Using electronic health records (EHR) for rigorous clinical research is finally becoming a reality. One may ponder what this has to do with standards; everything is the answer. When EHR vendors were initially approached with the ‘opportunity’ to add a new service for the research community, most of them figured out fairly rapidly that this was not a wise business decision. At that time, they realized that they would have to extract data and map them to every requested format, which varied by research study sponsor; or, the EHR system would need to be configured differently for each research study or research sponsor—not very feasible. In addition, there was a perception that the entire EHR would have to be validated to meet Good Clinical Practices (GCPs) and other regulations required of biopharmaceutical development companies such as 21CFR Part 11 in the U.S.—another relatively impossible option.
Now, however, standards and processes are coming into being to ease the burden on EHR vendors to conduct research, even regulated research. In addition, the increasing use of EHRs around the globe, and the disappointing results of other new technologies in not truly transforming the clinical research process, have strengthened the case for paving the road to demonstrate the value of using EHRs for research. There are a number of initiatives that can only be summarized briefly in this article, but readers are encouraged to seek further information on this subject. We finally have a real chance of succeeding in not only making it easier for EHR vendors and researchers to conduct studies but also in bringing additional value to clinicians and patients globally.
Streamlining research at the source (i.e. for the clinician) can increase the capacity for research and speed the cycle of research data acquisition and analysis to more rapidly inform clinical decisions. Friedman, Wong and Blumenthal  have referred to rapid use of research findings to inform clinical decisions as the Learning Health System. Standards and initiatives that are showing the way to such a System, which would leverage EHRs for research, include (but are not limited to):
- an EHR Clinical Research Functional Profile;
- key standards/integration profiles that comprise an Interoperability Specification documented by the Health IT Standards Panel (HITSP);
- the EHR4CR Program sponsored by the Innovative Medicines Initiative (IMI) in Europe.
The EHR Clinical Research Functional Profile was developed jointly through the eClinical Research Forum and a PhRMA EDC Task Force. This functional profile took into account the GCPs and global regulations to define the requirements for an EHR to do regulated research. This profile passed HL7 ballots to become an ANSI/HL7 standard. It was then reviewed and modified to become a set of criteria that EuroRec has published and uses in certifying EHRs for the purpose of research. Most recently, because there is interest in having one global functional profile (not two different profiles for the U.S. and Europe), the Joint Initiative Council (JIC) for Global Harmonization of Standards (which includes ISO, CEN, CDISC, HL7, IHTSDO, GS1) (http://www.jointinitiativecouncil.org) has recommended that this be a JIC project to meet global needs for an EHR Clinical Research Functional Profile upon which certifications can be based.
The HITSP Interoperability Specification has an interesting history in that the clinical research use case for EHRs was the last use case recognized through the HITSP process (i.e. by the American Health Informatics Community-AHIC) and there was actually no funding left for this use case. The research community, however, demonstrated their support for this project through contributions from 35 organizations, including academic research institutions and biopharmaceutical industry stakeholders. The use case was selected by a multidisciplinary research committee and developed to support the production of a core standard research dataset from EHRs. This was considered a ‘foundational’ use case upon which others (such as safety and pharmacogenomics for personalized medicine) could build. The three standards/integration profiles selected to support this use case through an Interoperability Specification (HITSP IS # 158) are
- the Continuity of Care Document (CCD), from the HL7 CDA and the CCR, constrained to a Clinical Research Document (CRD) ;
- the CDISC-IHE Retrieve Form for Data Capture (RFD) (http://wiki.ihe.net/index.php?title=Retrieve_Form_for_Data_Capture), which is an integration profile and a workflow enabler based upon requirements in the CDISC eSource Data Interchange (eSDI) document ( http://www.cdisc.org/esdi-document); and
- the CDISC CDASH (Clinical Data Acquisition Standards Harmonization) standard (http://www.cdisc.org/cdash).
CDASH represents a minimum core standard research dataset that has been developed through a global consensus-based standards development process and is based upon regulations and requirements for data common across all research studies. This Interoperability Specification (IS # 158) has been demonstrated in multiple interoperability showcases (IHE-HIMSS, DIA-CDISC-IHE). Along with a site-based eSource data archive with audit trail (in the CDISC ODM standard) and relevant IHE security profiles, this IS can support regulated clinical research processes without requiring that the entire EHR be validated. The FDA Office of Scientific Investigation (OSI) has expressed interest in observing a research study done by EHRs (at least at some sites) using these standards and the IS and eSource archive. The beauty of a standards-based approach is that the research sites for a given study do not all have to use the same data collection system to produce standard data (CDASH datasets) that can readily be aggregated into a common operational database.
The EHR4CR project (http://www.ehr4cr.eu) is a very large and exciting initiative that was selected from competing proposals by the Innovative Medicines Initiative (IMI). It was launched this year with 33 participating organizations, including ten global biopharmaceutical companies, and 21 academic institutions, non-profit organizations and small-medium enterprises. This team is developing a platform and processes to support the use of EHRs for four research use cases: protocol feasibility, subject recruitment, clinical research conduct and safety reporting.The EHR4CR project team held a 3 day meeting this past October in Frankfurt, Germany with an energized crowd of participants who are simultaneously working on a variety of work packages relating to requirements, a technology platform, semantic interoperability, confidentiality, pilot site validation and business modeling. These activities are managed through a central project management office and there is oversight through a Steering Committee, an Ethics Board and a high-level Advisory Committee. This project promises to take into account all of the various stakeholders who will be impacted by such a cross-cutting initiative that will involve the whole of Europe as well as global research communities. The goal is to ensure that this benefits not only biopharmaceutical companies, but academicians, clinicians and certainly patients. The results will also contribute to next-generation standards in a few years time.
Another initiative of this kind, in the US, is the SHARP project for the Secondary Use of EHR Data, based at Mayo Clinic, Intermountain Healthcare and with other partners.
A vision that was documented in slides at least 15 years ago --at least as early as 1997, the year CDISC was formed as a volunteer group--- is finally coming true. Standards that provide a common ‘target’ for formatting and accepting data from EHRs and ‘profiles’ to identify functional requirements and to streamline workflow processes are making it increasingly attractive and easy for EHR vendors to support true research (i.e. not just data mining projects). In addition, researchers are realizing that data must be generated easily by investigative sites, without time-consuming and error prone transcription; otherwise, many researchers will continue to do one study and no more. The opportunity to use a variety of different technologies at clinical sites to support a standard set of research data for sponsors is increasingly attractive for all stakeholders. Researchers are finding many ways to leverage EHRs to improve the entire research process, and patients will benefit from a more rapid use of health information for research to, in turn, inform clinical decisions.
Watch for more news about these important initiatives in the coming months.
Author: R. D. Kush
- Friedman CP, Wong AK, Blumenthal D. Achieving a nationwide learning health system. Sci Transl Med. 2010 Nov 10;2(57):57cm29.