Explaining the Bomb-Like Dynamics of COVID-19 with Modeling and the Implications for Policy

Using a Bayesian approach to epidemiological compartmental modeling, we demonstrate the "bomb-like" behavior of exponential growth in COVID-19 cases can be explained by transmission of asymptomatic and mild cases that are typically unreported at the beginning of pandemic events due to lower prevalence of testing. We studied the exponential phase of the pandemic in Italy, Spain, and South Korea, and found the R0 to be 2.56 (95% CrI, 2.41-2.71), 3.23 (95% CrI, 3.06-3.4), and 2.36 (95% CrI, 2.22-2.5) if we use Bayesian priors that assume a large portion of cases are not detected. Weaker priors regarding the detection rate resulted in R0 values of 9.22 (95% CrI, 9.01-9.43), 9.14 (95% CrI, 8.99-9.29), and 8.06 (95% CrI, 7.82-8.3) and assumes nearly 90% of infected patients are identified. Given the mounting evidence that potentially large fractions of the population are asymptomatic, the weaker priors that generate the high R0 values to fit the data required assumptions about the epidemiology of COVID-19 that do not fit with the biology, particularly regarding the timeframe that people remain infectious. Our results suggest that models of transmission assuming a relatively lower R0 value that do not consider a large number of asymptomatic cases can result in misunderstanding of the underlying dynamics, leading to poor policy decisions and outcomes.

Pooling RT-PCR or NGS samples has the potential to cost-effectively generate estimates of COVID-19 prevalence in resource limited environments

BackgroundCOVID-19 originated in China and has quickly spread worldwide causing a pandemic. Countries need rapid data on the prevalence of the virus in communities to enable rapid containment. However, the equipment, human and laboratory resources required for conducting individual RT-PCR is prohibitive. One technique to reduce the number of tests required is the pooling of samples for analysis by RT-PCR prior to testing. MethodsWe conducted a mathematical analysis of pooling strategies for infection rate classification using group testing and for the identification of individuals by testing pooled clusters of samples. FindingsOn the basis of the proposed pooled testing strategy we calculate the probability of false alarm, the probability of detection, and the average number of tests required as a function of the pool size. We find that when the sample size is 256, using a maximum pool size of 64, with only 7.3 tests on average, we can distinguish between prevalences of 1% and 5% with a probability of detection of 95% and probability of false alarm of 4%. InterpretationThe pooling of RT-PCR samples is a cost-effective technique for providing much-needed course-grained data on the prevalence of COVID-19. This is a powerful tool in providing countries with information that can facilitate a response to the pandemic that is evidence-based and saves the most lives possible with the resources available. FundingBill & Melinda Gates Foundation Authors contributionsRL and KRN conceived the study. IF, KT, KRN, SB and RL all contributed to the writing of the manuscript and AH and JJ provided comments. KRN and AH conducted the analysis and designed the figures. Research in contextO_ST_ABSEvidence before this studyC_ST_ABSThe pooling of RT-PCR samples has been shown to be effective in screening for HIV, Chlamydia, Malaria, and influenza, among other pathogens in human health. In agriculture, this method has been used to assess the prevalence of many pathogens, including Dichelobacter nodosus, which causes footrot in sheep, postweaning multisystemic wasting syndrome, and antibiotic resistance in swine feces, in addition to the identification of coronaviruses in multiple bat species. In relation to the current pandemic, researchers in multiple countries have begun to employ this technique to investigate samples for COVID-19. Added value of this studyGiven recent interest in this topic, this study provides a mathematical analysis of infection rate classification using group testing and calculates the probability of false alarm, the probability of detection, and the average number of tests required as a function of the pool size. In addition the identification of individuals by pooled cluster testing is evaluated. Implications of all the available evidenceThis research suggests the pooling of RT-PCR samples for testing can provide a cheap and effective way of gathering much needed data on the prevalence of COVID-19 and identifying infected individuals in the community, where it may be infeasible to carry out a high number of tests. This will enable countries to use stretched resources in the most appropriate way possible, providing valuable data that can inform an evidence-based response to the pandemic.