News
Clarification on the interpretation of daily point prevalence of the HKU real-time community surveillance initiative

03 January 2023

During the fifth wave of the rapidly evolving COVID-19 epidemic in Hong Kong, LKS Faculty of Medicine, The University of Hong Kong (HKUMed) has launched a real-time community surveillance initiative using rapid antigen tests (RAT) since 1 March 2022.

A total of 10,000 participants have been recruited from all 18 districts around Hong Kong, divided into 7 subcohorts and carried out RATs regularly, irrespective of clinical symptoms, to assess and monitor the underlying epidemic dynamics of COVID-19.

Basing on the RAT results submitted by the participants, this surveillance programme has estimated the real-time point prevalence on a daily basis. The daily point prevalence is estimated by dividing the number of individuals reported positive results by the number of valid test results submitted on the same day. It reflects the estimated proportion of the population that has COVID-19 infection on a particular day, including newly occurring cases and existing cases that have not yet recovered. This estimate has been age-standardised by relative age distribution of the population structure in Hong Kong according to the 2021 Population Census, with its confidence interval calculated to show the uncertainty around the estimate.

With the persisting concern on the local epidemic situation, we hope the surveillance data, irrespective of going up or down, would help to inform a proper understanding of the epidemic level for the general public, and to provide real-time evidence for the governmental public health and epidemic control policy. In the last week, the daily point prevalence was ranging from 2% to 6%. Given that RAT is having an average sensitivity of 75% for detecting an infection, and that patients infected by the Omicron variant generally remained RAT positive for 7 days, the 2% to 6% daily point prevalence indicated that the number of daily new infections in the community shall be around 26,700-80,000 (7,000,000 x 2% ÷7 ÷0.75 = 26,667 and 7,000,000 x 6% ÷7 ÷0.75 = 80,000). As considerable uncertainty in the estimate is suggested by the wide confidence intervals, continuous monitoring and a comprehensive consideration of all relevant surveillance data would be necessary to determine whether the lower figures over the recent two days did suggest a trend of subsiding epidemic activity after reaching the peak.

The point prevalence estimated from COVID-19-Real-time Community Surveillance Initiative is released daily on the real-time online dashboard maintained by the School of Public Health, HKUMed (https://covid19.sph.hku.hk). We are most grateful to generous support of Henry Fok Foundations, Hong Kong Health Bureau and all the participants in this surveillance initiative.