Estimation of unreported SARS-CoV-2 cases in Italy using a Susceptible-Exposed-Infectious-Recovered-Dead model

Background: An important epidemiological characteristic that might modulate the pandemic potential of severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) is the proportion of undocumented cases. Methods: Here, we employed a Susceptible-Exposed-Infectious-Recovered-Dead (SEIRD) model to estimate the proportion of unreported SARS-CoV-2 cases in Italy from the reported number of deaths prior to the adoption of national control measures. Results: We estimated 115 894 infectious individuals (95% confidence interval (CI) = 95 318-140 455) and a total of 144 116 cases (95% CI = 119 030-173 959) on 20 March, 2020. These estimates resulted in 67.3% (95% CI = 60.3%-73.0%) unreported infectious individuals and in 67.4% (95% CI = 60.5%-73.0%) total cases. As such, given the substantial volume of undocumented cases, the case fatality risk would drop from an apparent 8.6% to an estimated 2.6% (95% CI = 2.2%-2.9%). Conclusions: Our findings partially explain the case fatality risk observed in Italy with a high proportion of unreported SARS-CoV-2 cases. Moreover, we underline that the fraction of undocumented infectious individuals is a critical epidemiological characteristic that needs to be taken into for a better understanding of the SARS-CoV-2 epidemic.