Covid-19: How to Conduct a Community Study in a Pandemic

Epidemiologist Shruti Mehta has spent decades studying HIV and hepatitis C in Baltimore. She’s used to running cohort studies that follow thousands of hard-to-reach individuals over the course of years, tracking how the diseases spread. But when she set out to learn about COVID-19 transmission in Baltimore, the tried and true methods of recruiting subjects, building trust within communities, and gathering data in person were no longer options. 

“Most of what we have done in the past is through direct face-to-face outreach and word of mouth recruitment—things that we cannot do in these times,” she says.

So, Mehta and her colleagues face a new question: How do you study a community when it’s not safe to be around people?

Last spring, Johns Hopkins created the COVID-19 Research Response Program and designated nine subject areas urgently requiring study, from genetic sequencing of the SARS-CoV-2 virus to modeling effects of public health policies. Mehta, PhD ’02, MPH ’98, was asked to help lead a study on how COVID affects communities.

A lot of research was focusing on people who were positive, their contacts, and “really understanding the natural history of the disease, clinical outcomes, duration of positivity,” says Mehta. But there was less understanding of asymptomatic infections, less-severe cases, and overall prevalence of the disease within communities. Nor was much known about the pandemic’s mental health effects, economic consequences, and impacts on food security, says Mehta. 

Mehta and her colleagues—including project co-leads Jason Farley, PhD, MPH, from the School of Nursing, and Jacky Jennings, PhD ’03, MPH ’98, from the School of Medicine—were asked to develop a community study that could fill in those gaps.

Her team was after the big picture: “the total impact of the pandemic, not just the disease.”

They received $950,000 from the University and later $4.5 million from NIH. The funding provided a foundation for the two-year project, Community Collaboration to Combat COVID-19, or C-FORWARD. The study aims to recruit 1,300 households in Baltimore to get tested for COVID and participate in monthly follow-up surveys. 

From the very start, they had to adapt to the pandemic’s reality. Traditional methods like knocking on doors to recruit a representative sample from different census blocks of the city had to be tossed, Mehta says. The door-to-door method only has a response rate of around 15%; it’s even harder to recruit people if you can’t meet them face to face.

“We know that access to health care is not equal, and the more we can liberate patients to seek the health care they need on their own terms, the more likely we are to get an engaged population.”

So they sought professional help. “We actually worked with a marketing company, which is something we've never done as researchers,” Mehta says. The marketing agency Subject Matter helped the team develop an entire campaign that included postcards and letters, as well as announcements on social media and local radio stations, with messages emphasizing that participation can help Baltimore fight COVID-19. (In exchange for joining the study, participants will receive a $50 stipend.)

C-FORWARD’s community engagement team—led by Jacky Jennings and Suzanne Grieb, PhD, MSPH, a medical anthropologist at the School of Medicine with a joint appointment at the Bloomberg School—also plans to work with leaders and organizations within the Baltimore community. The hope, says Grieb, is that C-FORWARD will become familiar to people and make these collaborations easier to initiate.

Public trust is key to a study like this, Grieb notes. It’s crucial for the researchers to include vulnerable community members who are at greater risk of contracting COVID-19, and part of reaching these community members entails building their trust in medical researchers. Collaborating with community leaders and being consistent and transparent are key—as is the fundamental principle of “respecting people just as people and not viewing them solely as participants in a research study.” 

The survey methodology, too, needed a different approach. Surveys are typically administered in a uniform manner: all self-administered written questionnaires or all phone interviews, for example. But an early pilot of this study convinced Mehta and her colleagues they needed to ensure “that we didn’t put up barriers to people participating.” Some low-income participants didn’t have reliable phone access because they ran out of monthly call minutes or data, for example. Some participants trusted printed surveys that were clearly official and from Hopkins, while others preferred the personal nature of phone calls. So, for the larger study, participants can choose to answer the survey questions by phone, by mail, or online.

The testing itself, meanwhile, is still happening in person for some. Farley led the logistics of setting up safe testing sites. The team piloted a testing protocol with an RV and a tent in the School of Nursing’s parking lot, with hourlong windows for each household of participants. As the study evolved, the protocol has expanded to include testing at clinics, at participants’ homes, and at mobile sites with the RV traveling to different Baltimore neighborhoods. Farley explains that the different options might give the team insights into which methods people are likeliest to participate in: “We know that access to health care is not equal, and the more we can liberate patients to seek the health care they need on their own terms, the more likely we are to get an engaged population.”

Mehta and her colleagues hope to bring some of C-FORWARD’s flexibility into future community studies. For instance, doing the follow-up behavioral surveys online rather than in person could prove less troublesome for participants. “I actually think online data collection ... requires a lot less in terms of infrastructure, so something like that is something I’d carry forward,” she says. 

As the pandemic evolves, the team is finding opportunities to learn more than what was envisioned by the project’s initial scope. For example, when the project was conceived, COVID-19 vaccines were still months away. Now, the vaccine rollout presents a chance to collect “longitudinal information on people's behaviors before and after they take the vaccine,” says Mehta.

In the meantime, though, this study could show how COVID-19 impacts different communities and serve as a prompt for addressing inequality in resources—during this pandemic and new ones down the line. 

“I hope that what we learn here informs us for the next time,” Mehta says.