By Deborah Borfitz

April 26, 2021 | The pandemic has served to highlight the value of real-world evidence (RWE) in personalizing the delivery of care and potentially reducing disparities in access to treatment options, including clinical studies, according to Ingrid Oakley-Girvan, Ph.D., senior vice president of research at Medable. To explore this “new frontier,” Medable has contracted with Kaiser Permanente to develop mobile apps designed to capture insights from informal caregivers, and digital biomarkers from wearable devices with patients, in fully decentralized trials.

This should help improve diversity in studies by reassuring would-be enrollees that they are going to be safe in a trial between televisits, in addition to eliminating the participation barriers associated with studies conducted at bricks-and-mortar sites, Oakley-Girvan says. RWE can additionally be used to determine who is most likely to benefit from an experimental therapy, which could counter distrust of medical research that has been particularly widespread among African Americans.

Large healthcare facilities have been using the growing number of real-world datasets to generate RWE that supports the concept of diversity and inclusion in trials, she says, to ensure medicines work in a representative sample of people with the targeted disease. This, in turn, has highlighted “the necessity of offering more remote or decentralized types of trials so a broader, more diverse group of individuals can participate than typically has been the case.”

One of the biggest challenges in accessing trials currently is proximity to a clinical study site, Oakley-Girvan says. “If you live in a more rural location, you are less likely to have physical access to the site, it’s more costly to get there, [and] takes more time.” Individuals who are working two or three jobs trying to make ends meet have neither time nor money to spare, and single parents face added financial and logistical challenges.

Moving forward, her expectation is that RWE will be providing “greater insights and knowledge” for clinical studies starting at the protocol design stage. Most immediately, the biggest benefits are going to come from an analysis of real-world data (RWD) on different disease-specific populations to inform clinical trial eligibility criteria, says Oakley-Girvan.

In theory, this could result in less stringent eligibility requirements, so trials accommodate a greater diversity of patients. Trials would be designed from the get-go to factor in social determinants of health (SDOH) that exist in real-world environments and are known to impact outcomes.

Using ZIP code data, for example, it is possible to assess if someone lives in a food desert with limited access to healthful and affordable food options—or once did and is now suffering the repercussions, she cites as an example. Diet and exercise habits, factors impacting drug metabolism, might similarly influence how eligibility criteria are written.

“We are just starting to contemplate and evaluate the different types of data that can be combined,” she continues. People living near a freeway rather than the ocean, whose lungs are being overwhelmed with toxic particulates and experiencing more asthmatic episodes, might need a higher dosage of an experimental therapy or a companion drug. Various organizations are producing data from environmental monitoring of particulate matter to better describe the burden of disease by location, much as the tracking of UV levels has been used as a barometer of skin cancer risk.

Heath Profiles

Decentralized trials, on their own, are unlikely to completely solve the problem of uneven access to clinical trials, says Oakley-Girvan. But they are certain to be an “important and major contributing factor” to broadening trial entry for people who have no other treatment option, which is why minimizing the burden of participation is so important.

 As a scientist, Oakley-Girvan says, she wants to also use robust RWD to create better health profiles of individuals that capture the whole person, likening the endeavor to marketers who know which online sites have been visited and if people are looking at towels, candy bars, or housecleaning supplies. Adaptive trial designs are a good fit for studies seeking to account for differing conditions in that they allow for flexibility with dosing arms as studies unfold and RWE develops.

By linking different kinds of data to create robust health profiles, study sponsors can get a big-picture view of who might benefit the most from a drug under study as well as the dose they should receive—e.g., 5 micrograms or 1 microgram—based on their expected profile differences, she notes. If both groups are instead put on the lower dose, “you may end up seeing no impact because folks living near the freeway didn’t get any better because they had greater particulate exposure that required a higher dose.”

Back in the 1990s, Oakley-Girvan helped document disparities between African Americans and Caucasians when it came to risk of mortality from prostate cancer. While clinicians subsequently narrowed that gap, the situation might have improved sooner if patients were being monitored remotely, she says.

As with prostate cancer, the risk of comorbid condition development and early mortality from chronic conditions like diabetes can be lowered by more closely monitoring those who have the greatest risk and face financial impediments to healthy eating and intervening early to assist, Oakley-Girvan says.

This more thorough approach to remote monitoring is a direction Medable helps enable through decentralized trials and its exploratory work with Kaiser and others, she says.

App Development

Oakley-Girvan serves as principal investigator for a pair of studies, funded by small business innovation research awards from the National Cancer Institute. One involves development of a mobile app to empower care partners (e.g., a spouse, parent, or adult child) to better monitor patients and know the symptoms to watch for, as well as the importance of attending to their own health and wellness—which is critical to keeping patients enrolled in trials, she says.

“I think it will be pretty incredible if we can put that into clinical trials because there is often a concern about the side effects of drugs and we may not know all of them, but the care partner can help provide insight into that and… then the clinical team can respond back quite rapidly and directly on what to do,” says Oakley-Girvan.

“It’s a whole new way to help support clinical trials,” she says. This work is being done with subcontracts to Stanford, Duke, and Kaiser.

The other project, involving a subcontract with Kaiser, is “looking at the digital biomarkers patients themselves can provide through a different app, together with a connected device that captures specific types of information,” she continues. The benefit here will come from “longitudinal analysis of data over time so you are comparing to baseline.”

Changes attributable to a bad night’s sleep could thereby be teased apart from multiple bad days in a row that more likely need clinical attention, Oakley-Girvan explains. The response might be to prescribe a sleep aid or, if insomnia is a side effect of a drug in an adaptive trial, to lower the dosage.

Knowing a bad night’s sleep, and not a study medication, is causing patients not to feel well might prevent them from dropping out of the trial, she says. “A lot of folks would also like to participate if they know they will be a little more closely monitored or tended to.” This would include studies that use wearables to collect certain physical or physiological parameters from participants, in addition to surveys they complete between televisits.

It is a bit like giving people a fitness coach to help them achieve their goals versus having them go it alone at the gym, says Oakley-Girvan. “The idea is to reassure participants and also monitor them accurately… it’s a decentralized approach certainly, and COVID has really helped folks appreciate the practicality of it.”

Emerging Data

The American Medical Association has come out in support of RWD and RWE to improve outcomes, as well as telemedicine in managing patients with chronic diseases, says Oakley-Girvan. Many researchers and scientists are eager to use real-world data and evidence specifically to address health disparities, improving patient outcomes equitably across the board, she adds.

Medable’s science team is quantifying the impact of its work on the recruitment and retention of diverse populations, she says. Patient and caregiver factors influencing adherence to digital tools in trials are among the aims of a research study underway by the Digital Medicine Society.

In the remote monitoring arena, in particular, “solid data” is emerging for clinical trialists to think about, particularly when it comes to managing chronic diseases, Oakley-Girvan says. When COVID-19 hit, for example, cardiologists also had success mailing their patients wearable heart monitoring devices resembling large Band-Aids in lieu of routine checkups at the clinic.

A study published last year in JAMA (DOI: 10.1001/jamanetworkopen.2020.1074) found a remote patient monitoring tool helped predict 30-day hospital readmissions, discharge location, and length of stay with over 80% accuracy—in this case among cardiac surgery patients in a progressive care unit. The University of Michigan Hospital more recently reported that a home telehealth program helped it preserve hospital capacity through multiple COVID-19 surges, while decreasing readmissions and emergency department utilization among mostly 60-plus patients with multiple comorbidities.

“As we emerge out of COVID-19 and get back to some semblance of what we used to consider ‘normal’ interactions, I think that the medical community and the clinical trial community are really aligned in trying to support new ways to increase diversity and participation that is less burdensome,” says Oakley-Girvan.

Among study sponsors, a more decentralized approach is clearly part of that vision, she adds. In a survey recently commissioned by Oracle Health Sciences, 76% of sponsors report that they have already decentralized at least some of their trials.