of their time)

Strategic Value that Data Steward activities bring:

• Aligning data management with business goals and transformation initiatives
• Supporting the implementation of advanced analytics and artificial intelligence
• Helping predict and prepare for future compliance needs

Malcolm Jeffers: Daniel, would you like to add to that and share your perspective?

Daniel Carraher: ""I would echo Dan's thoughts. In my team, the Statistical Process Control, Data Management, Data Engineering, and Manufacturing Intelligence folks are within my organization. We've tried to build a fluid resource model where the roles and responsibilities are blended to a certain extent. Each team member has sufficient competency in the roles of the other teams to function there if necessary. This enhances data governance because the expertise in data handling from each team is shared with the others, and knowledge of applications and solutions is also shared. This gives the team more knowledge about the beginning of the data stewardship process and its end use and consumption, allowing us to more effectively adopt a unified data strategy that works for everyone.

A lot of the questions we have about how to proceed come down to: Can I find the data I need? Does it exist? Can I trust it? Can I use this data, and what can I use them for, and how can I use them?

How are you implementing data stewardship practices?

Malcolm Jeffers: Dan, as you hear Daniel talk about the shared thinking around data stewardship responsibilities in his team at Takeda, do you see an advantage in everyone having a sense of ownership for these responsibilities?

Dan Hill: "I think the awareness of the data lifecycle - understanding what it takes to capture, store, use, and repeat that cycle - is really important for the majority of folks to have. That shared awareness is helpful to ensure we have a standard way of capturing data and are aligned on how we're storing it. However, I think the core role of a data steward, in either a process sciences or manufacturing context, is in shepherding or owning that whole process and making sure folks are moving in the same direction probably comes down to one person within a function. It's part of everybody's role, but there's one person really responsible for making sure that program is being executed."

Malcolm Jeffers: Daniel, does this conversation about owning the responsiblity of data steward activities raise any new questions for you?

Daniel Carraher: "I'd love to know how that works in different settings. In my organization, we don't have a formal data steward position, but the data stewardship role is perceived as very important. The responsibilities involve coordinating cross-functional partnerships at both the technical and non-technical levels.

Everyone in the organization has some data steward responsibilities because it's most effective when everyone does their part.

My team implements a lot of good data practices, and we try to align with the intended use and user requirements for different datasets. We even go down to the level of writing a service level agreement for how each product is going to be used, including the level of regulatory compliance and the required data refresh period. Formalizing the data steward role could enhance our coordination and data governance efforts. We do this all the time, but we don't have a specific role.

We take a cross-functional team approach with a fluid resource model and socialize different aspects of good data management, such as sharing tidy data structures, how to minimize touch points and interactions, and deciding on what level of formalization is needed through establishing or applying standards for example."

Dan Hill: "Just to add to that, I envision like the role of a data steward as being the one that drives a lot of those conversations within a functional area, for example a lab lead, working with the SMEs in their lab space to understand the business processes, how data is being generated, what alignment is needed and then working with someone else in manufacturing to make sure that those pieces are now able to tie together to create shared use cases.

How can the Data Steward role amplify value?

Malcolm Jeffers: Do you see the real value begin to accrue when someone can see beyond their own function?

Dan Hill: "I think it's about how you amplify value. You can create a lot of value within a function or a small part of the business. But these stewards can certainly amplify their value by working across functions and making sure that we're able to draw additional insights from different parts of the value chain. That's where we start getting into higher impact types of analytics. Being able to take manufacturing and process data, and blend them together, now you're doing types of analytics you couldn't do before. I think that's really the next step in the evolution of a data steward."

Daniel Carraher: "I like the points that Dan made. Absolutely, I do see real value accruing when someone can see beyond their own function. Cross-functional understanding enhances our data stewardship efforts and brings significant benefits to the organization. While there's a more formal approach to data stewardship and a more informal approach, it's important to balance them appropriately.

In my team, we have experts in systems like SAP ERP, LIMS, and TrackWise. While I rely on their specialized knowledge and wouldn't dictate specific standards to them, we recognize the need for some level of formal data governance to ensure consistency, compliance, and data quality across the organization. There's general guidance that I can contribute, but there's also valuable information that flows back from these experts, so we strive to maintain open communication.

If we made our processes too formal, it might be restrictive for those deciding the best standards for their data domains. Therefore, we aim to balance formality with flexibility to prevent inconsistencies and silos. You can't apply one data table for every purpose. I appreciate formality in the right environment with the right goals, and we endeavor to find that balance to enhance our data practices and overall organizational value."

Dan Hill: "I'd suggest this is more context driven, where you know you're probably going to need to have just a range of approaches just to make sure that people are aligning in the right way at the right time."

How are you communicating data stewardship practices?

Malcolm Jeffers: I'm curious about how you communicate good practices. Is it more incidental or more structured? How do you communicate those good practices?

Daniel Carraher: "When my team provides training sessions, which includes colleagues who do data engineering in various platforms, we started talking about certain ways of working that people tend to call tidy data structures. Establishing standards or even just making people aware of these things can improve usability of data within the organization. We use training sessions and knowledge management tools to communicate these good practices, ensuring that data structures are usable by both technical and non-technical experts.

We tend to take this cross-functional team approach and socialize different aspects of good data management. While our communications are somewhat informal, they are structured through these training sessions and tools to ensure consistency across the team.

When we're diving into discussions and defining business processes, that's when internal best practices and standards get discussed. This approach contributes to our overall data governance and compliance within the organization. You have to recognize the starting point of where different systems and practices are and start from there, asking how to take steps in the right direction. We're committed to continuous improvement in enhancing our communication and data practices."

What do you expect to see in a data steward?

Malcolm Jeffers: What do you expect to see in a data steward at the performer and go-to-expert levels?

Daniel Carraher: "Understanding the environment they're working in and taking a risk-based approach is important for data stewards at both the performer and go-to-expert levels. At the performer level, I expect data stewards to comprehend failure modes and ask about typical error rates in different scenarios. They should implement guardrails and checks to avoid specific data errors, aligning their actions with the organization's data governance policies and standards.

Another aspect is adopting good practices that make life easier and increase the likelihood of success. For instance, knowing whether to work with one type of data table versus another, or if multiple data tables are needed, demonstrates proficiency with data tools and technologies—a skill essential at the go-to-expert level. Go-to-experts not only focus on quality and compliance but also contribute to solutions that support the organization's strategic objectives.

Strong communication and collaboration skills are crucial, as data stewards often work across functions to promote continuous improvement in data practices. While some people think of the data steward role as being more on the quality and compliance side, it is equally about providing solutions and innovations. Therefore, balancing these responsibilities is key to effective data stewardship."

Dan Hill: "I think a data steward will want to keep both sides in tension knowing the types of insights to generate to produce a high enough value but also putting more data quality around a certain object if dealing with data that could have a a good amount of risk if we get it wrong. So, I think being able to kind of manage that tension is key. For example, improving productivity outcomes could justify putting more governance around data because we know that it is going to drive higher productivity output. The data steward needs to be able to understand the use cases, what those use cases will provide, the amount of effort it is going to take to get to that value and manage a portfolio around that."

Malcolm Jeffers: Can you share some advice on how leaders can develop data steward skills in others who are starting out in their career?

Dan Hill: "When I bring in co-ops or interns, I try to have them understand the use cases or outcomes we're trying to drive toward. It's about them understanding how they're driving the outcome and why it's important to apply certain standards to their data. I encourage folks to understand the bigger picture. When I started in the industry, the focus was often just on executing tasks as an entry-level scientist or chemist. But it's really hard to do that well if you don't understand how that data is going to be used. Giving them context and having them focus on the bigger picture is a way to train them and help them understand why they should apply more rigor in their day-to-day work."

Daniel Carraher: "I completely agree with what Dan said about needing to understand the applications and desired outcomes. I'd also encourage developing key skills and competencies in data stewardship, like understanding data governance, data quality management, and data policies. Leaders can help by ensuring there are clear data standards and guidelines in place.

It's important to align these skills with our organizational goals, so newcomers understand how their work contributes to the bigger picture. I'd also emphasize the importance of communication and collaboration. Encouraging them to ask another layer of questions can be very beneficial. Given those desired outcomes and applications, what aspects of the data being worked with are most important? Sometimes it might be certain parameters or keys, or how to join different data sources. Other times, it might be about trusting a certain part of the process, like ensuring the refresh rate is appropriate.

There are a lot of questions they can ask about what options within the data flow matter with respect to the application and desired outcomes. By fostering an environment where open communication and collaboration are encouraged, leaders can help new team members develop essential data steward skills that are aligned with our organizational goals."

What are your thoughts on the future significance of this role?

Malcolm Jeffers: How much more significant will the skill set of a data steward be to the pharmaceutical industry in years to come?

Dan Hill: "I see it becoming very significant. The big question we're going to have over the next five years is whether it's worth the hype. Are we really delivering the value that we think we can? If we are, then this role is going to be extremely critical. If companies are able to really harness the value of data, it's not a question of if you should, but when - you must, just to be competitive. But it all comes back to whether we can really deliver the value that we think is there."

Daniel Carraher: "As we introduce more sensors, collect more data, and start doing more with it, the progress is accelerating—much like what's happening with digital technology in general. I think the skill set of a data steward is going to become significantly more important in the pharmaceutical industry.

Everywhere from R&D to manufacturing, the amount of data we're collecting is exploding. Things are being automated with high-throughput automation, and we're integrating emerging technologies like AI and machine learning. So there's just going to be more and more data to work with.

Data stewards will play a critical role in ensuring data quality and integrity as these volumes increase. They'll need to be proficient not just in data governance and regulatory compliance—which are crucial in our highly regulated industry—but also in advanced data analytics and integration techniques.

With the challenges of data privacy concerns and complex data integration, data stewards will be essential in implementing robust data management solutions. Their role will contribute to strategic decision-making and provide a competitive advantage by making sure all this data we're collecting is reliable and can be effectively leveraged.

Ongoing skill development and training will be necessary to keep up with technological advancements. And I think there will be a growing need for global data standards and collaboration across different regions and departments. So overall, the significance of the data steward role is only going to increase as we move forward."

The full set of seven IAAE Transformation Profiles, including that of data steward, was developed and reviewed by the IAAE Life Sciences Advisory Board and IAAE team, and can be downloaded via the Transformation Profiles link within the IAAE Knowledge Center.

This role of data steward will be presented by Malcolm Jeffers, Transformation Strategist and Head of Research with IAAE on behalf of Rob, Dan, and Daniel at the upcoming 2024 ISPE Pharma 4.0™ and Annex 1 Conference in Rome, Italy, and virtually, 10-11 December 2024.

Please note that the views of each contributor below are not the stated views of the organizations they work for.

#datasteward #myIAAE #LifeSciencesManufacturing