By Dalvir Gill, CEO of TransCelerate BioPharma, Inc.
Electronic Health Records (EHRs) have significant potential for improving the way we conduct clinical trials, although this potential has yet to be fully realized. EHRs were designed to support payers, providers and hospital systems, not clinical research. Indeed, EHRs have proven to be valuable in the doctor’s office and hospital setting, adding important efficiencies in a space that was once inundated with paper and burdened by overflowing files. Clinical trials, too, have a history of paper-based inefficiencies. But the drug development world, much like the doctor’s office, is changing: we are digitizing, we are sharing data, we are re-examining historical practices and identifying areas where we can improve. We acknowledge the opportunities EHRs could provide in a clinical research setting, as well as the challenges we’ll have to overcome when it comes to variety in systems and user-friendliness. Leveraging EHRs, with their large repository of patient data that have yet to be tapped for research, could ultimately be an important part of the technology evolution in clinical research.
Key ways that EHRs can positively impact clinical research.
The first is use of EHRs to support trial feasibility and protocol optimization. Increasingly, access to insights derived from large and diverse data sets are critical to ensure a study is practical in design and that appropriate patients exist. EHR data today represent an important source for feasibility and optimization insights, as sponsors can access aggregated de-identified data that reflect eligibility criteria and other underlying patient population data. The successful use of EHR data for feasibility requires access to appropriate patient data that matches the study population and the geographies of a given protocol, along with an organization whose culture is supportive of data-driven decision making.
EHRs also have great potential to impact clinical trial recruitment. A persistent challenge in clinical trials today, EHRs could help identify specific patients who may meet study criteria. Even where data may exist as unstructured free-text notes, natural language processing (NLP) technologies continue to mature allowing matches to occur even if the data is not tightly structured. The emerging use of artificial intelligence and cognitive computing will create even greater opportunities to match patients despite the density of free text notes. These technology improvements could create smarter EHRs, essential to elevating clinical research.
In fact, the EHR4CR project, which involved 35 academic and private partners and 11 hospital sites in France, Germany, Poland, Switzerland and the United Kingdom, has recently worked to develop a platform that can utilize de-identified data from hospital EHR systems, in full compliance with the ethical, regulatory and data protection policies and requirements of each participating country. EHR4CR reported that such a platform can significantly improve the efficiency of designing and conducting clinical trials, reducing time, costs and administrative burdens and enabling the participation of European hospitals in the more clinical trials.
While EHRs may be able to help find that a patient exists within a medical dataset, work remains to help achieve the “last mile” and enable EHRs to go from purely data in a database to a useful tool in the identification of appropriate patients that could be referred to matching study investigators and coordinators.
While an investigator may be able to view the identity of patients in their own practice who match for their study, the policies and laws within their medical system may not allow them to view the larger number of patients under the care of other physicians just down the hall. In many cases, only the treating physician for the patient may contact them if there is a possible study match, opening the issue of lack of motivation among treating physicians to support referrals into the clinical trials of others. Alerts and notifications embedded within an EHR can notify the provider of a study match, but without the right awareness and motivation, these alerts will fall on disinterested eyes – the average primary care physician is managing over 75 EHR notifications, consuming one hour of time every day creating a phenomenon being referred to as ‘alert fatigue,’[i] hindering the great promise EHRs hold for patient recruitment.
Perhaps the greatest use for EHRs in clinical research is the collection of patient data for prospective clinical trials. While 21st Century Cures legislation charts pathways for the Food and Drug Administration (FDA) to support regulatory decision making based upon historical EHR data as real-world evidence (RWE), prospective trials will also benefit from electronic sourcing of study data (eSource). Last year, the FDA issued a notice expressing interest in demonstration projects to test the capability and evaluate performance of using an end-to-end EHR-to-Electronic Data Capture (EDC) single-point data capture approach signifying the importance and potential positive impact of data collection in clinical trials. It is considered that EHRs, with digital applications and monitoring tools, can help “improve reliability, quality, traceability, provenance and integrity of data from electronic source to regulatory submission,” as the FDA notice on EHRs states.
Sourcing electronic data for trials via EHRs may bring great efficiency for study conduct (including data management and monitoring), but the diversity of EHRs across the thousands of sites supporting trials around the world challenge the ability to consistently pull and map data for prospective trials. Data standard initiatives along with new platforms meant to work across different EHRs may create opportunities across this otherwise disconnected ecosystem.
The final opportunity for EHR use involves patients opting to contribute their personal health data to clinical research. Meaningful use criteria contained with the Affordable Care Act and other legislation require that patients in the US can access their health data in the format in which it is acquired and maintained by their provider – if your doctor uses an EHR then you as a patient have a legal right to access your data electronically.
Multiple studies have found that over 90% of patients with access to their health data are willing to share that data to support research[ii]. With the launch of the Precision Medicine Initiative by the US National Institutes of Health, an expectation was set to enable over one million American to share their health data into a large cohort study. That Initiative brought together the country’s largest EHR vendors including Epic, Cerner, and Allscripts to work together to develop and pilot an application program interface (API) that will let patients share their health data for medical research in the “Sync for Science” Project. In addition, with clinical trials historically suffering from a lack of patient engagement in clinical trials, the API/Sync for Science program can get patients excited about clinical research as the industry moves to become more attuned to the needs of the patient. These programs contribute to the elimination of the silos that separate health data by EHR vendor and slow scientific progress. Activating EHRs in this collaborative manner and engaging patients to get involved can provide researchers with a mechanism to exchange data, and understand patient experiences with medical conditions, as well as clinical trial design. Ultimately, these advances will improve drug development inefficiencies, and deliver lifesaving drugs to patients faster.
The ability to collect and share data is only part of the solution. For EHRs to function as a vehicle for more efficient clinical trials, there are critical R&D stakeholders that must contribute to and collaborate on a potential roadmap to success:
Collaboration, as we know, is key to the transformation within a modern R&D community; we have seen that when we work together, we’re able to more effectively move the needle on clinical trial success. We’re shifting way from silos, and moving more toward an integrated research ecosystem. This will be especially evident when it comes to EHRs in clinical research, with each of the efforts and roles described above working in tandem to bring EHRs closer to generating meaningful change in clinical research. When that happens, we could realize the dream of end-to-end data flow. This will take time, determination and involve significant technological, regulatory and process challenges, but the change is inevitable, and welcomed.
Dalvir Gill, PhD, is the Chief Executive Officer of TransCelerate and serves on its Board of Directors. Dr. Gill has more than 25 years of drug development experience, and prior to his appointment as CEO of TransCelerate served as the President of Phase II-IV Drug Development at PharmaNet-i3, an international contract research organization.