Making the most of RWD: Data strategy helps increase the value of analytics-generated insights

By Karla Feghali and Seshamalini Srinivasan

We are routinely asked by life sciences companies to help strengthen their technology capabilities and operations to make more effective use of real-world data (RWD), generally defined as patient data?that is routinely collected from sources outside of a controlled clinical trial. Examples include electronic health records (EHRs), claims and billing activities, product and disease registries, wearables, mobile devices, and other sources that can inform on health status.[1]

RWD is playing an increasing role in life sciences and health care decision-making across disease types, and its value—as well as the value of the technology that can analyze the important trends/patterns it contains—has been most recently apparent in the ecosystem’s response to COVID-19. RWD, combined with clinical trials data, has been used to identify how the coronavirus impacts patients, conduct COVID-19 vaccine effectiveness studies, and help regulators determine dosage timing and the need for boosters.[2] RWD also is being used to understand the impact of COVID-19 on treatment patterns for other therapeutic areas (TAs). Advanced artificial intelligence and machine learning (AI/ML) algorithms were used to increase the odds of vaccine trial success by identifying the best site locations in the United States to run the trials and recruit the right patient populations–ultimately enabling the vaccines to be available to the public at an accelerated pace. Compiling various sources’ RWD data on populations with COVID-19 (in particular, Israel) has produced one of the largest RWD sets ever available.[3]

Understandably, life sciences companies are making sizeable investments in licensing RWD to support disease understanding, clinical trial designs, augment trials with external control arms,[4],[5] and generate innovative treatments and devices. What is less clear is why many of these companies appear to be more focused on licensing different types of data from different sources than first thinking about the types of analytics-generated insights they want to gain from the data. Setting data strategy as an afterthought rather than a prerequisite can make it difficult to demonstrate value from newly achievable analytics. ?

Companies planning to license RWD can increase its effectiveness and value by proactively developing a data strategy that clarifies the types of data they're licensing and the subsequent analyses they want to conduct with that data. Important considerations include focusing broadly on data that generates evidence across indications, therapies, and use cases and allows hundreds of analyses to be run, and linking internal clinical trials data with RWD to answer novel questions at an accelerated pace. In addition, the strategy should address the following key questions:

• Maximizing data vendor partnerships: Have we conducted a landscape assessment of current and potential data vendors, academic institutions, disease-specific foundations, and niche players based on the types of analytics we want to conduct?
• Investing in resources to help structure the data: Are the investments we’re making in licensing data also helping to structure the data for use across a variety of fast-changing analyses, from R&D through commercial launch?
• Understanding the data supply chain: Where is the data we’re purchasing coming from? Does the vendor who is selling the data also manage where the data is being collected or does the vendor get the data second hand and have to clean and standardize it?
• Educating patients on the value of sharing their data: Are patients informed about how sharing their data can help advance science? Are they clear that their identify will be protected? Have they consented to use of this data?

More data, more availability, more potential

The quantity and types of RWD data—and that data’s availability—have expanded substantially over the last several years, offering ever-more potential for reaping value-added insights from data analysis. This is why it is important that the types of analyses company leaders want to conduct should determine their data strategy so they can a cost-effectively license the mix of foundational, therapeutic area (TA)-specific, and transformative data that best meets their needs.

• Foundational data—Large data sets, multiple TAs, less complex data models, good for exploratory analysis
• TA-specific data—Extensive data models, deep dive into the TA of interest (e.g., oncology)
• Transformative data—Often accessed and built through partnerships (both traditional and non-traditional/disruptive vendors); takes effort to structure the data to be usable; when done right, can truly differentiate the organization

Over time and across analyses guided by thoughtful data strategy, life sciences companies can assemble a robust, cost-effective repository of RWD from multiple commercial, academic, and non-traditional sources to enhance current offerings and enable future innovations.

In our next blog, we will discuss the importance of AI/ML in data analysis and how it presents a unique opportunity to increase companies’ speed and agility in bring lifesaving therapies to market.

____________________________________

References

[1] Real-World Evidence | FDA.

[2] “Real-world evidence confirms high effectiveness of Pfizer-BioNTech COVID-19 vaccine and profound public health impact of vaccination one year after pandemic declared,” Pfizer new release, March 11, 2021, https://www.pfizer.com/news/press-release/press-release-detail/real-world-evidence-confirms-high-effectiveness-pfizer.

[3] Jennifer M.?Polinski et. Al., Effectiveness of the Single-Dose Ad26.COV2. S COVID Vaccine. medRxiv. October 2020, Effectiveness of the Single-Dose Ad26.COV2.S COVID Vaccine | medRxiv.

[4] https://pharmaphorum.com/r-d/external-control-arms-real-world-data/.

[5] https://www2.deloitte.com/us/en/blog/health-care-blog/2021/can-external-control-arms-improve-clinical-trial-diversity-participation.html.