November 1, 2023
The Evidence Base
In a recent presentation at the DIA Real-Word Evidence (RWE) Conference (October 16–17, 2023, Baltimore, MD, USA), Sonja Wustrack, Managing Director of Integrated Evidence Generation at OM1 (MA, USA), discussed strategies for optimizing evidence generation in clinical research. She described the OM1 approach to overcoming longstanding challenges through integration, flexibility and automation.
Wustrack began by highlighting the evolution in how observational research is viewed, emphasizing the need for accelerated progress in this field. Defining integrated evidence generation, she discussed both technological infrastructure and study design to support multiple research objectives over time. Integration is key, enhancing efficiency and enabling adaptability. Currently, integrated evidence generation frameworks leverage diverse real-world data (RWD) sources and innovative data collection methods to generate fit-for-purpose real-world evidence (RWE).
The presentation focused on three key objectives:
- Utilizing RWD and RWE in clinical research
- Exploring the role of registries in prospective RWD collection
- OM1’s approach to automated and integrated evidence generation
Utilizing RWD and RWE in clinical research
- Drug development is a time-consuming, costly and risky process
- RWD needs to be integrated to complement randomized controlled trials (RCTs) and enhance evidence quality
Wustrack highlighted the challenges of drug development, taking an average of 12 years to get an agent to market and costing over a billion US dollars. Almost 90% of molecules fail, with 30% after Phase II and 58% after Phase III. The chance of success is also compounded by operational problems across the lifecycle, including inefficiencies, low participation rates and a lack of access to new patient populations. There is opportunity for RWD to play a greater role in assessing benefits/risks for drugs that receive accelerated approval and for use in diverse patient populations that are not included in pivotal trials. She highlighted that RWE, which is more representative of the impact of the product and the evolving standard of care, can be complementary to RCTs and contribute to a robust evidence package.
One of the key issues being addressed by RWD and integrated RWE generation is a reduction in burden to all stakeholders, including clinical trial site administrators, clinicians, patients and caregivers. Wustrack noted that, to overcome these often-frustrating aspects in trial site processes, automation is key to reduce burden, while achieving large sample sizes to support efforts in diversity and generalizability.
Wustrack acknowledged the challenges associated with RWD, including diverse sources, lack of standardization and potential data quality issues. However, RCT data, while controlled, may not fully represent real-world scenarios. The presentation highlighted the vision of using existing RWD as a foundation for prospective data collection.
Moving on to discuss regulatory progress in RWD usage, Wustrack provided a timeline of guidances from the FDA since the 21st Century Cures Act of 2016, which underscores the evolving acceptance and integration of RWD in pre- and post-launch settings. The FDA continues to release guidances across different types of RWD sources that are available and how these can be incorporated into study programs. Wustrack specifically highlighted the guidance related to patient registries and electronic health records (EHRs) as mechanisms for RWD collection. She also noted that the use of RWE is being explored beyond the regulatory setting; for instance, by the Centers for Medicare & Medicaid Services, who are exploring the use of RWE in their decision-making.
Wustrack provided an overview of RWD and the foundations for its use, discussing the differences in complexity of this data, from well-structured EHRs to unstructured sources, and how these can be processed using technologies such as machine learning and natural language processing, to make the data useful. She explained that patient-reported outcomes (PROs), which provide the patient voice, are increasingly being collected and can be an important source of RWD to be used.
“PROs are an element of prospective data collection that can be added on to complement and supplement the RWD that exists; and that voice of the patient is really important regarding quality of life, effectiveness, tolerability and safety from the patient perspective… this is an emerging area that we see as really important to supplement other data elements being collected.”
Exploring the role of registries in RWD collection
- Patient registries have emerged as a crucial mechanism for gathering prospective RWD
- Over 10,000 registries are logged in ClinicalTrials.gov, demonstrating their growing significance
- There are challenges with registries, which are becoming increasingly ‘trialized’
There is an increasing awareness of using patient registries for several use cases; for instance, to generate RWD, support health systems, conduct patient-centered outcomes assessments and to provide the infrastructure for embedded or nested studies. However, whilst patient registries provide a rich source of data, they also face some challenges. This can include biased enrollment, operational burden, and limited flexibility, such as being unable to support multiple use cases.
“The original intent of registries has morphed into something that’s a little bit more challenging to work with. We need new approaches for registries with a focus on the integration of data sources, flexibility and automation, all while maintaining rigor around traceability of the data and the right processes, so the data can be ready for regulatory submission if needed.”
Wustrack outlined various use cases for RWE, including disease burden, adherence, utilization, regulatory purposes and safety, and emphasized the importance protocol-driven data collection whilst maintaining as flexible infrastructure to develop a framework for generating evidence to support decision-making in health care.
She shared a case study using OM1 technology; first, OM1 Origin™ for data acquisition. Data is ingested via site networks, including patient-consented data from electronic medical record feeds, structured and unstructured data, and prospective data collection that can be bolted on along with supplemental data sources.
Data is then processed via OM1 Engine™, where data is normalized, curated and standardized. Unstructured-to-structured data manipulation takes place and end points can be developed. All of which is traceable for regulatory submission. A process of enrichment also takes place, linking to other data sources that can be internal to the customer, external to OM1 and the customer, or include linkage to the OM1 RWD data cloud, which includes 300 million patient records. This volume of information can help manage issues such as data missingness.
Wustrack also highlighted the importance of expertise, and the need to have services to understand how to leverage the data and design the study to extract the information needed. This includes regulatory and registry guideline experts; study design and data analytics support; clinical expertise to put the information into context; technology and data privacy specialists; and project operations experts.
OM1’s approach to automated and integrated evidence generation
- Automation is a pivotal solution to:
- Reduce burden on stakeholders
- Enhance site and patient engagement
- Support diversity and generalizability
Wustrack concluded the presentation by summarizing the OM1 solution – Aspen. Aspen leverages OM1 Origin and OM1 Engine to revolutionize evidence generation by integrating diverse data sources and providing a scalable, automated solution. Wustrack highlighted the need to move from a linear approach to the concept of a research ecosystem, and the Aspen platform’s flexibility, efficiency and stakeholder engagement were highlighted as key strengths.
Speaking to The Evidence Base after the session, Wustrack called out the flexibility of the OM1 Aspen system, and how crucial automation is to reduce burden and make systems available at scale – something that supports stakeholder engagement:
“We can connect to sites of all different shapes and sizes, ranging from big academic medical centers who have resources to support stakeholder engagement but perhaps don’t leverage this, to small private practices that just don’t have the resources. The fact that, following an initial engagement, you don’t really need to think about it is hugely appreciated by participants, and, in particular, by study coordinators.”
She also highlighted the evolution current taking place in regulation:
“I think the most interesting part of all this is the evolving landscape around regulation and regulatory guidelines on the use of RWE – and the need for proof points to build from. And the more guidance and direction coming from regulators, the more important it will be to move things quickly.”
Overall, the presentation provided a comprehensive overview of OM1’s approach to evidence generation in clinical research. By leveraging integrated evidence generation, stakeholders can navigate the complexities of drug development, enhance data quality, and, ultimately, bring therapies to patients more efficiently.