W-01: Developing Research Partnerships with Health Networks for Real World Evidence Generation
Nancy A. Dreyer
Global Chief, Sci Affairs; Sr VP, Head, Center for Advanced Evidence Generation
IQVIA United States
Based on 9 years of experience in the US, we offer core lessons learned from building a productive network of health system partners to deliver real-world insights for diverse stakeholders.
Since its inception in 2009, the COMPASS network, a US-based group of integrated delivery health systems with de-identified EMR data, and physician and patient outreach capabilities, has evolved by expanding member institutions with different contract structures and capabilities.
COMPASS was initially developed as a distributed data network that repurposed electronic health records and administrative data to support real-world studies. The concept was derived from a desire to characterize continuity of care to capture a complete understanding of the patient population including healthcare utilization, as well as meaningful exposures (treatments and diagnostics) and clinical outcomes. Key partner characteristics initially included an integrated data warehouse, established IT infrastructure, and research staff available to support data extraction. A standardized query model estimates potential patient subgroups and the availability of data elements of interest as well as a post-hoc common data model to support creation of analytic datasets from multiple sites. As COMPASS evolved, contracts were revised and streamlined to fit institutional and research needs.
COMPASS eventually grew to 10 partners across 15 States and was further adapted to support enriched studies, which complement using existing data with customized primary data collection. Enriched studies can also support direct-to-patient, health care provider surveys, and pragmatic studies. Lessons learned through adaptation include clear roles and accountability for engagement with sites and providers and training required to better understand observational and pragmatic trial research at the site level. Indeed, a clear value proposition for partnership, tailored to site interests and capabilities, is another key determinant of success including meaningful research that benefits their patients and provides opportunities for thought leadership recognition.
Successful collaboration has been built through shared goals and transparency in the exchange of information, including sharing results that may influence medical care practice. The ability to link to additional data, whether within the health system via a health plan or external claims data from an insurer, adds value for multiple stakeholders, including the partners, by providing a more complete evaluation of patient care to support evidence-based decisions for improving patient care.
Establishing efficient pathways for study start-up and operation is critical when partnering with what are often large, complex health systems. Mapping out the key decision-makers and approval process is a helpful step to determine the critical path forward. Establishing an effective and flexible governance structure outlining roles and responsibilities for contracting, IRB approval, site training, clinic and physician recruitment, patient enrollment and primary data capture at both the central and local level is necessary. These processes are likely to differ for each health system and at times by therapeutic area focus, and are best to be conducted in advance of a specific study.
Large provider networks must meet evolving needs of research sponsors as well as their own needs and interests. A network developed to support primarily EMR-based studies looks different from a network that aims to conduct prospective studies with primary data collection and study-specified interventions. The balance of seeking information about how treatments work in “everyday” medical care has to be balanced with the sophistication of an institution in terms of provider and patient identification, mechanisms for their recruitment into studies, and a fairly rapid and reliable process for recruitment and data assemble and transfer.