The Art of Being Data-Enabled: UCB’s Approach to Delivering Innovative Patient Solutions

As UCB’s Chief Data Officer, I combine my passion for data, analytics, and medicine to create meaningful solutions and innovative treatments for patients who put their trust in UCB. We use data analysis to answer questions and inform better decisions on behalf of patients and caregivers in need.

This work uses analytics, machine learning, and AI throughout our drug discovery, development, and innovation processes. For example, we’re optimizing the way we design clinical trials, selecting patients, automating the analysis of trial data, and optimizing our logistics operations. This all has a tremendous impact on how we work and how efficiently we can deliver results. And we’re expanding this work, building a platform to maximize the value of our data for the patients we serve.

Advancing the Health Solutions Lifecycle

UCB’s Data Office Strategy is built around four core principles:

• We are question-led and purposeful
• We determine fit for purpose data operating models in ways that are responsible and efficient
• We build strategically to unleash the full potential of data
• We create a strong technology foundation that connects people, systems, and data.
  

I want to focus on the first principle – we are question-led, which contrasts with most other pharma companies who are mostly “data-driven” – meaning they take large pools of data and try to see which learnings can be extracted.

We believe our approach better serves patients, focusing us on the answers to questions that are most important to them. Questions like, “How can we improve symptoms and how quickly?”, “What is the appropriate dose of a medicine?” or “Can we predict future patient response/relapse to treatment?”

UCB’s work with Stanford on severe diseases is a great example of our question-led approach. The collaboration leverages our expertise in clinical, real-world, omics, and other data sources to advance learnings in certain important areas.

Our first project with Stanford focuses on Hidradenitis Suppurative (HS), also known as acne inversa. HS is an immunological skin disease that results in debilitating quality of life for people living with the disease. The treatment journey is often long and complex with delays, misdiagnoses, and ineffective treatment. With our Stanford collaboration, we plan to further examine phenotyping, computational discovery of pathogenic mechanisms, as well as the disease burden and societal experience for people living with severe disease like HS.

The last decade has introduced new computational methods which increase the cost-effectiveness and efficiency in the pharmaceutical industry. The digitization of patient records and clinical trials, the advent of cloud computing, the growth of wearable devices, and the medical Internet of Things (mIoT) has transformed clinical development with an explosion of data. The opportunity and challenge is to remain intensely curious and creative as industry leaders in unleashing the potential of data.

For us, that’s taking form in the development of a state-of-the-art approach to large-data analytics that combines an AI approach with a Bayesian framework. In practice, it allows us to leverage the power of large data analytics even when the sample size is relatively small. In fact, our Statistical Science & Innovation group recently filed a patent application centered around an improved machine learning method that can search through a large hypothesis space to identify the most meaningful pieces of information and then use AI to predict an outcome or provide decision guidance. This innovation has already been used to inform patient stratification, clinical trial design, and precision medicine approaches for Parkinson’s disease and epilepsy.

Looking Toward the Future