eSource, the electronic sources of clinical research data, has been a goal of the industry for more than a decade and recently, this goal has made great strides towards reality. The reuse of eSource versus traditional data transcription in clinical trials has been shown to be associated with a significant reduction in data entry time and data transcription errors.
Often limited in scope to the reuse of EHR data for clinical research, in reality the sources of data collection for clinical research have multiplied with the growth of healthcare digitization. TransCelerate BioPharma’s eSource initiative outlines four core modalities:
Research sponsors are now receiving data for a single study not just from multiple sites across the globe, in compliance with multiple laws and regulations, but also from multiple digital sources. Furthermore, they return that data electronically to principal investigators, regulatory agencies and patients. The gold standard of randomized clinical trials is facing a digitization of clinical research leading to a paradigm shift unlike the industry has ever seen.
Despite its growth, it is important to highlight the challenges of scaling across the clinical ecosystem and propose mitigation strategies. Pharmaceutical companies and clinical trial sponsors, consortia, healthcare technology firms, government and regulatory agencies, standards developing organizations (SDOs), academic medical centers and clinical research staff all have unique interests, roles and responsibilities in scaling eSource to reach its full potential and maximize the benefits. It is essential that each of these groups has a seat at the table to ensure the right steps are being taken in this collective effort.
The industry has seen great progress since the inaugural eSource STARBRITE study in 2007 with contributions globally from thought leaders. In the past decade, government and regulatory agencies have supported the development and progress of digitizing research through policy, guidance and directly funding projects such as the Common Data Model Harmonization Project, OneSource and Duke Data FITS. SDOs such as Health Level Seven (HL7) are bringing the community together to create semantic and structural formats for interoperable exchange of data. The development of the HL7 FHIR® standard is a disruptive innovation that provides an avenue to not only reuse the data in the EHR but also to reuse the clinical care interoperability functionality, thereby laying the groundwork for scalability. The realization of full-scale production of an interoperable exchange of data for clinical research is within the grasp of the scientific and research community; however, this will require collaboration.
Representational quality issues refers to the gap between data collected and used for clinical care versus data for clinical research. Not only are there separate systems and processes for collecting and utilizing such data, the definitions and value sets often do not align. For example, research submissions to the regulatory agencies are often required in Clinical Data Interchange Standards Consortium (CDISC) standards, which often are not harmonized with clinical systems. Such misalignments create work to share data between clinical care and clinical research. For example, electronic data exchange from a clinical trial site’s EHR system to a sponsor would require data mapping to convert clinical care standards to CDISC. Beyond standard mapping, clinical trial protocols and data collection tools are often developed without direct knowledge of the way data is collected for clinical care, and therefore, subtle but fundamental differences (such as value sets) are misaligned, requiring additional translation and potentially even machine learning algorithms to gain alignment.
Fortunately, there are some current efforts to resolve this and close the chasm between clinical care data and clinical research data. For example, the Common Data Harmonization project, funded by U.S. DHHS, has successfully harmonized various data models used by key government research stakeholders to a single model called BRIDG. Additionally, the NIH CTSA institutions, through the project CD2H are striving to harmonize the varying data models used by sites. Also, funding by the U.S. Food and Drug Administration to map HL7 FHIR to BRIDG is another example of great work being done to bridge data collection between clinical care and clinical research. Furthermore, Learning Health Systems and American Medical Informatics Association (AMIA) have been engaging the community in clinical research informatics discussions for years. While each of these initiatives are producing great progress, these stakeholder groups are often working on similar if not the same use cases without visibility into each other’s work. The opportunity to bring these groups together may be the key to successfully closing this chasm.
eSource is a fast-paced, bleeding-edge area of clinical informatics. The environment is changing so rapidly that there are questions around the interpretation and differences in regulations across agencies, and specifically relating to clarity on expectations, requirements and auditing of these systems. For sponsors that have low risk tolerance, lack of data standardization and evolving regulatory requirements add a barrier to implementation. Sites and sponsors must ensure that issues such as maintaining audit trails, control of data, and data security are addressed. While the industry still needs to build trust in exchanging patient data electronically, paper records are also limited by lack of security and lack of an audit trail to identify unauthorized access to records.
Continued engagement of global regulatory and government agencies as partners will be key to the adoption and implementation of eSource and the digitization of clinical research. Inclusion of stakeholder perspectives as equitable partners in environments where open discussions and collaboration can occur is imperative to prevent one-off solutions and to collaboratively develop a single solution with enough flexibility to meet the varying use cases and prevent additional administrative burden to the sites while allowing for multiple vendors to engage in product development and delivery for the clinical research industry.
The entire paradigm of healthcare is shifting toward patient centricity and greater data interoperability. Reticence to change exists in all organizations and is understandable in an industry that directly contributes to patients’ health. Sponsors may be risk-averse because of time and resources necessary for implementation. Academic medical researchers may also have multiple initiatives and responsibilities that compete for time and funding. Vendors have multiple opportunities for product development. Patients may be provided with volumes of data, which, without additional context, may not always be easily converted to actionable knowledge for the patient. eSource is a bleeding edge paradigm shift for all stakeholders engaged in clinical research and healthcare.
The path forward is collaboration. These problems are not individual to any one stakeholder. Together, stakeholders can work towards an environment where eSource is leveraged to close the gap between clinical care and clinical research, building on and contributing to a rapid and continuous learning health system that benefits patients.
Many collaboration efforts are already occurring, are moving the needle forward and making eSource and interoperability part of the conversation. Stakeholders are coming together through groups like consortia and SDOs. Consortia serve as a place for people across organization types with similar interests to get together and discuss issues and solutions.
But there is still opportunity for further collaboration. The clinical research and clinical care data collection chasm needs to be bridged. Sponsors should work with patients, sites and SDOs towards technology platforms bridging this gap. Academic medical centers can partner with pharmaceutical companies to provide outcomes metrics showing efficiency of clinical trial processes and new technologies like eSource. Researchers can join and engage in consortia or SDOs to lead the conversation. Healthcare technology firms should seek to understand the full scope of their market and consumers and build technology platforms that address needs of each group recognizing opportunities in the concerns of other groups. Participating in events such as the HIMSS Interoperability Showcase can help vendors learn about customer specifications and potentially build out interoperable solutions.
Collaboration is key. Much progress has been made, and now is the time for stakeholders to come together and push for true interoperability. We need to all align on the path forward and we will all get there faster.
Sponsored content. The views and opinions expressed in this blog or by commenters are those of the author and do not necessarily reflect the official policy or position of HIMSS or its affiliates.
March 9 – 13, 2020 | Orange County Convention Center | Orlando, Florida
The world of health information and technology is evolving and so is HIMSS. Find out what's next for health and be part of the transformation.