Up to 87% of COVID-19 cases in Wuhan, China, between January and March 2020 may have gone undetected, according to a modelling study published in Nature. These findings are consistent with recent serological studies in the United States and Europe. Undetected, or unascertained, infections — which may have included asymptomatic or presymptomatic individuals, or those with mild symptoms — probably had a major role in the rapid spread of the disease, and could lead to a resurgence of infections upon lifting of restrictions too early.
People with COVID-19 who are asymptomatic, presymptomatic or have mild symptoms are thought to have an important role in the rapid spread of infection because they are difficult to detect and isolate. Reconstructing the full dynamics of an outbreak allows a better understanding of the proportion and the effects of unascertained COVID-19 infections, which could inform efforts to monitor and control the resurgence of the disease.
Chaolong Wang and colleagues studied the transmission dynamics of the COVID-19 outbreak in Wuhan and evaluated the impact of interventions using data from 32,583 laboratory-confirmed cases from 8 December 2019 until 8 March 2020. They used these data to model the outbreak from 1 January 2020 and divided it into five time periods based on key events and interventions, such as Chinese New Year and the imposition of centralized isolation and quarantine.
Their analysis reveals that the initial rate of transmission was very high, with an estimated reproduction number (R0) of 3.54 in the first period, falling to around 0.28 by the end of the study period. This finding suggests that progressive and multi-faceted public health interventions that were put in place between late January and March 2020 reduced the number of total infections in Wuhan by 96.0% by 8 March.
By fitting their models to epidemiological data, the authors demonstrate that extensive undetected infections are likely to have been present in Wuhan. They estimate that 87% of infections were undetected during the study period, with a lower bound of the estimate being 53% under an extreme assumption that all cases were detected at the initial phase. Public health interventions, such as quarantine and social distancing, seem to be an effective way to block the transmission from unascertained cases and control the outbreak, the authors suggest. They emphasize that further investigations, such as serological studies, are needed to confirm these estimates.
Using the fitted model of these data, the authors go on to predict the chance of a second wave of infections. If all restrictions are lifted after 14 days from the first day on which no cases are reported, the chances of disease resurgence are expected to be very high (up to 97%), owing to the role of undetected cases with mild or no symptoms. They predict that the surge in cases would occur 34 days after restrictions were lifted. Under a more stringent scenario in which all restrictions are lifted only after 14 consecutive days without cases, the probability of resurgence drops to 32%, and the surge could be delayed to 42 days after the lifting of restrictions.
Epidemiology: A website to assess COVID-19 event risk in the US in real timeNature Human Behaviour
Materials: Making strong bio-based replacements for plasticsNature Communications
Pterosaur teeth reveal dietary preferencesNature Communications