Data and a Good Data Culture

I often joke that as a biotech company employee I am primarily responsible for the manufacture of data (and water) first and foremost, and as a result we get a byproduct of a pharmaceutical drugs.

Many of us face challenges within organizations when it comes to effectively managing data. There tends to be a prevailing mindset that views data handling as a distinct activity, often relegated to the responsibility of someone else, rather than recognizing it as an integral part of everyone’s role. This separation can lead to misunderstandings and missed opportunities for utilizing data to its fullest potential.

Many organizations suffer some multifaceted challenges around data management:

Lack of ownership: When data is seen as “someone else’s job,” it often falls through the cracks.
Inconsistent quality: Without a unified approach, data quality can vary widely across departments.
Missed insights: Siloed data management can result in missed opportunities for valuable insights.
Inefficient processes: Disconnected data handling often leads to duplicated efforts and wasted resources.

Integrate Data into Daily Work

Make data part of job descriptions: Clearly define data-related responsibilities for each role, emphasizing how data contributes to overall job performance.
Provide context: Help employees understand how their data-related tasks directly impact business outcomes and decision-making processes.
Encourage data-driven decision making: Train employees to use data in their daily work, from small decisions to larger strategic choices.

We want to strive to ask four questions.

Understanding: Do people understand that they are data creators and how the data they create fits into the bigger picture?
Empowerment: Are there mechanisms for people to voice concerns, suggest potential improvements, and make changes? Do you provide psychological safety so they do so without fear?
Accountability: Do people feel pride of ownership and take on responsibly to create, obtain, and put to work data that supports the organization’s mission?
Collaboration: Do people see themselves as customers of data others create, with the right and responsibility to explain what they need and help creators craft solutions for the good of all involved?

Foster a Data-Driven Culture

Fostering a data-driven culture is essential for organizations seeking to leverage the full potential of their data assets. This cultural shift requires a multi-faceted approach that starts at the top and permeates throughout the entire organization.

Leadership by example is crucial in establishing a data-driven culture. Managers and executives must actively incorporate data into their decision-making processes and discussions. By consistently referencing data in meetings, presentations, and communications, leaders demonstrate the value they place on data-driven insights. This behavior sets the tone for the entire organization, encouraging employees at all levels to adopt a similar approach. When leaders ask data-informed questions and base their decisions on factual evidence, it reinforces the importance of data literacy and analytical thinking across the company.

Continuous learning is another vital component of a data-driven culture. Organizations should invest in regular training sessions that enhance data literacy and proficiency with relevant analysis tools. These educational programs should be tailored to each role within the company, ensuring that employees can apply data skills directly to their specific responsibilities. By providing ongoing learning opportunities, companies empower their workforce to make informed decisions and contribute meaningfully to data-driven initiatives. This investment in employee development not only improves individual performance but also strengthens the organization’s overall analytical capabilities.

Creating effective feedback loops is essential for refining and improving data processes over time. Organizations should establish systems that allow employees to provide input on data-related practices and suggest enhancements. This two-way communication fosters a sense of ownership and engagement among staff, encouraging them to actively participate in the data-driven culture. By valuing employee feedback, companies can identify bottlenecks, streamline processes, and uncover innovative ways to utilize data more effectively. These feedback mechanisms also help in closing the loop between data insights and actionable outcomes, ensuring that the organization continually evolves its data practices to meet changing needs and challenges.

Build Data as a Core Principle

Focus on quality: Emphasize the importance of data quality to the mission of the organization
Continuous improvement: Encourage ongoing refinement of data processes,.
Pride in workmanship: Foster a sense of ownership and pride in data-related tasks, .
Break down barriers: Promote cross-departmental collaboration on data initiatives and eliminate silos.
Drive out fear: Create a safe environment for employees to report data issues or inconsistencies without fear of reprisal.

By implementing these strategies, organizations can effectively tie data to employees’ daily work and create a robust data culture that enhances overall performance and decision-making capabilities.

FDA Continues the Discussion on AI/ML

Many of our organizations are somewhere in the journey of using AI/ML some where in the drug product lifecycle, so it is no surprise that the FDA is continuing the dialogue with the recently published draft of “Considerations for the Use of Artificial Intelligence to Support Regulatory Decision-Making for Drug and Biological Products.”

This draft guidance lays out a solid approach by using a risk-based credibility assessment framework to establish and evaluate the credibility of AI models. This involves:

Determining if the model is adequate for the intended use
Defining the question of interest the AI model will address
Defining the context of use for the AI model
Assessing the AI model risk based on model influence and decision consequence
Developing a plan to establish model credibility commensurate with the risk
Executing the plan and documenting results

I think may of us are in the midst of figuring out how to provide sufficient transparency around AI model development, evaluation, and outputs to support regulatory decision-making and what will be found to be acceptable. This sort of guidance is a good way for the agency to further that discussion and I definitely plan on commenting on this one.

Semantic Meaning

Over on Squire to Giants, Steve Schefer, writes about the semantic drift of the word triage in business talk.

What’s in a word? Triage…

I think it can be a really valuable exercise to consider, and align on semantic meaning of words, even words that may seem to everyone to mean one particular thing, and triage is a great example of that. When we spend time agonizing over words in documents, arguing about glossaries, what we are doing is aligning over semantic usage for terms that may have drifted a lot.

And don’t even get started on cultural appropriation of words.

The technical nature of our work means that semantic change, which is already a natural and inevitable process in language evolution, is going to happen. Words that we regularly use acquire new meanings or shift in their usage over time. Look what we’ve done to the poor word leverage or pipeline for just to examples.

Like data, we need word stewards, the keeper of the glossary. This role is in service to the process owners to enforce them agreeing on terms and using them the same way as possible. This is why I strongly believe in central glossaries. The dangers of not doing this can be impaired communication, with the message being lost or misinterpreted. And that leads to inefficiencies, and errors, and history has shown us those errors can get pretty significant.

FDA Draft Guidance on “Considerations for Complying with 21 CFR 211.110”

Usually I expect the FDA to publish some basic primer material as a webinar, so I was a little surprised when “Considerations for Complying With 21 CFR 211.110” was recently published as a draft. I’ve been rereading it, looking for what is actually worthy of a guidance here, and quite frankly, struggling.

It literally is a refresher course on 21CFR211.110. Maybe I should read it as “No we were serious about ICH Q8 and critical quality attributes.” Or maybe it is just the result of one too many bad Type C meetings lately.

Anyway, good refresher on product quality, in-process controls and samples. Still I think this would be better as a webinar with some graphics. Maybe I’ll better understand why this was published based on what sort of crazy comments are made and I can scratch my head and wonder what shenanigans some of these companies are up to.

Understanding Policies

The article “Research: Do New Hires Really Understand Your Policies?” by Rachel Schlund and Vanessa Bohns in HBR does a great job discussing consent that I think have real ramifications on GMP training, especially of the read-and-understood variety. Really gets me thinking on GxP orientation and the building of informed consent.

GxP New Hire Orientation

Building Effective Consent

1. Transparent Communication

Provide clear, detailed information about the why’s.

2. Staged Introduction

Instead of overwhelming new hires with all that GxP training at once, introduce them gradually over time. This approach gives employees the opportunity to digest and comprehend each requirement individually.

3. Interactive Training Sessions

Conduct engaging training sessions that explain the rationale behind each major requirement set and allow employees to ask questions and voice concerns.

4. Regular Policy Reviews

Implement periodic reviews with employees to ensure ongoing understanding and address any evolving concerns or questions.

5. Clear Benefits Communication

Explain the benefits of each requirement to the employee and the organization, helping new hires understand the value and purpose behind the requirements.