Epidemiology: Underdetection of COVID-19 after the first lockdown in France

An estimated nine out of ten symptomatic cases of COVID-19 were not detected by the surveillance system in France shortly after the nationwide lockdown ended in May 2020. The findings, reported in Nature, indicate that the testing system did not achieve the detection rates needed to contain the pandemic, although detection levels did increase over time. Increased targeted testing with easier access is needed to make test-and-trace an effective tool for containing the pandemic, the authors suggest.

Strategies to detect and monitor the spread of SARS-CoV-2 are considered critical so that cases can be identified and isolated to prevent transmission within a community — in particular at low disease prevalence levels after lengthy and costly lockdowns. This approach is hindered somewhat by the large proportion of cases in which individuals are asymptomatic or have mild symptoms, which may go undetected. Ascertaining the frequency at which symptomatic infections are successfully detected will help our understanding of the limitations of testing systems and could help to improve the management of the COVID-19 pandemic.

Vittoria Colizza and colleagues estimated the number of symptomatic SARS-CoV-2 infections that occurred in France during a seven-week period (from mid-May until the end of June 2020) using mathematical models of transmission. These models were calibrated using regional data on hospital admissions and were validated in two ways: first, against serological studies; and second, by comparing the projected number of symptomatic cases in a single week with corresponding data on virologically confirmed cases and estimates from a self-reported symptom-tracker database for that week. Nearly 104,000 symptomatic infections were predicted to have occurred during the study period, compared with just over 14,000 confirmed cases. Detection rates increased over the study period, but only 5 of the 12 regions studied exceeded a 50% median detection rate by the end of June.

Data from the symptom-tracker database indicate that only 31% of individuals with COVID-19-like symptoms consulted a doctor in the study period. Together, these findings suggest that the majority of SARS-CoV-2 infections went undetected in the first few weeks after lockdown. The authors propose that test-and-trace strategies need to be improved substantially to control the spread of COVID-19. This could be achieved by increasing screening (for example, by removing barriers such as the need for a doctor to prescribe a test) and encouraging more people to seek medical advice if they present with COVID-19-like symptoms, even if these symptoms are mild. Overall improvements in the capacity and resolution of testing approaches are also required.