covid-19

Frequent Rapid Testing May Reduce COVID Spread in Low- and Middle-income Countries

Brooke Nichols co-leads a new report and modeling consortium that has examined whether rapid diagnostic tests can mitigate COVID transmission in distinct circumstances, particularly in low- and middle-income countries.

Diagnostic testing remains a key part of mitigating the spread of COVID-19, particularly in low and middle-income countries where access to the vaccines is still limited. Health and science experts widely consider polymerase chain reaction (PCR) tests to be the “gold standard” for diagnosing COVID-19 infections due to the test’s high sensitivity and specificity, meaning its ability to detect true positive and negative COVID cases, respectively.

Although highly accurate, PCR tests are performed in labs and the turnaround time for results is longer than Antigen Rapid Diagnostic Tests (RDTs). They’re also largely unavailable in LMICs.RDTs are less accurate than PCR tests, but also less costly, and they can yield results in minutes. So the question becomes: do RDTs provide an opportunity to bring high-income testing capacity to LMICs, where PCR tests are scarce?

According to a new report co-led by a School of Public Health researcher, RDTs appear to be a valuable test in many settings—particularly in LMICs—due to the frequency and speed with which they can be administered and processed, and to their ability to avert outbreaks when combined with other mitigation measures.

Titled “Expanded Use Case Analysis for Rapid Antigen Diagnostics for SARS-CoV-2 Mitigation,” the report details the first phase of results of a COVID-19 rapid antigen testing modeling consortium that Brooke Nichols, assistant professor of global health, co-led with Colin Russell, professor of applied evolutionary biology at Amsterdam University Medical Center. The modeling consortium has served as the diagnostic arm of the Access to COVID-19 Tools Accelerator (ACT-Accelerator), an initiative that the World Health Organization launched in April 2020 with partners in academia, government, and the private sector to conduct COVID-19 research and develop strategies to distribute tests, treatment, and vaccines across the globe.

The report quantifies the impacts of RDTs on COVID-19 infection spread using different mathematical models in seven distinct settings (called “use cases”), including community testing, mass gatherings, K-12 schools, universities, border crossings, and testing to exit quarantine and isolation following contact tracing. The researchers aimed to understand when and in which settings RDTs can best be utilized to reduce COVID transmission, and determine the resources and policies that would support these goals.

“RDT tests have the ability to fill a large gap in testing between high-income countries and low- and middle-income countries,” say Nichols, a health economist and infectious disease mathematical modeler who has also conducted modeling of HIV and hepatitis C prevention strategies. “High-income countries have typically had the advantage of substantial PCR capacity to do testing from the start, and now we have technology that theoretically can fill the gap.”

The researchers assumed an RDT sensitivity rate of 80 to 85 percent, based on data from the Foundation for New Innovative Diagnostics. While not as sensitive as PCR tests, “what is important about RDTs is that the test is highly sensitive when a person is actually infectious—making them a very useful tool for pandemic mitigation.”

The researchers found that, in general, more frequent and widespread testing in most of the settings produced greater impact in terms of infections averted. Testing strategies were most effective when the Rt (which represents how quickly the virus is spreading) or the prevalence of COVID-19 cases were also low. Routine asymptomatic testing of individuals in a variety of settings may also substantially reduce COVID transmission within a given setting. 

However, an increase in testing frequency and population tested poses challenges for LMICs with very limited resources and funding for tests, the researchers caution.

“In order for population-wide routine community testing to be impactful, major resources would be needed,” says epidemiology PhD student Reese Sy, who designed the model for community testing (along with Nichols), which modeled the impact of community testing among nearly 59 million people in South Africa. “However, health systems in resource-constrained settings may have difficulty scaling up testing to a large percentage of the population, and thus population-wide testing may not be feasible. Testing resources may be better spent in other, more defined use-case settings where routine testing can be implemented, in order to maximize the utility of each test.”

The model for the mass gathering setting, where mass gatherings were defined as “any gatherings for which the number of people attending are enough to place additional strain on planning and response resources where these events take place,” was developed to estimate number of individuals who would be expected to attend a mass gathering while infected.

“One major use for COVID-19 testing is to stop infectious people from attending a gathering,” says Stephen Kissler, postdoctoral fellow in the Department of Immunology and Infectious Diseases at Harvard T.H. Chan School of Public Health, and who designed the model for this setting with Yonatan Grad, associate professor of immunology and infectious diseases at T.H. Chan. “There’s a huge range of ‘gatherings’ we might imagine: sports games, concerts, church services, flights—anything where a group of people comes together for a fixed amount of time.” 

The researchers found that higher COVID-19 case prevalence, longer event duration, and more time elapsed between testing and the mass gathering were associated with higher numbers of infectious individuals at these gatherings. Using RDTs to screen mass gathering attendees the day before, or day of, an event appeared to offer the greatest reductions in disease transmission at the gathering, compared with testing at earlier points in time.

“Our models show that it’s usually better to use the faster test,” Kissler says. “Since SARS-CoV-2 levels in the body can ramp up so quickly, a person can be undetectable three days before a gathering, but actively infectious when the gathering happens. Testing with a rapid test on the day of the gathering reduces this risk by a lot, even if the test needs more virus to turn positive.”

The model for the university setting was originally developed by Boston University researchers—including Laura White, professor of biostatistics, MPH student Joshua Chevalier, and Eric Kolaczyk, professor of statistics in the College of Arts & Sciences—to inform COVID-19 interventions for the fall 2020 reopening strategy. The model used a sample population of 3,681 faculty, staff, and students to project COVID-19 cases and outcomes within that group.

Interestingly, the model showed that an RDT strategy at a university would prevent the largest percentage of infections when testing is conducted only twice weekly—less than most of the other scenarios. When a setting’s infection rate is low, the results showed that testing once weekly or biweekly would be sufficient at preventing a comparable number of infections. White says that the size of the population may explain the different in testing frequency. “This model is a scaled-down version that has a very small population relative to a typical university, so that likely impacted some of the dynamics,” she says.

The results in each scenario discussed in the report only refer to the effectiveness of testing strategies within each individual setting. The next phase of the project will examine the potential effectiveness of each setting in different LMIC settings based on a variety of demographics, geographical settings, and available public health resources.

“What became clear from the modeling of all of the testing strategies was the need for one comprehensive model that can investigate the community-level impact of any use-case specific strategy,” says Nichols. “How can routine asymptomatic testing in K-12 schools affect community transmission? How can testing at mass gatherings affect community transmission? These testing strategies all form part of a comprehensive mitigation strategy—and understanding their synergies is essential in designing efficient programming.”

This report included modelling support from Boston University, New York University, London School of Health & Tropical Medicine, A*STAR Singapore, and Harvard T.H. Chan School of Public Health. Members of the working group included the World Health Organization, the Foundation for Innovative New Diagnostics, Clinton Health Access Initiative, the Health Economics and Epidemiology Research Office (HE2RO), National Health Laboratory Service in South Africa, and the Amsterdam University Medical Center.