Electronic health records (EHRs) aren’t simply paper-based files copied to computers. They are shared systems of medical records that give clinical researchers access to vast amounts of patient data. This puts a larger patient populations within their reach. It also allows researchers to better match patient eligibility with research areas, thus advancing the development of precision medicine.
Despite the promised benefits, however, integrating data from EHRs into clinical trials has not been as fruitful as expected. In this post, we explore some of the challenges using EHR data presents and how these can be overcome.
EHRs allow researchers in clinical trials to efficiently share patient data with primary care providers. That means data from patient outcomes and research requirements — lab reports, progress notes, adverse events reporting, and randomization of subjects — can be tracked and shared by using EHRs and electronic data capture, explain Adam Donat, Martin Hamilton, Irfan Khan and Nichole Chamberlain at Socra.
The benefits of this approach include real-time data analysis, reportage of adverse events, and more cost-effective studies. Another is that trial data can be quickly gathered from multiple sources and processed.
EHRs are necessary if researchers wish to advance precision medicine. This is why so much money and time has been invested in collating multiple data sources. The FDA and NIH have spent $1.45 billion to gather health records and the DNA of 1 million people in the U.S., writes Dr. John Danaher, president of clinical solutions at Elsevier.
Existing patient data will be treated like a large clinical trial, providing insight into how patients have responded to care over the years, explains Danaher. While EHR data adds complexity to compliance issues, he’s confident that the FDA will develop a suitable approval process regarding digital health and medical technologies.
Part of the appeal of EHRs is that they offer clinical trial sponsors and staff greater ability to match prospective patients with treatment areas. This is why the Medical University of South Carolina partnered with the health tech company TriNetX to develop an EHR participant cohort assessment tool. The aim of the tool, along with other TriNetX pursuits, is to use technology to optimize the use of real world evidence (RWE) in clinical research by accessing and analyzing longitudinal clinical data.
TriNetX has also helped use RWE on patients with diabetes to replicate results from randomized clinical trials, writes Samara Rosenfeld at iDigital Health. All of the existing data can be analyzed to determine patients’ responses to different treatments.
One example of what using an EHR network looks like is when TriNetX evaluated the effects of two drugs for the treatment of diabetes by analyzing tens of millions of EHRs. The company accessed medical data from 38 million patients in 35 different healthcare organizations across the U.S. It contrasted 47,000 patients who had taken a drug called SGLT2 with 189,000 patients who’d taken dipeptidyl peptidase 4 inhibitors.
The use of EHRs has been successful in certain instances. Switching from paper-based records has resulted in fewer medication errors, better adherence and improved patient safety. But there have been plenty of issues too, according to the Patient Safety Network.
Some of the failings include poor usability in terms of information display, complicated screen sequences and difficulty navigating, along with incongruence between user and clinical workflow. Software functionality, data entry errors and insufficient clarity in sources and date information are also on the list.
The quality of EHR can have an impact on tracking and following up abnormal test results. In a study examining the difference between a basic EHR and an enhanced EHR in a simulated EHR user environment, the former EHR’s interface was left unchanged while the enhanced EHR was manipulated to improve tracking of abnormal test results.
The study headed by Lukasz M. Mazur, Ph.D. analyzed physicians’ interaction and follow up of abnormal test results. Those using enhanced EHR systems performed better and picked up more abnormal test results than those using basic EHR systems.
While EHR data provides much promise, its use has not always been effective. The first misconception is that data from EHRs is not gathered following the same protocol as that of clinical trials, says John Seeger, chief scientific officer at health services and innovation company Optum.
Consider that EHR data, which comes from routine care patients receive, won’t necessarily have captured all of the relevant data concerning a patient’s received care or treatment. Also the standards of care might differ. So trial researchers need to reconsider the way they analyze this data, he explains.
“To effectively use EHR, it’s important to recognize that patient services that do not come from providers who contribute to the EHR database will not be visible in the same way that they are in a closed system, such as a health insurance claims database,” adds Seeger.
To overcome these challenges, researchers should consider contemporaneous cohort designs with relative incidence measures and be sure of the outcomes they are looking for.
Real-world evidence depends on the data from EHRs and is vital to designing trials and developing treatments — especially in the field of precision medicine. Indeed, former FDA commissioner and resident fellow at the American Enterprise Institute Dr. Scott Gottlieb refers to the wealth of data available to researchers and healthcare providers through EHRs.
And technology such as AI and machine-learning algorithms can help maximize the data from EHRs for better development of medical treatments and improved accuracy of patient data.
EHRs were set to significantly change the way patient data is used in medical treatment but the reality has been different. Costly and difficult to integrate with research software and systems, EHRs need to work better, according to Thomas H. Davenport, Tonya M. Hongsermeier and Kimberly Alba Mc Cord at Harvard Business Review.
Currently, EHRs tend not to capture and present data efficiently, often causing clinicians timely delays. But technology — specifically AI — can improve these processes by increasing flexibility and intelligence of systems, the researchers write.
There’s a difference between RWE and real-world data. The latter needs to be accessed, evaluated and analyzed using standardized approaches in order to support regulatory decision-making and satisfy requirements that it is robust enough for clinical investigation, explains dermatologist Art Papier, CEO of healthcare informatics company VisualDX.
Clarifying this distinction is why recent FDA guidance on how clinical trial professionals should use RWD from EHRs is so important. Papier argues that the guidance has brought the agency closer to the demands set out in the 21st Century Cures Act.
What’s promising, Papier says, is that regulators have begun to broaden their focus on standardized data to include standardized approaches for gathering data from non-traditional sources. Doing so will improve the relationship between clinical researchers and healthcare services and lead to greater interoperability between systems.
Data from EHRs can provide an overarching view of patient populations and not just in randomized clinical trials. It can also match patients to treatments, determine patients’ interest in participation and pre-populate a case report form, says Wayne Kubick, chief technology officer of standards organization Health Level Seven International.
But the problem with the clinical trial industry is that it has been slow to adopt new technology. Unsurprisingly, deviating from current (and often historical) best practice can seem daunting and a risk-averse industry tends to be cautious. Kubick suggests that using a standards framework can help overcome some of the regulatory and temporal challenges.
He cites the Fast Healthcare Interoperability Resources as a key framework through which to access EHRs. Doing so will improve contact with patients as well as the data feedback loop.
Part of the value of EHRs is that clinical researchers can better identify patients that are eligible for certain treatments. And according to research from Laura M. Beskow, Kathleen M. Brelsford and Catherine M. Hammack at BMC Medical Research Methodology, most patients want researchers to contact them directly. Three-quarters said it would also be acceptable for researchers to contact patients via their physicians.
The result is promising, showing patients willingness to share this data and have it be used for improved treatment. The researchers say that trust and transparency of motives and data must remain paramount. But the belief is that patients would enjoy increased decision-making power — such as the ability to opt in and out — and better access to research and improved treatment.
Patient attitudes towards this recruitment approach should be encouraging to sponsors, which have had a long history of recruitment and retention challenges.
Bigger patient pools and improved understanding of patients’ eligibility are two key wins to be had from integrating EHR data into clinical trials. While challenges persist, new technology, FDA guidance and improved understanding from researchers is increasing the value of EHR data to clinical trials.
Images by: pandpstock001/©123RF Stock Photo, sheeler/©123RF Stock Photo, elnur/©123RF Stock Photo