Bidirectional contact tracing could dramatically improve COVID-19 control

Contact tracing, alongside isolation and testing, is a powerful public health intervention. Current "forward-tracing" protocols focus on identifying and isolating individuals exposed to known cases. However, this approach misses asymptomatic or undiagnosed cases and those involved in superspreading events, leading to persistent transmission chains. The study hypothesizes that "bidirectional" contact tracing, which also identifies the infectors of known cases, can address these limitations. Furthermore, it explores the potential benefits of integrating bidirectional tracing with digital exposure notification systems using smartphones. The study aims to evaluate the efficacy of bidirectional tracing, both alone and in combination with digital exposure notification, using a stochastic branching-process model to simulate SARS-CoV-2 transmission under various epidemiological scenarios. This research is important because it directly addresses a critical gap in current COVID-19 control strategies, evaluating a method that could significantly reduce transmission and improve public health outcomes. The study seeks to quantify the potential benefits of bidirectional contact tracing and provide evidence-based recommendations for improving its implementation.

The paper references existing literature on the effectiveness of contact tracing, the prevalence of asymptomatic COVID-19 cases, and the role of superspreading events in transmission. It highlights the limitations of current forward-only tracing protocols and cites examples of bidirectional tracing implemented in countries like Japan and Singapore. The authors also review existing literature on digital exposure notification systems and their potential advantages in speed, scale, efficacy, and confidentiality. The literature review underscores the need for more effective contact tracing strategies and the potential of bidirectional tracing and digital tools to address the challenges posed by asymptomatic transmission and superspreading events.

The study utilizes a stochastic branching-process model of SARS-CoV-2 transmission, extending a previously published model of forward-tracing. The model incorporates parameters such as the proportion of asymptomatic carriers, the generation time distribution, the incubation period, the ascertainment rate for symptomatic cases, test sensitivity, and the compliance rate with isolation. Different tracing strategies are simulated, including forward-only tracing, bidirectional tracing with varying temporal windows (2 days and 6 days pre-symptom onset), digital exposure notification (with varying smartphone coverage and data-sharing rates), and hybrid systems combining manual and digital tracing. Each outbreak is initialized with 10 index cases, and the model assesses outbreak control based on extinction before reaching 10,000 cumulative cases. The effective reproduction number (R_eff) is computed as the mean number of child cases per case. The model is used to explore the efficacy of these strategies under various scenarios, including different levels of case ascertainment, test sensitivity, and basic reproduction number (R0). The parameters used in the model are based on published literature and reflect a range of plausible values (both optimistic and pessimistic scenarios). Sensitivity analyses are conducted to assess the robustness of the findings to changes in key parameters.

The key findings demonstrate the superior performance of bidirectional tracing compared to forward-only tracing across various scenarios. Bidirectional tracing significantly reduces R_eff, more than doubling the reduction achieved by forward-tracing alone. Extending the manual tracing window to 6 days pre-symptom onset dramatically improves the effectiveness of bidirectional tracing, leading to a much larger reduction in R_eff. The study also shows that while digital exposure notification alone can be highly effective with near-universal participation, even small decreases in smartphone uptake drastically reduce its effectiveness. A hybrid approach combining manual and digital tracing shows promising results, particularly when combined with a longer manual tracing window. Bidirectional tracing is shown to be more robust to low case ascertainment and test sensitivity compared to forward-only tracing. The benefits of bidirectional tracing are consistent across a range of R0 values. The probability of outbreak control is highly non-linear with varying R0, with even small reductions in R0 resulting in large increases in control probability for bidirectional tracing strategies. Comparing the performance of different tracing strategies relative to current practice, the study shows that bidirectional tracing with an extended window or high-uptake digital exposure notification can significantly improve COVID-19 control.

The findings strongly support the implementation of bidirectional contact tracing for improved COVID-19 control. The significant improvements in R_eff reduction and robustness to parameter variations highlight the potential of this approach. The study underscores the importance of addressing the challenges of asymptomatic transmission and superspreading events by actively identifying and tracing infectors. The results emphasize the need for extended temporal windows in manual tracing or near-universal participation in digital exposure notification systems. While the optimal approach might be a hybrid system combining both manual and digital tracing, the study's findings highlight the critical need for improved contact tracing methodologies to effectively control the spread of COVID-19 and other infectious diseases. The relative effectiveness of bidirectional tracing increases significantly in settings with low case ascertainment, emphasizing the importance of this strategy in areas with limited testing capacity. The findings provide valuable insights for policymakers and public health officials in optimizing contact tracing strategies.

This study demonstrates the significant potential of bidirectional contact tracing to enhance COVID-19 control. Expanding the manual tracing window or implementing high-uptake digital exposure notification can drastically improve the effectiveness of current protocols. The results highlight the importance of considering both infectors and infectees in contact tracing efforts and suggest that hybrid approaches combining manual and digital strategies could offer the most effective control. Further research should explore the practical challenges and cost-effectiveness of implementing bidirectional tracing strategies in various settings and investigate the optimization of resource allocation for manual tracing efforts.

The study relies on a stochastic branching-process model, which may not fully capture the complexities of real-world transmission dynamics. The model makes assumptions about parameters such as compliance rates, test sensitivity, and the proportion of asymptomatic cases. The accuracy of the model's predictions depends on the validity of these assumptions. Additionally, the study focuses on the theoretical benefits of bidirectional tracing and doesn't explicitly address the practical challenges and costs associated with implementing such strategies. Future work should involve field studies to validate the model's predictions and evaluate the cost-effectiveness of bidirectional tracing in real-world settings.