Abstract
SummaryBackgroundMaintaining effective contact tracing to control COVID-19 is challenging. Rapid growth in the number of infected cases can overload tracing and testing capacity, resulting in failure to trace contacts and delays in confirming an infection until after symptom onset (confirmation delay), hence increasing transmissibility. A substantial outbreak in Hong Kong, which was suppressed with non-pharmaceutical interventions (NPIs), provided an opportunity to assess the impact of overloading contact tracing and of efforts to improve its efficiency.MethodsUsing epidemiological-link (epi-link) data, we calculated the probability and duration of confirmation delay for cases with and without an epi-link, among all 3,148 confirmed cases between 5 July and 15 August 2020. Logistic regression was performed to determine the relationship between the number of recently confirmed infections and the probability of confirmation delay for epi-linked (contact-traced) cases. We estimated the impact on this relationship of targeted testing of at-risk groups.FindingsThe probability and duration of confirmation delay were associated with the rise in daily case number during growth of the outbreak. The proportion with confirmation delay among contact-traced cases increased from about to nearly as the number of cases grew from 1 to 50 per day (p-value = 0.003). The subsequent introduction of testing services for at-risk groups substantially reduced the proportion and it did not approach again until the daily number of cases exceeded 125. This 2.5-fold improvement in capacity contributed crucially to suppression of the outbreak.InterpretationThe number of recently confirmed infections is an indicator of the load on the contact-tracing system, the consequence of which can be assessed by the probability of confirmation delay. Measures to monitor and improve contact-tracing efficiency, alongside social distancing interventions, can enable outbreaks to be controlled without lockdown.FundingCity University of Hong Kong and Health and Medical Research Fund.
Highlights
There is real hope that vaccination will eventually terminate the COVID-19 pandemic, new outbreaks are likely to occur, perhaps for many years
We searched PubMed, bioRxiv, and medRxiv for articles published from 1 January, 2020 to 31 March, 2021, with the following keywords: (“2019-nCoV” OR “COVID-19” OR “SARS-CoV-2”) AND (“contact-tracing efficiency” or ”effectiveness of contact tracing” or ”confirmation delay” or “isolation delay”). 7 recent population-level modelling or simulation studies of COVID-19 have demonstrated theoretically the value of minimizing confirmation delay
We present the dynamics of the proportion of COVID-19 cases that were traced and of delays in confirmation during a substantial outbreak that was suppressed by non-pharmaceutical interventions (NPIs) without lockdown
Summary
There is real hope that vaccination will eventually terminate the COVID-19 pandemic, new outbreaks are likely to occur, perhaps for many years. Alongside social distancing measures, maintaining or increasing the efficiency of contact tracing and testing plays a crucial role in constraining the spread of infection.[1−3] rapid growth in the number of infected cases during expansion of an outbreak can overwhelm contact-tracing capacity,[4] which might reduce efficiency. Prolonging the time taken for tracing and testing will increase the probability of confirmation being delayed until after symptom onset (confirmation delay). 7 recent population-level modelling or simulation studies of COVID-19 have demonstrated theoretically the value of minimizing confirmation delay. No previous observational study has documented a change in the probability and duration of confirmation delay associated with an increase in number of cases, or determined to what extent different approaches of tracing and testing can mitigate such effects
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
