There’s a lot of excitement building right now about the possibilities of using data derived from electronic health records (EHRs) in pragmatic clinical trials (PCTs)—and for good reason. EHRs, after all, not only represent a potential “treasure trove” of data for researchers,[i] but are now sufficiently ubiquitous that the long-cherished vision of the efficient, affordable pragmatic trial seems to be within reach. Indeed, we are already seeing the first mature instances of this approach in the field. PCORnet’s ADAPTABLE aspirin trial[ii] serves as a notable example, as does the National Institutes of Health’s Collaboratory project, which is demonstrating scaled-up versions of PCTs[iii],[iv] that leverage EHRs as the primary instrument for data collection in outcomes studies and comparative effectiveness research. Such approaches are designed to answer important research questions quickly and with relatively minimal infrastructure.
However, it’s important that these genuinely exciting and impressive forays into EHR-driven pragmatic outcomes trials not be allowed to eclipse the full potential of the EHR as an instrument of research and quality improvement. And while it’s great that pragmatic trials may usher in a new era for discovery, unless we get systems of care that help us consistently apply what we know to patients, such discovery will be for naught. Thus, we need to find ways to tap into the efficiencies of EHR and learn to apply these data and IT systems across the entire spectrum of research activities.
Leveraging EHR data across the research process
EHRs offer assets at every stage of the research process, starting when the investigator is in the planning stages of developing a clinical study. The data stored in these records can be leveraged for basic feasibility analysis and to examine research sites’ levels of readiness to participate in a study. EHR data can also prove invaluable for identifying and refining potential study cohorts and for characterizing their interactions with their respective health systems. In addition, these data are highly useful for developing study eligibility criteria that strike an optimal balance between rigor and inclusivity.
With regard to the actual study, EHRs can be used to identify patients who may be eligible for a study. At my home institution, Duke Health, we are actively deploying and refining tools that leverage EHR data to identify potential research cohorts and to automatically alert investigators when potential study participants have encounters with the health system. EHR data can also be used directly to replace or augment clinical information needed for a study. Finally, investigators can use information from electronically stored systems to limit follow-up requirements and for event ascertainment. Below, we will expand on these potential roles of EHRs in the clinical research process.
During study setup
The potential of EHRs as a tool for enabling more efficient randomized clinical trials (RCTs) grows as projects transition from concept to the first stages of execution. As a trial moves into the study setup phase, the cohort identification and recruitment capabilities that are so valuable for establishing basic feasibility can be applied to focusing on local populations of interest, further refining the potential study cohort and modeling outcomes. Tools designed to take advantage of enterprise data can be distilled into a “feasibility dashboard” for researchers that simplifies, both conceptually and practically, the process of study setup within a given site or health system.
And while much discussion of EHR capabilities focuses on the technology as a source or substrate for data extraction, it’s important to remember that the flow of information is bidirectional. For instance, the same basic capabilities used to create recruitment and feasibility dashboards can also be applied to embed encounter instructions and other content into the EHR, including site- or system-specific informed consent or pre-consent.
As noted above, participant recruitment is an area particularly ripe for EHR-based strategies. Multiple health systems are moving forward with toolkits that incorporate study screening criteria directly into EHRs. This facilitates nearly every aspect of the process, from contact and recruitment to visit scheduling. And, as we and others have demonstrated, these same systems can be configured to provide highly refined and specific alerts to investigators and other study personnel whenever potentially eligible subjects have an encounter with the health system.
Further, the same near-ubiquitous health portal systems that now afford patients access to their own health records can be used to allow patients to opt-in or opt-out of a wide array of studies. This particular application has already attracted significant attention in the world of outcomes-based PCTs, where less conventional study design elements such as cluster randomization are especially relevant, but the advantages of such capabilities apply broadly across the range of RCTs, observational studies and quality improvement programs.
Once a trial has commenced, EHRs continue to offer potential for streamlining multiple aspects of trials conduct and operations. First and foremost, they allow trials-specific data to be captured efficiently at the point of care delivery without the need for the extensive parallel data-collection systems that have in the past been used for many RCTs. By linking the EHR and the case report form, key data fields can be automatically populated at the time of entry. Automated validation and quality assurance checks can then be applied to ensure that data are complete and meet quality standards. Such linkages can be used to capture and report relevant data about hospitalizations and adverse events, as well as collect information about clinical event rates. Study data can also be extracted in ways that facilitate and expedite the tasks of the study coordinator and other support personnel.
And just as researchers are already leveraging the potential of EHRs to facilitate cohort identification and patient recruitment at the outset of clinical trials, they are also finding them useful as instruments for retaining and maintaining contact with study participants throughout follow-up. This is an especially important application, considering the logistical challenges of maintaining adequate follow-up over long intervals (such as those typical of cancer research) or with research participants whose circumstances present challenges for study follow-up. They can also serve as vehicles for patient education,[v] with information tailored to each patient’s specific circumstances and needs.
Realizing the promise of EHRs in the research environment
Taking an expansive view of EHRs as a fundamental resource for efficient clinical research—a tool that can be applied to everything from basic feasibility analysis and cohort identification for a small first-in-human study to a nationwide pragmatic outcomes trial and to everything in between—will allow us to realize the fullest potential of this technology. Indeed, EHR-based research is already thoroughly ensconced in safety surveillance and in epidemiological and observational research. In my own field of cardiology, numerous registry and quality efforts are tapping into EHR data.[vi]
However, as technologies and methodological approaches mature further, we will be able to leverage EHR data to dramatically expand capabilities in the worlds of outcomes research and implementation science. In short, we believe that the EHR is relevant and applicable across the entire spectrum of research activities, from RCTs to epidemiological studies to large-scale post-market surveillance studies to implementation and finally to policy evaluation. At the same time, however, it’s important to remain mindful of the challenges that will need to be addressed.
Despite the roadmap created by the standards of meaningful use and the incentives embodied in the 2009 HITECH Act and the recent issuance of a draft FDA guidance on EHR use in research contexts, it is clear that we are not yet at the point of seamless integration of EHRs with the world of clinical research. For one thing, a common complaint heard among investigators is that many EHRs are not optimally configured for conducting research—a problem further exacerbated by differences among the various platforms being adopted across the healthcare landscape. Many of the solutions for leveraging EHRs for research applications have been “home-brewed” at different institutions and health systems and necessarily represent parochial, not universal, approaches to specific needs and problems. And finally, although there is widespread understanding of the need for true interoperability, data standards and “electronic phenotypes,”[vii] the work in this arena is ongoing.
Still, all of these challenges, significant as they are, are ultimately solvable. Healthcare and clinical research are moving into the era of Big Data. There are a number of ways to respond to the rising flood of information: We can be overwhelmed by it, we can run from it—or we can learn to channel and harness it effectively for the benefit of patients.
Dr. Eric Peterson is the executive director of Duke Clinical Research Institute (DCRI) and a professor of medicine in the Division of Cardiology at Duke University Medical Center. He served as vice chair for quality in the Department of Medicine from 2004-2010.
[i] Zozus MN, Richesson R, Hammond WE, Simon GE. Acquiring and using electronic health record data. In: Rethinking Clinical Trials: A Living Textbook of Pragmatic Clinical Trials. Available at: https://sites.duke.edu/rethinkingclinicaltrials/acquiring-and-using-electronic-health-record-data/. Accessed October 19, 2016.
[ii] National Patient-Centered Clinical Research Network (PCORnet). ADAPTABLE, the Aspirin Study – A patient-centered trial. Available at: http://theaspirinstudy.org/. Accessed October 19, 2016.
[iii] NIH Health Care Systems Collaboratory. Demonstration Projects. Available at: https://www.nihcollaboratory.org/demonstration-projects/Pages/default.aspx. Accessed October 25, 2016.
[iv] Landro L. The ultimate battle against MRSA. The Wall Street Journal. September 12, 2016.
[v] Agency for Healthcare Research and Quality. Module 17: Electronic health records and meaningful use. In: Practice Facilitation Handbook. Available at: http://www.ahrq.gov/professionals/prevention-chronic-care/improve/system/pfhandbook/mod17.html. Accessed October 19, 2016.
[vi] Cowie MR, Blomster JI, Curtis LH, et al. Electronic health records to facilitate clinical research. Clin Res Cardiol. 2016 Aug 24. [Epub ahead of print]
[vii] Richesson R, Smerek M. Electronic health records-based phenotyping. In: Rethinking Clinical Trials: A Living Textbook of Pragmatic Clinical Trials. Available at: https://sites.duke.edu/rethinkingclinicaltrials/ehr-phenotyping/. Accessed October 19, 2016.