Between March 10 and April 3, 2020, the RECOVERY team developed the protocol, obtained regulatory and ethics approval, and enrolled 1,000 patients.
COVID-19 presents an unprecedented clinical challenge to the health system, staff, and patients. Even moderate effects from this study will be worthwhile in the generation of robust evidence.
RECOVERY follow up will involve linkage to national data sources for vital status and death certificates; coded hospital episode statistics (diagnoses, procedures); intensive care audit data and SARS-CoV-2 PCR laboratory results; and primary care and national outpatient prescribing data.
Informed consent for participants in the RECOVERY trial consists of a 2-page information sheet and single form written in plain language.
The urgency of this pandemic requires everyone to focus on what matters and leave orthodoxy, habits, and traditional practices behind. Our mindset has been altered by the COVID-19 disruptions and this public health crisis of extraordinary proportion.
Researchers in ADAPTABLE pursued the unanswered question of whether a low dose or high dose of aspirin is optimal for secondary prevention of atherosclerotic cardiovascular disease. Complexities of running a randomized clinical trial and the expenses associated with it have previously prevented researchers from answering this question.
The capabilities of PCORnet have aided in this research. Though not all randomized clinical trials can be designed as PCTs, ADAPTABLE demonstrates the possibility of incorporating pragmatic elements into future studies with the goal of producing real-world evidence.
A key objective of involving PCORnet in ADAPTABLE was the ability for large-scale recruitment (15,000 participants) using electronic health records and electronic informed consent. The PCORnet component permitted complete electronic participation, from randomization to data collection. Patients reported their own data during scheduled electronic follow-ups, which replaced conventional follow-up visits.
NIH Collaboratory Coordinating Center co–principal investigator Dr. Adrian Hernandez also serves as co–principal investigator of ADAPTABLE and contributed to the article.
A new document from the Ethics and Regulatory Core is available that provides considerations around determining which individuals or groups are engaged in research in pragmatic clinical trials (PCTs). Developed for investigators designing and conducting PCTs as well as institutional review boards overseeing them, the document introduces these questions in relation to research subjects, study team members, and service providers:
Which individuals/groups are included in the research?
Are these individuals/groups research subjects, study team members, or service providers?
Why does it matter how the individuals/groups are categorized for the research?
Meeting minutes and supplementary materials are available that summarize discussions related to the ethics and regulatory issues associated with each of the UG3 PRISM Demonstration Projects. These discussions, which took place by teleconference, included representation from study principal investigators and study teams, members of the NIH Collaboratory Ethics and Regulatory Core, NIH staff, and NIH Collaboratory Coordinating Center personnel as well as some IRBs responsible for oversight of the projects.
Shivan Mehta, MD, MBA
Assistant Professor of Medicine and Health Policy
University of Pennsylvania
Pragmatic Trials of Behavioral Economic Interventions to Increase Colorectal Cancer Screening
Behavioral economics; Colorectal cancer; Pragmatic clinical trials; Health technology; Communication modality; Informed consent
Colorectal cancer is the second leading cause of cancer deaths in the United States. Increased rates of screening can reduce mortality from colorectal cancer by 30% to 70%.
Use of behavioral economics can help us understand human motivation and behavior related to participating in clinical studies. How the message to patients is framed—and how choices are offered—can alter the response.
In the example pragmatic trial, the “choice architecture” for the colorectal cancer screening was designed by the study team in collaboration with health system stakeholders and clinical operations. Changing the framing from opt in to opt out had the effect of increasing participation in screening.
In some settings, choice overload can have a negative effect on participation.
When designing embedded pragmatic trials, researchers must be mindful not to increase burden on clinicians’ workflow.
While behavioral economics offers suggestions for how to increase colorectal cancer screening rates, its effectiveness in different contexts needs to be evaluated.
Six Collaboratory Demonstration Projects—ACP PEACE, EMBED, GGC4H, HiLo, Nudge, and PRIM-ER—have recently transitioned from the planning to implementation phase of their embedded pragmatic clinical trial (ePCT). During the transition, study teams reviewed and updated their ethics and regulatory meeting minutes from discussions with the Ethics and Regulatory Core. The minutes describe ethics and regulatory issues the trials have encountered, along with approaches the trials are using for informed consent, HIPAA, and monitoring and oversight:
Ethics and regulatory issues can pose challenges to embedded pragmatic trials because of the unique nature of clinical research conducted in the setting of routine clinical care. The Ethics and Regulatory Core provides assistance to study teams as they navigate the ethics and regulatory landscape of ePCTs.
We spoke with Dr. Angelo Volandes, co–principal investigator of ACP PEACE with Dr. James Tulsky, at the NIH Collaboratory Steering Committee meeting in May about what the study team has learned during the planning phase of the trial.
Were there surprises during the planning phase of the study?
There were lots of surprises. The biggest surprise was that most clinicians don’t use the structured variable in the electronic health record (EHR) that we were going to use to extract our primary outcome. The workaround, which I think is actually better, is to use natural language processing (NLP) to abstract our primary outcome from the free text of the clinical note in the EHR.
The other big surprise was that oncologists really enjoyed the intervention. They have been open to the skills training and, if anything, they’ve asked for more.
What is an example of a challenge you were able to overcome with the help of the NIH Collaboratory Core Working Groups?
One challenge we encountered is related to an issue discussed in a paper by NIH Collaboratory investigators Dr. Kevin Weinfurt and Dr. Jeremy Sugarman and colleagues. It has to do with the idea of “broadcast notification.” One of our 3 participating healthcare systems asks patients if they will allow their deidentified medical record data to be used for research purposes. Every patient in that healthcare system has the option to opt out of having their deidentified data used for research purposes. Our other 2 participating healthcare systems don’t do that as a routine matter. So we needed a different approach.
The idea of broadcast notification—which is new and was developed in the NIH Collaboratory—is to display posters or other notices in healthcare settings that let patients know they can opt out if they have a concern about their deidentified data being shared for research purposes. Our local institutional review board (IRB) was unfamiliar with this approach. Having the Ethics and Regulatory Core help us understand the approach and educate our IRB was incredibly helpful. It was especially helpful to be able to share a published, peer-reviewed paper showing that this was an issue the NIH Collaboratory had studied.
There are always competing priorities in real-world oncology clinics. For example, there are quality improvement projects all over the place. When you’re the clinician, how do you devote the appropriate attention and time to this particular project? We feel our project is at the crux of patient-centered care, about understanding the goals, values, and beliefs of patients when it comes to serious illness care. But there are competing priorities. There can be a tension between the time you need to get the project done, for the intervention to truly reach its fruition, versus what a clinic might be willing to do.
What advice do you have for investigators conducting their first embedded pragmatic clinical trial (ePCT)?
It’s really important to get the appropriate buy-in from people in high enough positions of authority so that the project happens. It is not enough to get the chief research officer of a healthcare system to say the project is a great idea. You need the chief marketing officer, the chief executive officer, the finance people to sign off on it. When you’re in the pragmatic research world, it’s no longer just research in a controlled environment. It affects things you didn’t think about—like patient flow, revenue—and everything has to be accounted for. Make sure you get appropriate buy-in from enough stakeholders to know that you’re going to get the project done.
Also, don’t underestimate the costs of information technology (IT). For example, we need a lot more resources for our IT infrastructure now that we have switched from using a structured variable to using NLP to obtain our primary outcome. Make sure you have thought through IT needs, especially in pragmatic trials, where so much is abstracted from the EHR. Think carefully, early on, about how much time you will need from the IT group.
Anything else you want to say about ePCTs or the NIH Collaboratory?
It’s critically important to participate in the regular meetings of the Core Working Groups, to take advantage of them to help you address challenges. For example, when we encountered the problem with obtaining the primary outcome, we presented it to the EHR Core. When we had the challenge with notification and the IRB, we presented it during a meeting of the Ethics and Regulatory Core. The Core Working Groups are most useful when you openly share the challenges you are facing. That’s the way to get help from the Cores. This is my second pragmatic trial, and I’m not perfect. I put it out there because I want help from the experts.
ACP PEACE is supported within the NIH Collaboratory by a cooperative agreement from the NIA and receives logistical and technical support from the NIH Collaboratory Coordinating Center. Read more about ACP PEACE in the Living Textbook, and learn more about the NIH Collaboratory Demonstration Projects.
Findings from a new article suggest that the majority of patients do not feel a personal responsibility to participate in clinical research. In the article, Kevin Weinfurt, Li Lin, and Jeremy Sugarman report the results of a national survey of nearly 3000 people regarding their attitudes towards research responsibilities as well as their trust in doctors, healthcare systems, and medical research. Ethical frameworks for learning health systems have suggested that patients have a responsibility to contribute to learning activities, including research. The findings from this survey suggest that most patients in the U.S. do not currently endorse such a responsibility.
“These data provide a useful snapshot of the public’s views toward the obligation to participate in research. It is unclear how, if at all, these views will shift with increased efforts to create mature learning health systems. And if such views do not shift, it is uncertain what that would mean for the success of learning health systems.” —Kevin Weinfurt, PhD
Robert J. Mentz, MD, FACC, FAHA, FHFSA
Director, Duke Cooperative Cardiovascular Society
Associate Program Director, Duke Cardiovascular Disease Fellowship
Duke University Medical Center and Duke Clinical Research Institute
Good Clinical Practice Guidance and Pragmatic Trials: Balancing the Best of Both Worlds in the Learning Health System
International Council for Harmonization (ICH); Good clinical practice (GCP); Learning health system; Pragmatic clinical trials; Institutional review board (IRB); Research oversight; Regulatory issues; Quality by design (QbD)
Good clinical practice (GCP) guidance details the responsibilities, procedures, and recording that are necessary for appropriate trial conduct; for example, conducting the trial in accordance with an IRB-approved protocol with appropriate adverse event monitoring and reporting.
There is an urgent need to streamline randomized trials. Key obstacles are lack of transparency, lack of representativeness, and lack of evidence of competence.
In the United States, clinical investigators must abide by guidance from FDA, HHS, and ICH-GCP. Yet it is hard for investigators to keep track and to know how GCP applies to their study.
GCP as an overall construct is useful, but it does not deal well with issues particular to pragmatic trials or trials outside the FDA-regulated world.
With embedded pragmatic trials, informed consent is more nuanced. New considerations and approaches for consent have arisen since ICH GCP first came into effect.
Establishing quality by design will take time, effort, and educating IRBs to understand how QbD can be used to avoid errors in a trial and collect data that is fit-for-purpose.
It’s crucial that trials address an important question, answer that question reliably, and keep participants safe.