Individual quarantine versus active monitoring of contacts for the mitigation of COVID-19: a modelling study

Abstract

Background: Voluntary individual quarantine and voluntary active monitoring of contacts are core disease control strategies for emerging infectious diseases such as COVID-19. Given the impact of quarantine on resources and individual liberty, it is vital to assess under what conditions individual quarantine can more effectively control COVID-19 than active monitoring. As an epidemic grows, it is also important to consider when these interventions are no longer feasible and broader mitigation measures must be implemented.

Methods: To estimate the comparative efficacy of individual quarantine and active monitoring of contacts to control severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2), we fit a stochastic branching model to reported parameters for the dynamics of the disease. Specifically, we fit a model to the incubation period distribution (mean 5·2 days) and to two estimates of the serial interval distribution: a shorter one with a mean serial interval of 4·8 days and a longer one with a mean of 7·5 days. To assess variable resource settings, we considered two feasibility settings: a high-feasibility setting with 90% of contacts traced, a half-day average delay in tracing and symptom recognition, and 90% effective isolation; and a low-feasibility setting with 50% of contacts traced, a 2-day average delay, and 50% effective isolation.

Findings: Model fitting by sequential Monte Carlo resulted in a mean time of infectiousness onset before symptom onset of 0·77 days (95% CI -1·98 to 0·29) for the shorter serial interval, and for the longer serial interval it resulted in a mean time of infectiousness onset after symptom onset of 0·51 days (95% CI -0·77 to 1·50). Individual quarantine in high-feasibility settings, where at least 75% of infected contacts are individually quarantined, contains an outbreak of SARS-CoV-2 with a short serial interval (4·8 days) 84% of the time. However, in settings where the outbreak continues to grow (eg, low-feasibility settings), so too will the burden of the number of contacts traced for active monitoring or quarantine, particularly uninfected contacts (who never develop symptoms). When resources are prioritised for scalable interventions such as physical distancing, we show active monitoring or individual quarantine of high-risk contacts can contribute synergistically to mitigation efforts. Even under the shorter serial interval, if physical distancing reduces the reproductive number to 1·25, active monitoring of 50% of contacts can result in overall outbreak control (ie, effective reproductive number <1).

Interpretation: Our model highlights the urgent need for more data on the serial interval and the extent of presymptomatic transmission to make data-driven policy decisions regarding the cost-benefit comparisons of individual quarantine versus active monitoring of contacts. To the extent that these interventions can be implemented, they can help mitigate the spread of SARS-CoV-2.

Funding: National Institute of General Medical Sciences, National Institutes of Health.

Figure 1

Simulated daily growth of infections and individuals under quarantine When the ratio of uninfected to infected contacts under quarantine is 1:1, the prevalence of infection among traced contacts is 0·5, and when it is 9:1, the prevalence is 0·1. The model assumes individual quarantine of contacts begins at a cumulative case count of 1000, in a low-feasibility setting with a basic reproductive number of 2·2, and a mean serial interval of 4·8 days (table). As can be seen in figure 2, exponential growth occurs in low-feasibility settings regardless of the longer or shorter serial interval scenario. The shorter serial interval and low-feasibility setting is a combination that has the clearest and fastest exponential growth, and has been used as an example to illustrate the differences in growth rates for cases and uninfected contacts.

Figure 2

Effective reproductive number under active monitoring and individual quarantine The effective reproductive number under active monitoring and individual quarantine increases with the basic reproductive number and in low-feasibility settings compared with high-feasibility settings in serial interval scenario 1 (A) with mean 4·8 days and scenario 2 (B) with mean 7·5 days. Equivalent control under individual quarantine and active monitoring would follow the y=x line.

Figure 3

Effect of presymptomatic infectiousness on effective reproductive number The grey borders around the Loess curves indicate 95% CIs. The dark grey line indicates the basic reproductive number. The effective reproductive number under active monitoring and individual quarantine decreases as the onset of infectiousness gets later with respect to the onset of symptoms in a high-feasibility setting, holding the basic reproductive number constant at 2·2. An offset of −2 days indicates infectiousness precedes symptoms by 2 days, an offset of 0 days indicates onset of both simultaneously, and an offset of 1 day indicates infectiousness onset occurs 1 day after symptom onset. The model assumes a mean serial interval of 4·8 days.

Figure 4

Effect of proportion of contacts traced on effective reproductive number The grey borders around the linear regression indicate 95% CIs. The dark grey line indicates the basic reproductive number. The effective reproductive number under active monitoring and individual quarantine increases as the proportion of contacts traced decreases, assuming a mean serial interval of 4·8 days and a basic reproductive number of 2·2. Intervention parameters other than fraction of contacts traced are set to the high-feasibility setting.

Figure 5

Synergistic effect of physical distancing and interventions based on contact tracing The grey borders around the linear regression indicate 95% CIs. Active monitoring and individual quarantine of 10%, 50%, and 90% of contacts provide incremental benefit over physical distancing when the mean serial interval is 4·8 days and the basic reproductive number is 2·2. Intervention parameters other than the fraction of contacts traced are set to the high-feasibility setting.