JMIR Publications

ABSTRACT

Background:

Contact tracing is a fundamental intervention in Public Health. When systematically applied, contact tracing enables the breaking of chains of transmission, an issue of special importance in the control of COVID-19 transmission. The capacity to perform contact tracing is influenced by availability of resources, prompting the need to estimate its effectiveness threshold relative to pandemic variations. This effectiveness threshold may be indirectly estimated by calculating the proportion of COVID-19 cases arising from high-risk contacts.

Objective:

To study the proportion of COVID-19 cases in high-risk contacts quarantined through contact tracing and its potential use as a pandemic control indicator.

Methods:

The research team used the data on COVID-19 collected by the Portuguese Directorate-General of Health and compiled by the Data Science for Social Good initiative. The team built an epidemiological compartmental model to simulate infection flow. We established parameters to assess infection dynamics, the influence of different variants, and vaccine efficacy. Two simulations were built: the first (A) adjusting for the presence and absence of variants or vaccination and a second (B) maximizing infection risk in individuals identified as high-risk contacts considering only one variant. The daily proportion of infected cases arising from high-risk contacts was calculated in both simulations, as was the effectiveness threshold of contact tracing.

Results:

An inverse relationship was found between the values of estimated proportion for high-risk contacts and the number of new cases in both simulations (correlations of -0.71 and -0.76, respectively). A parallel analysis of simulations’ results accounting for different variants and a potential protective effect from vaccination exhibits significant overlap. Simulation A had an effectiveness threshold for contact tracing of 1.93 (PPV = 71.7%) and simulation B had a value of 0.07 (PPV = 72.5%).

Conclusions:

Our results highlight that a diminishing proportion of cases in exposed individuals vis-Ă -vis new cases is an indirect indicator of diminishing efficacy of contact tracing. The models employed also allowed us to define a value for the effectiveness threshold of contact tracing considering its role as a key pandemic mitigation tool. The lessons derived from the application of our proposed methodology and the results we obtained are an important starting point to use data from identified high-risk contacts for COVID-19 to define the infection dynamics of the SARS-CoV-2 virus. We identified scenarios and limitations that this synthetic indicator might bring to support decision-making by health authorities and policymakers.