Clinical research in the United States does a poor job of including patients from minority groups. Credit: David Edwards/Natl Geogr./Getty

The clock is ticking for experts charged with designing a US government programme to collect genetic, physiological and other health data from one million volunteers over the next two decades. The plan for the effort, part of the US$215-million Precision Medicine Initiative (PMI) announced in January, is due in the next few weeks — a daunting deadline, in part because the effort’s priorities include filling racial and socio-economic gaps left by other long-term studies.

The US National Institutes of Health (NIH), which is leading the PMI, is weighing ambition against a desire to avoid the mistakes that torpedoed its National Children’s Study, which would have tracked 100,000 children from birth to adulthood. The agency abandoned that effort in December 2014, after recruiting just 5,700 participants at a cost of US$1.3 billion. It cited overly complex study design and goals.

Clinical trials in the United States have historically relied on enrolling white participants from higher socio-economic levels, despite the fact that ethnic minorities make up about 40% of the country’s population. Of the 58,160 lung-disease studies published between 1993 and 2013, for example, less than 5% reported the inclusion of participants from minority ethnic groups (E. G. Burchard et al. Am. J. Respir. Crit. Care Med. 191, 514–521; 2015). The disparity is especially problematic because many diseases are more prevalent among certain ethnic groups, and ethnicity may also influence which therapies are effective, says Esteban Burchard, a physician scientist at the University of California, San Francisco.

“I would argue that it’s a scientific question that needs to be addressed; it’s not about the social reasons,” adds Burchard, a member of the PMI working group. The team has decided to over-represent minority groups in the study relative to their share of the US population. Doing this should increase researchers’ ability to draw statistically significant conclusions about small groups.

For example, alcoholism is particularly prevalent in Native American communities, and a study such as the PMI could help to reveal genetic and environmental factors that might underlie this vulnerability. But Native Americans make up just 1.6% of the US population, and if they were represented proportionally in a 1-million-person study, that would amount to just 16,000 participants — and focusing on subgroups determined by socio-economic status or age would further reduce the sample size.

Sarah Gehlert, who researches health disparities at Washington University in St. Louis, Missouri, hopes that the PMI will focus not only on ethnic minorities but also on other under-represented groups, such as poor people and those in rural areas. When minorities are included, they tend to be upper-class, educated and urban. But recruiting and retaining members of under-represented groups presents a challenge — they may not have the resources to find information about the study online, and may lack experience in using the fitness trackers and mobile apps with which the PMI plans to collect physiological data.

Just because you can study patients at an ivory-tower academic institution doesn’t mean you can do it in rural Appalachia.

The NIH has not yet decided how it will recruit PMI participants. Gehlert is concerned that the agency could rely too heavily on patient-advocacy organizations such as breast-cancer support groups, which tend to attract white, affluent city dwellers. She also notes that the PMI plans to use data from medical records. Because poor people are more likely to seek care at emergency departments than to have regular doctors, their records are often fragmented.

Such people are also more likely to be distrustful of researchers. “Just because you can study patients at an ivory-tower academic institution doesn’t mean you can do it in rural Appalachia,” Burchard says. With this in mind, the NIH is consulting researchers experienced in recruiting under-represented groups into clinical trials.

At the PMI working group’s meeting on 2 July, public-health researcher Donna Antoine-LaVigne of Jackson State University in Mississippi talked about her work with the 5,300-participant Jackson Heart Study, the largest survey of cardiovascular disease in African Americans. It includes both urban and rural populations, and has relied heavily on health workers going into their own communities to recruit participants.

Although this approach is labour-intensive, Antoine-LaVigne believes that it is cost-effective. “Having people on the ground that do this would cost a lot less in the long run, because otherwise you’re taking investigators away from their research,” she says. “And a lot of them don’t have a clue about bringing folks in.”

Striking a balance between community-based approaches and conventional research studies at hospitals or universities is a priority for the PMI working group, says its co-chair Bray Patrick-Lake, who works in patient engagement in research at Duke University in Durham, North Carolina. The NIH has not decided how to allocate the project’s resources, but “I don’t see this as a landscape for only the traditional players in research”, she says.