One Day, We Will Have Layered Datasets for High-Nuance, High-Impact Outcomes Research
To articulate the impact of clinical interventions in the real world, HEOR teams need to power their analyses with many kinds of data. Clinical, administrative, economic, and other types of real-world data (RWD) may all add value when generating insights into patient journeys and understanding treatment adherence in practice.
During Komodo Health’s The One Day Summit, Komodo Health’s Chief Operating Officer, Aswin Chandrakantan, MD, hosted a breakout session to dive into these topics and more. He was joined by Kristen Hahn, PhD, MPH, Head of Real-World Evidence (RWE) Research at Picnic Health, and Robert Sanchez, PhD, Executive Director of HEOR at Regeneron Pharmaceuticals, as well as a number of Life Science leaders, who provided valuable perspectives.
The conversation focused on how we can integrate multiple data sources to support HEOR teams in developing real-world treatment recommendations. Below are some of the key takeaways from the discussion.
Integrating multiple types of data provides HEOR teams with the most robust picture of patient journeys.
Participants agreed that different types of data generate different types of evidence and therefore may or may not be fit for the purpose of answering different research questions. While navigating the nuances of datasets can be a challenge, it’s crucial to combine various data types in order to create a full picture of real-world treatment scenarios. The conversation included real-world examples of linking commercial, lab, EHR, claims, and patient-reported outcomes data to bridge gaps in understanding and help teams answer questions about patient journeys holistically. These insights enable teams to match patients with the right clinical trials more efficiently, accelerate patient access to new treatments, and facilitate more effective conversations with providers.
When working across data types, tokenization helps connect patient journeys to unlock new insights.
Through the process of tokenization, artificial intelligence algorithms assign unique “tokens” to de-identified versions of patient information, allowing researchers to track patient journeys across datasets without compromising privacy. With the proper patient consent obtained, tokenization allows researchers to track their health over time. Leaders discussed the potential value of this tool, especially when working with multiple linked datasets.
While tokenization may not yet be widely accepted for regulatory purposes, it offers a key first step for HEOR teams to explore data comprehensively and visualize what’s possible for the patient population or medical intervention they are studying.
Life Sciences leaders must decide on the right organizational models to facilitate data management and effective use of RWE.
As the demand for RWE continues to grow, it’s essential for any Life Sciences company to elevate RWE experts within their organization to ensure high-quality insight generation. Depending on an organization’s needs or existing structure, leadership may decide to consolidate RWE expertise into a center of excellence (CoE) or distribute roles across teams. The discussion included the pros and cons of each model: A centralized approach to RWE simplifies questions of data access and ensures a unified approach to RWE generation, while a distributed approach may require a larger budget for data purchasing across teams. Regardless of the structure, understanding how to manage RWE effectively enterprise-wide is a key piece of the insight-generation puzzle.
As regulators and payers continue evaluating where and how to use RWE in review processes, the industry can help set the bar for high-quality evidence.
Regulatory agencies and payers are turning to RWE increasingly to inform decision-making, but it will take time for these bodies to reach conclusive decisions about how and where to use RWE. Participants shared current operational roadblocks to regulatory RWE generation: For example, the FDA requires sponsors to conduct prospective studies with 10 years’ worth of data comparing patients’ experiences with their drug versus a competitor’s — an unreasonable timeline for most sponsors, and one which delays key insights. As industry, payers, and regulators continue to learn together, Life Sciences companies can help set standards for RWE by exploring how to navigate issues of data confounding or gaps in RWD analyses. As one participant pointed out, the use of RWE doesn’t have to be “all or nothing,” so the industry can also test how RWE can complement or enhance findings from randomized controlled trials to create more robust evidence packages.
Patient engagement and decentralized trials are key to developing representative datasets and comprehensive evidence.
A high burden of disease does not translate to clinical trials being available. Decentralized trials — which are operated remotely and can involve patients across geographies — provide a unique bridge to health equity and representative insight generation. RWE can help teams match patients to the right decentralized trials, but they must also consider how best to engage patients. Patient consent is table stakes, but it’s also important to establish trust by being transparent about how their data will be used to improve outcomes.
One Day: More Impactful HEOR Analyses
As discussed during this session, integrated datasets combined with top-notch analytics capabilities allow HEOR teams to unlock deeper findings about patient journeys — which could ultimately impact patient care and access to treatment.
There is still work to be done to define how to best use and operationalize RWE within Life Sciences organizations and among regulators, payers, and patients. But, as we look ahead, our teams are eager to see how efforts to deliver more comprehensive HEOR insights will transform our industry.