Abstract
Recent reports have documented substantial proportions of asymptomatic or mildly symptomatic Covid-19 infections that may also be infectious [1-3] as well as the potential role of respiratory droplets and contaminated surfaces in driving SARS-CoV-2 transmission [4]. While facemask use is promoted in healthcare settings as part of infection control protocols [5-8], its use during the COVID-19 pandemic varies across countries [9,10]. As the COVID-19 pandemic rages on, widespread facemask use has been recently recommended by Dr. George Gao, the director-general of the Chinese Center for Disease Control and Prevention (CDC) [11].
Any plan for stopping the ongoing 2019-nCov pandemic must be based on a quantitative understanding of the proportion of the population that needs to be protected by effective control measures such that each infected person infects no more than one other person on average (the effective reproductive number, R < 1), at which point transmission contracts and eventually burns out. Based on a modeling study, we show that the pandemic may be readily controllable through a combination of testing, treatment if necessary, and self-isolation after testing positive (TTI) of symptomatic individuals together with social protection (e.g., facemask use, handwashing).
We used an SEIR-type model incorporating asymptomatic but infectious individuals (40%) [1-3] to investigate how individual protective behaviors, different levels of testing, and isolation influence the transmission and control of the COVID-19 pandemic (Appendix). When the basic reproduction number, R0, is 2.4 [12], 65% effective social protection alone (35% of the unprotected transmission) brings the R below 1 (Appendix). Alternatively, 20% effective social protection brings the reproduction number below 1.0 so long as 75% of the symptomatic population is covered by TTI within 12 hours of symptom onset (Figure A and Appendix). Even with 20% effective social protection, TTI of 1 in 4 symptomatic individuals can substantially “flatten the curve”, cutting the peak daily incidence in half (Figure B, Appendix).
(A) Total number of infections in a population of 10 million individuals for varying levels of social protection and testing (TTI) at 12 hours testing delay and assuming 40% asymptomatic transmission. (B) Daily incidence in a population of 10 million individuals for 20% effectiveness of social protection and varying levels of testing (TTI) at 12 hours testing delay and assuming 40% asymptomatic transmission. Initial conditions were set as follows, susceptible individuals: 9,943,400; symptomatic individuals who underwent testing: 40,000; asymptomatic individuals: 16,000; and deceased individuals: 600.
Data Availability
All data is included in the appendix
Supplementary Appendix to
Basic reproduction number, R0
The basic reproduction number, R0, is defined as the average number of secondary cases generated by primary infectious individuals during the early transmission phase in a completely susceptible population and in the absence of control interventions. This is a key metric to gauge the intensity and type of interventions that need to be implemented in order to bring the epidemic under control. For the ongoing epidemic of 2019-nCov, R0 has been estimated at 2.4 (3.9, upper bound of 95% confidence interval).
Reproduction number with testing, isolation, and social protection
The reproductive number, R, quantifies the potential for infectious disease transmission in the context of a partially susceptible population. When R > 1, infection may spread in the population, and the rate of spread is higher with increasingly high values of R. If R < 1, infection cannot be sustained and is unable to generate an epidemic. As there are seven classes contributing to new infections, the reproduction number is the sum of the contributions of the infectious classes:
The contributions of the individual compartments are as follows
Modeling the effect of testing, isolation and social protection of infected individuals
Compartmental transmission models are commonly used in infectious disease epidemiology as a population-level modeling approach that subdivides the population into various classes based on epidemiological status. Compartmental models are specified by a set parameters and ordinary differential equations that track the progression of the number of individuals within each state. Here, we developed an extension of the commonly used SEIR transmission model for modeling the transmission of COVID-19.
Individuals within the model are classified as susceptible (S), latent (E1), partially infectious but not yet symptomatic (E2), asymptomatic and will not be tested (An), asymptomatic and will be tested and will be isolated (As), infectious and will not be tested (In), infectious and will be tested and isolated (Is), hospitalized/isolated infectious (J), isolated asymptomatic (Ja), recovered (R), and deceased (D). Constant population size is assumed, so N is equal to the sum of individuals in all of the compartments. Further, seven classes can contribute to new infections: E2, An, As, In, Is, J, Ja. Susceptible individuals move to E1 at rate β[qeE2(t) + qaqAn(t) + qaAs(t)+qIn(t) + qIs(t) + haJa(t) + hJ(t)]/N, where β denotes the overall transmission rate. The transmission rate, β, was calibrated based on the baseline value of the basic reproductive number, R0 = 2.4, which had been calculated.
Individuals in E1 progress to E2 at rate κ1. Individuals from E2 are partially infectious, with relative transmissibility qe, and progress at rate κ2, where a proportion ρa become asymptomatic and partially infectious (relative transmissibility qa), and 1 – ρabecome fully infectious. Among the proportion ρa who become asymptomatic, ns will be tested, while 1-ns will be undetected. Further, among the proportion 1 – ρa that become fully infectious, ρs will be tested, while 1 – ρs will be undetected. Asymptomatic individuals who are not tested and symptomatic individuals wear personal protective equipment (PPE) (such as wearing masks in public, handwashing, etc) and thus have relative transmissibility q, which is proportional to the level of effectives of PPE. Individuals within An and In classes (who are not tested) recover at rate γ1. Those who are tested (Is and As) will progress to the hospitalized and isolated class at diagnosis rate α, where relative transmission within hospitals and isolated places, h, can occur. However, we assume perfection isolation in our analyses. Individuals who are hospitalized and isolated progress to the recovered class at rate γ2 or to the deceased class at rate δ.
Therefore, the system is defined by the following system of non-linear differential equations:
Schematic of the model diagram
The population is classified into 10 epidemiological states: susceptible (S), latent (E1), partially infectious but not yet symptomatic (E2), asymptomatic and will not be tested (An), asymptomatic and will be tested and will be isolated (As), infectious and will not be tested (In), infectious and will be tested and isolated (Is), hospitalized/isolated infectious (J), isolated asymptomatic (Ja), recovered (R), and deceased (D). Model parameters are described in Table 1.
(**Indicates the value was estimated)
References
- 1.
- 2.
- 3.
- 4.
- 5.
- 6.
- 7.
- 8.
- 9.
- 10.
- 11.
