DATAx presents: How Bristol-Myers Squibb is innovating pharma through data and AI

Ahead of DATAx New York, we spoke to Marisa Co, head, business insights & analytics - R&D at Bristol-Myers Squibb, to find out how the company is using emerging technologies to revolutionize patient care


Data, machine learning (ML) and artificial intelligence (AI) are unleashing newer and increasingly more effective capabilities for healthcare providers. Companies such as Bristol-Myers Squibb are sitting on the very forefront of a new age of patient care, discovering, developing and delivering innovative medicines that help patients prevail over serious diseases.

To provide insight into the groundbreaking work the global biopharmaceutical company has achieved over the last few years through deeper and more enriched utilization of data, we caught up with Marisa Co, head, business insights & analytics - R&D at Bristol-Myers Squibb, who is also delivering the keynote speech on Day One of the AI and Big Data in Pharma summit, part of DATAx New York festival happening this week at New York's Hilton Midtown.

DATAx: How have you used data analysis in clinical development programs at Bristol-Myers Squibb?

Marisa Co: Using data and advanced analytics is crucial to how we approach our clinical trials. We can segment patient populations, look for optimal treatment responses and potentially accelerate trial executions. Regulatory agencies like the US Food and Drug Administration (FDA) continue to work with companies like Flatiron Health, a healthcare technology company focused on accelerating cancer research and improving patient care, and other data aggregators to more deeply understand the depth and quality of de-identified patient level oncology data.

Additionally, their level of comfort with the use of real-world data for regulatory submission has increased. Bristol-Myers Squibb recently submitted an analysis in which we created a synthetic control arm (control arm using RWD instead of a traditional study) supporting a regulatory filing the FDA accepted. The synthetic control arm saved us two to three years where, in most cases, traditional trial paths may have caused a significant delay in filing. By expediting the process, we provided patients quicker access to an innovative therapy. The control arm enhanced our ability to deliver high-quality data to the FDA based on existing regulatory-grade real-world data.

Digital health is clearly transforming our industry, which accelerates our ability to deliver and execute value for patient care. Overall, deeper data analysis sets not only provide a profound understanding of the disease and patients in clinical trials, but it expedites the process.

DATAx: How have you leveraged data analysis to improve your R&D insights?
MC: Over the last few years, our analytical capabilities have enabled us to make better-informed and more insightful decisions that support Bristol-Myers Squibb's strategy. We've used analytics in our cancer research using our system to determine ways to improve patient care and outcomes. We did this by evaluating and leveraging real-world data to provide insight on patients outcomes, to support inquiries from providers, regulatory agencies and payers.

We use an integrated, diverse data platform in combination with advanced analytics to support robust study design and execution through an embedded decision-making process that aligns analytical groups with the business strategy. Access to a large number of real-world databases lets us geographically broaden and deepen research.

Additionally, in partnership with Flatiron Health, we're able to analyze larger genomic datasets and identify therapeutically relevant mutations to further develop remedies. This capability aligns with Bristol-Myers Squibb's goal to target the right therapy, to the right patient at the right time.

DATAx: How do you see advancements in AI and ML influencing the pharma industry in 2019?

MC: This area is continually growing and is a huge range of focus for our industry. For example, we anticipate that AI and ML, combined with deep linked and curated datasets, including imaging, will eventually help create something like an "AI-assisted tumor evaluation score".

From a manufacturing perspective, AI will help identify improvements in the drug development cycle and permit us to actively anticipate quality issues. We've only begun to use ML to design compounds and pre-assess toxicity, potency, metabolism and selectivity.

What is unique about Bristol-Myers Squibb is that we don't just "do AI" or get swayed by the buzz of it. Instead, we look at end-to-end processes and assess whether we could apply AI as a tool to close gaps. Thus, we deploy AI – or any other advanced analytical technique – as a means to an end, not as an end in and of itself.

We started to use AI to collaboratively clean up data, identify risks, stratify patient populations and/or find new biological paths and discover new compounds, which will push the industry further in the years to come. The future of collective technology advancements will (and has) started linking relevant business information to the discovery and development of therapies, which will assist with making better treatment decisions, drive healthier outcomes and patient experience, while maintaining patient privacy.

DATAx: How do you balance the importance of insights generated through data analysis when making business decisions?

MC: At Bristol-Myers Squibb, we have a centralized analytics group. The value of this approach is that we leverage multiple databases and analytics groups such as R&D, Commercial, Manufacturing and Enabling Functions to solve business issues.

The analytic groups partner closely with development teams and operational functions, and due to this process, we've identified key strategic business questions. We then use a variety of databases and complex analytical capabilities throughout the drug development and commercialization continuum; translational medicines, clinical development and operations, medical affairs, regulatory, value access and health economics.

As these experts and others execute our digital health strategy, they are supporting investment decisions and the development of insights, supported by four key capabilities:

  • Centralization and standardization of real-world data.
  • Integrated platforms.
  • New processes to access data and share knowledge across the organization to support data-driven decision-making.
  • Advanced analytical capabilities to drive standard methodologies, including working with partners and the FDA in using real-world data for regulatory submissions.

DATAx: Finally, what innovations do you see disrupting the pharmaceutical industry the most in 2019?

MC: Digital health is in its infancy in our industry. In 2019, we will experience:

  • More availability of data both in the US and globally.
  • A deeper use of digital to improve manual processes (such as data cleanup, study startup).
  • Enabling better patient stratification through a combination of clinical, genomics, pathology and PRO data.
  • Identifying new biological paths leveraging higher access and linkages of genomics, digital pathology, biomarkers and bioinformatics.
  • Identifying and selecting sites through larger datasets and better analytics.
  • Providing a better patient experience through e-consent, health apps and sensors to improve patient outcomes.

In addition, the true value of digital can only be unleashed through new and untraditional partnerships (such as health networks and data aggregators), which may require thinking differently about the type of value-based relationships we want to establish. These partnerships will evolve similarly to the ones established within various biotech or smaller pharma companies and will necessitate their own support infrastructure.

A major challenge we all face is the shortage of talent. Acquiring and developing talent is a significant challenge requiring meaningful investment and resources. Pharmaceutical companies not only compete among themselves for talent, but also with a thriving technology sector and other industries. As part of Bristol-Myers Squibb's digital efforts, we have developed a talent strategy and are currently working with our faculty acquisition team to incorporate the findings within our enterprise aptitude acquisition strategy. In addition, we have created the concept of a shared advanced analytics curriculum to upscale our internal analytics staff's capabilities.

Beyond 2019, new applications such as voice and facial recognition could also play a role in improving compliance or studying patient outcomes. Blockchain may be deployed in clinical research to prevent deviations and improve compliance, although this may need to be reconciled with other policies like the "right to be forgotten" within the EU's new General Data Protection Regulation (GDPR).

Digital health poses exhilarating new opportunities for the pharma industry and we've only just begun to reap the benefits of it. As we continue to develop new integrated platforms, increase the volume of data and partner with regulatory agencies, patient groups, payors and healthcare systems, we may be able to develop new innovative therapies, match the best treatment to the right patient and improve patient outcomes.

We expect that these efforts will eventually lower the cost of drug development and healthcare. We're excited about the future and how these new developments will unfold in the industry.

Marisa Co will be on a panel on Day One of the AI & Big Data for Pharma Summit, part of DATAx New York, taking place on December 12–13 at the Hilton Midtown. To attend and hear more great insights from other data experts from some of the biggest and most influential organizations, register here today before it's too late.

How big data can protect infrastructure small

Read next:

How big data can protect infrastructure