Paying research participants — a lot — may be a key to increasing diversity in studies

Paying people to participate in medical research has been proposed as a way to narrow racial and ethnic gaps in who signs up for clinical trials and other research studies. But a new study finds that if the amount is too small, it could have the opposite effect.

Offering $100 to participate in a Covid-19 study enticed affluent and white people while doing little to increase the participation of people from low-income or non-white households. A $500 incentive closed the participation gap completely, however, according to the study, which was released today by economists seeking ways to improve equity in clinical research.

Earlier this year, a report by a National Academies of Sciences, Engineering, and Medicine committee found that the exclusion and underrepresentation of many disadvantaged groups remained a widespread problem in medical research and that “large swaths of the population, and those that often face the most challenges, are less able to benefit from discoveries.” The report stated that little research existed on solutions for boosting participation, and it urged all sponsors of human research to provide adequate compensation to participants and called on federal agencies that oversee research to provide explicit guidance about equitable compensation to study participants and caregivers.

To test whether cash payments could help, the economists mailed invitations to participate in a survey to detect Covid antibodies, an indicator of past infection with the SARS-CoV-2 virus, to nearly 900 households in neighborhoods across Chicago. They chose communities in zip codes that varied by both income and racial and ethnic makeup.

The invitations were randomized to either include no incentive, a $100 incentive, or a $500 incentive for those who agreed to take a small sample of their blood at home and mail it to the researchers in a prepaid envelope.

The researchers found that participation was low but similar to other mail-in studies — just 6% — when no incentive was offered, and most participation in this group came from higher-income and white households. When $100 was offered, overall participation increased to 17%, but the incentive widened the gap between groups: More than 20% of higher-income and white households participated, while participation of groups from lower-income and non-white households was about 10%.

Offering a $500 incentive narrowed the gap almost completely; participation in all households increased to nearly 30%, which is considered a high rate for mail-in surveys.

Many studies have linked lack of participation in research to people from disadvantaged communities not being invited to be part of trials, perhaps because they do not receive medical care at institutions conducting studies. The new study sidestepped this issue by sending invitations to a variety of households, allowing the team to explore why some people may be hesitant to participate, or think the risks outweigh the benefits.

The fact that $100 did not vastly increase participation in lower income households “was super striking,” said Michael Greenstone, a co-author, professor of economics at the University of Chicago, and director of the school’s Becker Friedman Institute, which conducted the study. “I think it speaks to some real discomfort in responding to these surveys, which is not equally spread across populations.”

The study began during the Covid pandemic “when there was an all-hands-on-deck feeling and we wanted to find a way to help,” said Greenstone. “As economists, we were not going to come up with a vaccine, but we thought we could help answer public health questions.”

The study has been released as a working paper and is not yet peer-reviewed; the authors plan to publish it in an economic journal. (Posting working papers to solicit feedback before publishing is common in their field, the authors said.) But they said the true value of their work may be in prompting outside-the-box thinking on how to reduce racial and income disparities in who is included in medical research.

“We think about incentives and how people make decisions all the time,” said Winnie van Dijk, a coauthor and assistant professor of economics at Harvard who worked on the project while she was a postdoctoral researcher at the University of Chicago. “I think it’s always good when people look across fields.”

Van Dijk said the study could help researchers better understand what causes problems of inclusion in medical research, or in other important studies, such as Census surveys conducted to determine the amount of federal funding that is sent to low-income areas. More research on possible solutions is needed, she added. “Financial compensation is one way to overcome hesitancy. It’s not necessarily the only way,” she said.

Many researchers and institutional review boards, which ensure studies are conducted ethically, are uncomfortable with paying participants, despite the fact the practice has a long history: In 1900, military surgeon Walter Reed paid people who were willing to be bitten by yellow fever-infected mosquitoes $100 in gold (an amount worth roughly $9,000 today) to take part in studies, and $100 more if they became infected.

Though paying subjects is quite common, it remains contentious, with some scientists seeing it as wrong or coercive, writes Christine Grady, a bioethicist with the National Institutes of Health. In a review article on the topic, Grady wrote that more research was needed to assess both the ethics and impact of paying volunteers, including answering questions about whether high payments could unduly influence subjects, lead to exploitation of vulnerable people, or increase medical distrust.

As an economist, Greenstone is aware that resources are finite and researchers must think carefully about how to budget their funds and determine both whether to offer financial incentives and how large they should be. But they must take some measures to ensure their results are representative of the entire population, he said. “Financial incentives,” he said, “seem a powerful way to do so.”

Arleen F. Brown, a general internist and professor of medicine at UCLA who served on the National Academies committee that produced the report on increasing diversity in clinical research, said that she found the new study interesting, but that her work showed that a number of factors other than money could influence who participates, such as transportation, child and elder care, and compensating people for taking time off from work. Yet researchers don’t always think about the needs of participants.

“The most important thing is asking people who you want to participate what they really need,” she said, adding that one key step is for researchers to partner with community organizations that know and understand the populations they serve.

Brown said she learned a valuable lesson from community leaders while working on a study on how to increase diversity in Covid-19 vaccine trials. They pointed out that recruitment fliers stating participants needed “documentation of a negative Covid-19 test” would be off putting to people who are not citizens or legal residents. “They told us that word is a deal-breaker and will keep people from participating,” she said. “Words matter.”

Her study, published last October, also found that in addition to providing transportation and culturally sensitive recruitment materials, providing good “customer service” — being welcoming to participants, thanking them, giving them tours and introductions to testing staff, and supplying them with water, snacks, or goodie bags — helped increase participation. Her study noted that other factors that may help but be more difficult or expensive to achieve include providing research sites closer to participants’ homes and providing materials in a wide variety of languages.

This story has been updated to correct the amount that $100 in gold from 1900 would be worth today.

This is part of a series of articles exploring racism in health and medicine that is funded by a grant from the Commonwealth Fund.