The health effects in the US of quarantine policies based on predicted individual risk of severe COVID-19 outcomes

BackgroundSocial distancing, testing and public health measures are the principal protections against COVID-19 in the US. Social distancing based on an accurate assessment of the individual risk of severe outcomes could reduce harm even as infection rates accelerate. MethodsAn SEIR dynamic transmission model of COVID-19 was created to simulate the disease in the US after October 2020. The model comprised 8 age groups with US-specific contact rates and low- and high-risk sub-groups defined in terms of the risk of a severe outcome determined by relevant comorbidities and a genetic test. Monte Carlo analysis was used to compare quarantine measures applied to at risk persons identified with and without the genetic test. ResultsUnder the piecemeal social distancing measures currently in place, absent a vaccine the US can expect 114 million symptomatic infections, 4.8 million hospitalisations and 262,000 COVID-19 related deaths. Social distancing based solely on comorbidities with 80% compliance reduces symptomatic infections by between 1.2 and 2.2 million, hospitalisations by between 1.2 and 1.3 million, and deaths by between 71,800 and 80,900. Refining the definition of at risk using a test of single-nucleotide polymorphisms further reduces symptomatic infections by 1.0 to 1.2 million, hospitalisations by 0.4 million and deaths by between 20,500 and 24,100. ConclusionsModels are now available that can accurately predict the likelihood of severe COVID-19 outcomes based on age, sex, comorbidities and polygenetic testing. Quarantine based on risk of severe outcomes could substantially reduce pandemic harm, even when infection rates outside of quarantine are high.