Stringent containment measures without complete city lockdown to achieve low incidence and mortality across two waves of COVID-19 in Hong Kong

Abstract

Introduction: An international city, Hong Kong, in proximity to the first epicentre of COVID- 19, experienced two epidemic waves with different importation pressure. We compared the epidemiological features of patients with COVID-19 in the context of containment policies between the first and second waves.

Methods: We retrieved information on the first 1038 cases detected in Hong Kong (23 January to 25 April 2020) to analyse the epidemiological characteristics including age/gender-specific incidence, clustering, reproduction number (R_t ) and containment delay; in relation to the containment measures implemented. Factors associated with containment delay were evaluated by multiple linear regression analysis with age, gender, epidemic wave and infection source as covariates. A time series of 5-day moving average was plotted to examine the changes across the two epidemic waves.

Results: The incidence and mortality (135.5 and 0.5 per 1 000 000 population) was among the lowest in the world. Aggressive escalation of border control correlated with reductions in R_t from 1.35 to 0.57 and 0.92 to 0.18, and aversions of 450 and 1650 local infections during the first and second waves, respectively. Implementing COVID-19 tests for overseas returners correlated with an upsurge of asymptomatic case detection, and shortened containment delay in the second wave. Medium-sized cluster events in the first wave were family gatherings, whereas those in the second wave were leisure activities among youngsters. Containment delay was associated with older age (adjusted OR (AOR)=1.01, 95% CI 1.00 to 1.02, p=0.040), male gender (AOR=1.41, 95% CI 1.02 to 1.96, p=0.039) and local cases (AOR=11.18, 95% CI 7.43 to 16.83, p<0.001), and with significant improvement in the second wave compared with the first wave (average: 6.8 vs 3.7 days). A higher incidence rate was observed for males, raising possibility of gender predilection in susceptibility of developing symptoms.

Conclusion: Prompt and stringent all-round containment strategies represent successful measures in pandemic control. These findings could inform formulation and implementation of pandemic mitigation strategies.

Keywords: SARS; epidemiology; public health; respiratory infections.

Figure 1

Number of confirmed cases by illness-onset date, time-varying reproduction number and arrival statistics at border points of entry. The epidemic curve in the upper panel includes 844 symptomatic cases. One symptomatic case with uncertain illness-onset date and 193 asymptomatic cases were excluded. The middle panel shows the estimated R_t of each day with a black line, and their 95% CIs with shaded area. The lower panel shows the number of persons arriving in Hong Kong via land and sea borders, and airport.

Figure 2

Estimated number of local cases averted by control measures implemented during the two waves of COVID-19 outbreak. Simulated (dash lines) versus observed (solid line) cumulative number of local cases before and after border controls and quarantine policies for entrants in the (A) first wave and (B) second wave of epidemic using the infection generation equation from Thompson and his colleagues. In the first wave (A), the initial median reproduction number of the simulated curve was 1.35, whereas in the second wave (B), the initial median reproduction number of the simulated curve was 0.92. A 30-day projection of cumulative local infections was simulated in the periods before and after the border controls and quarantine policies in the first (before and after 8 February) and second (before and after 20 March) waves. Shaded regions indicate the 95% confidence bands of the simulated curves.

Figure 3

The number and incidence of patients with COVID-19 by epidemic waves. Panel (A) shows the number of cases with respect to the first wave (left) and second wave (right) for each age and sex group, and overall (in thick borders). Panel (B) shows the incidence rates. The number of locally acquired infections was 70 and 397 in the first and second waves, respectively; while there were 82 and 935 local residents in the two waves, respectively. Local residents refer to individuals with Hong Kong identity cards.

Figure 4

Cumulative age/sex-specific incidence rates. All COVID-19 cases confirmed on or before 25 April 2020 who are Hong Kong residents were included. The mid-2019 population census data were used as denominators.

Figure 5

Number of asymptomatic SARS-CoV-2 infections detected and implementation of testing policies. All COVID-19 testing services provided by the Department of Health as part of the outbreak response measures were free of charge. Symptomatic persons were admitted to hospital where nasopharyngeal swab samples were taken. Self-collected deep throat saliva samples were used for asymptomatic adults. Self-collected stool samples were used for asymptomatic children arriving in Hong Kong.