Moving Data for Clinical Trials: Standards, Systems to Link EHRs with EDC
By Allison Proffitt
June 16, 2022 | “Thirty to 40% of clinical trial costs are associated with actually just the manual retyping of [EHR] data [into the clinical trial system], and it leads to reporting delays and a number of other inefficiencies,” Adam Asare, chief data officer at the Quantum Leap Healthcare Collaborative/UCSF, told the Bio-IT World Conference & Expo audience last month. Asare shared a solution to that problem: a collaboration with OpenClinica that won the team a Bio-IT World Innovative Practices Award.
Quantum Leap Healthcare Collaborative is a 501c(3) charitable organization established in 2005 as a collaboration between medical researchers at University of California, San Francisco, and Silicon Valley entrepreneurs. Quantum Leap Healthcare Collaborative is the sponsor for the I-SPY-2 clinical trial, the longest-running and successful platform trial in breast cancer which aimed to improve speed, efficiency, and quality. Phase two of that trial was focused on adverse events and patient-reported outcomes when COVID-19 hit. “That’s when the FDA helped support the phase three project which is to apply these standards of data interoperability,” to COVID-19.
For the COVID portion of the I-SPY trial, the study looked at more than 3,000 critically-ill COVID-19 patients in 30 sites comprised of academic centers, community settings, and non-academic trial centers. One goal was to apply OpenClinica’s OneSource solution to address the inefficiencies of gathering these patients’ data into the clinical trial.
From the first time he and Asare met, Cal Collins told the Bio-IT World audience, they had a shared vision for less manual data transcription and abstraction in clinical trials. “We’re not the only ones to do that. There’s plenty of published literature on this; there’ve been many demonstrations and pilot projects over the years showing that if you can automatically acquire data you reduce time, you can lead to enhanced site experiences, create more efficient clinical monitoring and reduce the amount of monitoring that has to be done, potentially rapidly identify safety signals and increase data integrity and quality,” he said.
But Asare and Collins wanted a solution that not only seamlessly ingested data from the EHR into the clinical trial environment in a way that was easy, accurate, and secure, but also something that incorporated patient-reported outcomes and scaled easily. “How can we do that not just at one site,” Collins added, “not just in a way that’s only going to work in one trial, but in a way that’s repeatable and scalable?”
OpenClinica’s OneSource platform pulls data directly from the EHR into the electronic case report forms, “and from these we’re also providing information and decision-support to clinicians,” Asare explained. “We worked closely with OpenClinica in developing the interface. This launches directly within the EHR in a single-sign on, providing an integrated environment providing decision support.” The environment shows the clinical trial staff not only patient demographic data that stays static, but also updated summaries of research and clinical findings, including onset and duration of adverse events.
Standards to Scale
“OneSource is designed to be deployable across heterogeneous sites and trials,” Collins and Asare wrote in their Innovative Practices entry. OpenClinica chose two standards to serve as the connection between the EHR and the clinical data environment: SMART (Substitutable Medical Applications and Reusable Technologies) on FHIR (Fast Healthcare Interoperability Resource).
“SMART on FHIR is an open-source integration standard,” Collins explained at the conference. “These two things working together have reached a level of adoption among platform providers on the healthcare side—the electronic health record vendors—that we felt would be a good basis for an approach to integration.”
Wooing hospital IT and security staff has been the biggest challenge, Asare said. He noted that implementation of OneSource takes about 14-21 hours. “The only customization is creating the launch button within the EHR environment,” he said. “However, the negotiation and meeting with governance and security, can range from two months to 14 months… It takes a lot of advocacy from the investigator.”
Collins agreed that getting buy-in is tough, but said it has gotten easier. The close collaboration with UCSF was a strong start. Early sites were more challenging, but, “since we’ve built us some success stories, it’s gotten a little more streamlined,” he said.
One particular point of success, Collins added, was that OneSource and this pilot have focused on limited datatypes. “We didn’t try to boil the ocean. We started with low-hanging fruit in terms of data that is available in a structured format. We’re dealing with ONC-certified EHRs and the data that they’re mandated to provide to the USCDI standards [United States Core Data for Interoperability].”
He’s not ruling out more complex datatypes in the future, but, he explained, the first focus was handling this data in a scalable, secure, and most useful way. “We’re not trying to pull in unstructured notes and process them. We have designs to do more of that down the road, but dealing with the straightforward data, we’re able to be confident that we’re getting structured, discrete data in a very repeatable way.”