Estimation of daily detection rate
To assess the completeness of the diagnosed new cases on a daily basis, we used Eq (4) first to obtain a time series of \documentclass[12pt]{minimal} 				\usepackage{amsmath} 				\usepackage{wasysym} 				\usepackage{amsfonts} 				\usepackage{amssymb} 				\usepackage{amsbsy} 				\usepackage{mathrsfs} 				\usepackage{upgreek} 				\setlength{\oddsidemargin}{-69pt} 				\begin{document}$$ F\left(\overline{x}\right) $$\end{document}Fx¯ to represent the estimates of cumulative number of potentially detectable cases; we then used the first derivative \documentclass[12pt]{minimal} 				\usepackage{amsmath} 				\usepackage{wasysym} 				\usepackage{amsfonts} 				\usepackage{amssymb} 				\usepackage{amsbsy} 				\usepackage{mathrsfs} 				\usepackage{upgreek} 				\setlength{\oddsidemargin}{-69pt} 				\begin{document}$$ F^{\prime}\left(\overline{x}\right) $$\end{document}F′x¯ to obtain another time series of observed new cases each day; finally, with the observed F ′ (xi) and model predicted \documentclass[12pt]{minimal} 				\usepackage{amsmath} 				\usepackage{wasysym} 				\usepackage{amsfonts} 				\usepackage{amssymb} 				\usepackage{amsbsy} 				\usepackage{mathrsfs} 				\usepackage{upgreek} 				\setlength{\oddsidemargin}{-69pt} 				\begin{document}$$ F^{\prime}\left(\overline{x}\right) $$\end{document}F′x¯, we obtained the detection rate Pi for day i as: 5 \documentclass[12pt]{minimal} 				\usepackage{amsmath} 				\usepackage{wasysym} 				\usepackage{amsfonts} 				\usepackage{amssymb} 				\usepackage{amsbsy} 				\usepackage{mathrsfs} 				\usepackage{upgreek} 				\setlength{\oddsidemargin}{-69pt} 				\begin{document}$$ {P}_i=F^{\prime}\left({x}_i\right)/{F}^{\prime}\left({\overline{x}}_i\right),\mathrm{i}=\left(12/8/2019,12/9,2019\dots, 2/8/2020\right) $$\end{document}Pi=F′xi/F′x¯i,i=12/8/201912/92019…2/8/2020
We used these estimated Pi in this study in several ways. Before January 20, 2020 when the massive intervention was not in position, an estimated Pi > 1 was used as an indication of detecting more than expected cases, while an estimated Pi < 1 as an indication of detecting less than expected cases.
During the early period of massive intervention, an increase trend in Pi over time was used as evidence supporting the effectiveness of the massive intervention in detecting and quarantining more infected cases.
During the period 14 days (latent period) after the massive intervention, Pi < 1 was used as evidence indicating declines in new cases rather than under-detection; thus, it was used as a sign of early declines in the epidemic.
The modeling analysis was completed using spreadsheet. As a reference to assess the level of severity of the COVID-19 epidemic, the natural mortality rate of Wuhan population was obtained from the 2018 Statistical Report of Wuhan National Economy and Social Development.