Contact tracing efficiency, transmission heterogeneity, and accelerating COVID-19 epidemics

Simultaneously controlling COVID-19 epidemics and limiting economic and societal impacts presents a difficult challenge, especially with limited public health budgets. Testing, contact tracing, and isolating/quarantining is a key strategy that has been used to reduce transmission of SARS-CoV-2, the virus that causes COVID-19 and other pathogens. However, manual contact tracing is a time-consuming process and as case numbers increase a smaller fraction of cases’ contacts can be traced, leading to additional virus spread. Delays between symptom onset and being tested (and receiving results), and a low fraction of symptomatic cases being tested and traced can also reduce the impact of contact tracing on transmission. We examined the relationship between increasing cases and delays and the pathogen reproductive number Rt, and the implications for infection dynamics using deterministic and stochastic compartmental models of SARS-CoV-2. We found that Rt increased sigmoidally with the number of cases due to decreasing contact tracing efficacy. This relationship results in accelerating epidemics because Rt initially increases, rather than declines, as infections increase. Shifting contact tracers from locations with high and low case burdens relative to capacity to locations with intermediate case burdens maximizes their impact in reducing Rt (but minimizing total infections may be more complicated). Contact tracing efficacy decreased sharply with increasing delays between symptom onset and tracing and with lower fraction of symptomatic infections being tested. Finally, testing and tracing reductions in Rt can sometimes greatly delay epidemics due to the highly heterogeneous transmission dynamics of SARS-CoV-2. These results demonstrate the importance of having an expandable or mobile team of contact tracers that can be used to control surges in cases. They also highlight the synergistic value of high capacity, easy access testing and rapid turn-around of testing results, and outreach efforts to encourage symptomatic cases to be tested immediately after symptom onset.
Author summary Contact tracing is a key tool in the control of infectious diseases. However, to successfully contact and quarantine individuals with traditional methods requires time and is limited by available capacity. As cases rise, limited capacity results in only a fraction of contacts being reached each day before the next set of cases is detected. Here we examine the relationships between increasing case numbers, contact tracing efficiency, and the pathogen reproductive number Rt (the number of cases infected by each case) and how these relationships vary with delays between symptom onset and tracing initiation, and incomplete participation in the testing and tracing process. We found that under conditions found throughout much of the world, where only 20–40% of symptomatic cases are detected, contact tracing can reduce Rt by a maximum of 20%, and this benefit quickly dissolves as contacts needing tracing exceed capacity. Increases in the fraction of symptomatic people being tested and reductions in the time between symptom onset and initiating tracing substantially increase the potential impact of tracing in reducing Rt to 60%, but this benefit is also lost as cases rise and contacts needing tracing exceed capacity to do so. Maintaining excess contact tracing capacity and the ability to shift it to areas most in need can substantially reduce pathogen transmission and limit epidemics.