Duke & PCORI Use EHRs To Transform Clinical Trials

By Ed Miseta, Chief Editor, Clinical Leader

PCORI (Patient-Centered Outcomes Research Institute) and the Duke Clinical Research Institute have launched the ADAPTABLE study, a patient-centric trial designed to assess the benefits and long-term effectiveness of aspirin dosing. With a budget of over $14 million, the trial will randomize and dose 20,000 patients while leveraging health systems, EHRs (electronic health records), and patients in an attempt to fundamentally transform clinical research.

Every year, over 720,000 Americans suffer heart-attacks and 380,000 die of coronary artery disease (CAD). While new therapies for CAD are needed, an increased use of aspirin is believed to have a positive impact on the rates of death and disability. Unfortunately, the dose of aspirin that would result in the lowest risk of ischemic and bleeding events has not been determined. The main objective of the ADAPTABLE study is to identify that optimal dose.

The trial will incorporate several essential aspects of the new genre of patient-centered comparative effectiveness trials, including:

• The use of existing data sources to gather baseline characteristics and a combination of existing data and patient reported outcomes during follow-up
• An Internet portal that will enable the collection and monitoring of data and allow for learning by both patients and clinicians
• The elimination of a placebo, instead treating patients with different doses of the active treatment
• The collection of patient reported outcomes (PROs)
• Use of the PCORnet infrastructure to streamline administrative aspects of the trial, including centralization of institutional review board (IRB) functions and contracts, electronic consent, and the use of EHR and claims data

“PCORI is investing over $300 million into a clinical research network that will unite patients, clinicians, health systems, and EHRs,” says Dr. Adrian Hernandez, associate director of the Duke Clinical Research Institute. “The network itself has medical information and data on over 100 million subjects, and the patient EHRs are mapped to the PCORnet Common Data Model (CDM), which essentially organizes data into a standard structure. That network will make ADAPTABLE the first pragmatic clinical trial to use the network. The trial will select patients via their EHRs, reach out to them electronically, consent them electronically, and then follow up with them electronically.”

Patients with heart disease will be identified electronically via inclusion/exclusion criteria based on a phenotype. Patients identified as suitable for the trial are then contacted and can learn about the trial via a web portal. After answering a few comprehension questions, they can enroll and be randomized to receive one of two doses of aspirin. Follow-up is also performed electronically via a patient portal as well as through data from the participating healthcare systems.

“The 20,000 patients will be identified across approximately 35 health systems,” says Hernandez. “For most trials that size, you would be looking at a global effort with 10 or 20 times the effort. By having the ability to identify and contact a large number of patients electronically, we are able to eliminate the traditional site approach.”

For a typical cardiovascular trial, Hernandez notes you might expect to recruit one patient per site per month. By making use of EHRs, the ADAPTABLE trial is looking to recruit 50 patients per site per month. The effort has been underway for a little over two months and is being done in phases to ensure recruitment is progressing as expected. The first two sites have already enrolled almost 60 patients in those two months. Over the next three months the trial will scale up to include the other sites as well.     

“We have pretty ambitious goals,” says Hernandez. “We are looking to rethink how we perform clinical research. Instead of waiting for patients to come into a clinic where a study coordinator just happens to be on duty, we are reaching out to the patients directly. When they come into a clinic, they will already know about the trial and can enroll if they have not already done so online. The very first patient who enrolled in the study did so from his home using an iPad. The consent process was created with patients and meets all regulatory requirements. Because the health system is reaching out to patients, no third-party needs to see or access patient information.”

Different facilities will use different systems to store patient information. Even when two facilities use the same system, they can be customized to individual needs. Part of the success of the PCORI network is that it is agnostic to the system used in health centers. “We don’t care what they’re using,” says Hernandez. “The centers are simply asked to map a common data model.  That places everyone on the same common playing field with a common language. We now have over 80 systems that have done that.”

Hernandez likes to state that good data is necessary, but it is not the only requirement to conducting a successful trial. First you have to have the right patients. While the patient population required for a trial certainly exists out there, they can be hard to find. Therefore, he believes proper patient identification and engagement is critical to meeting trial enrollment goals.

“Patients were vital to the protocol design process,” says Hernandez. “They helped design the interface for the patient portal, and they were instrumental in helping to create the patient consent form, which ensured it was patient-friendly. They are also helping us to spread the word about why it is so important to participate in this study. That is a level of patient interaction that you don’t generally see in a clinical trial.”

The electronic patient interactions that were incorporated into this trial also helped to make it possible. Under normal circumstances, this type of study might see 100 eligible patients enter a participating clinic or health center every week. Hernandez notes there is no way a study coordinator would have the time to meet with each one of those patients to explain the benefits of participating in the trial and then get them registered for it. “Not only are we going directly to the patients with the details and the consent, we are relieving the study coordinators of that administrative burden so they can spend their time focusing on the patients themselves,” he says.

Hernandez is hopeful the ADAPTABLE study will change the way the industry looks at patient recruitment. The Duke Clinical Research Institute has a new trial starting this fall that will test influenza vaccines using the same principle of harnessing electronic health data to speed recruitment. “The CDM plays a pivotal role in EHR-driven research,” he adds. “When working with multiple institutions and systems, it provides the foundation we can build upon going forward.”