Contact tracing, isolation, and testing are some of the most powerful public health interventions available, but they haven't been optimized to deal with the low case identification rates and high asymptomatic transmission characteristic of the COVID-19 pandemic.
Current “forward-tracing” protocols seek to identify and isolate individuals who may have been infected by the known case, preventing continued transmission. However, chains of SARS-CoV-2 transmission commonly go undetected, with even the best-performing health systems struggling to detect even 50% of cases.
We reasoned that when asymptomatic carriers are common and detection rates are low, “bidirectional” contact tracing could identify and isolate undiscovered branches of the viral family tree, preventing many additional cases.
To test this hypothesis, we created a mathematical model informed by a range of epidemiological parameters from the literature, which predicts than bidirectional tracing can more than double the efficacy of current protocols. Importantly, this result is robust across epidemiological scenarios, and it's actually more effective in areas that are struggling to find most cases.
Next, we examined whether adding smartphone-based exposure notification could help, and found that it can triple the effectiveness... but only if nearly every smartphone participates by logging exposure events. Anything less, and the benefits are marginal, strongly suggesting that digital exposure notification should be integrated into the operating system of each smartphone with voluntary sharing of exposure events upon diagnosis.
Our work is available as a preprint on medRxiv and has been submitted for publication.