Modeling the effectiveness of olfactory testing to limit SARS-CoV-2 transmission

A central problem in the COVID-19 pandemic is that there is not enough testing to prevent infectious spread of SARS-CoV-2, causing surges and lockdowns with human and economic toll. Molecular tests that detect viral RNAs or antigens will be unable to rise to this challenge unless testing capacity increases by at least an order of magnitude while decreasing turnaround times. Here, we evaluate an alternative strategy based on the monitoring of olfactory dysfunction, a symptom identified in 76–83% of SARS-CoV-2 infections—including those with no other symptoms—when a standardized olfaction test is used. We model how screening for olfactory dysfunction, with reflexive molecular tests, could be beneficial in reducing community spread of SARS-CoV-2 by varying testing frequency and the prevalence, duration, and onset time of olfactory dysfunction. We find that monitoring olfactory dysfunction could reduce spread via regular screening, and could reduce risk when used at point-of-entry for single-day events. In light of these estimated impacts, and because olfactory tests can be mass produced at low cost and self-administered, we suggest that screening for olfactory dysfunction could be a high impact and cost-effective method for broad COVID-19 screening and surveillance.

: Impact of the timing of olfactory dysfunction onset on its effectiveness to limit viral spread. Example of viral spread in fully-mixed community of 20,000 individuals performing olfactory dysfunction (OD) screening daily (A) or weekly (B). olfactory dysfunction is modeled to be present in 75% of infected individuals, and to last 7 days. We consider 80% participation in testing and that 20% of individuals would suffer COVID-19independent olfactory dysfunction which would exclude them from effective testing. Timing of olfactory dysfunction is varied from one to four days after virions levels reaching 1000 virions/ml (purple shaded lines as indicated). No mitigation is shown as black line. For comparison, weekly RT-PCR testing with a one day turnaround is shown.  Figure 3: Impact of duration of olfactory dysfunction on its effectiveness to limit viral spread.
Examples of viral spread in fully-mixed community of 20,000 individuals performing olfactory dysfunction (OD) screening daily (A) or weekly (B). Olfactory dysfunction is modeled to be present in 75% of infected individuals, and to begin two days after virion levels reach 1000 virions/ml. We consider 80% participation in testing and that 20% of individuals would suffer COVID-19-independent olfactory dysfunction which would exclude them from effective testing. Duration of olfactory dysfunction is varied from 7 days (lightest green), 5 days (light green), 3 days (green), to 1 day (dark green  olfactory dysfunction is modeled to be present in 75% of infected individuals, and to begin after 1, 2, 3, or 4 days from when virion levels reach 1000 virions/ml, indicated by varying shades of purple (see legend). We consider 80% participation in testing and that 20% of individuals would suffer COVID-19-independent olfactory dysfunction which would exclude them from effective testing. No mitigation is shown as black line. For comparison, RT-PCR testing with a one day turnaround and antigen testing are shown with testing every three days (left column) or weekly (middle and right columns).  The impact of repeated population screening on the reproductive number can be estimated by considering the ratio of population infectiousness with a screening regimen to population infectiousness with no screening. However, note that the impact of a population screening policy may depend on two additional factors.
First, not all individuals may wish to participate in a testing program. Let the fraction of individuals who participate be given by .
Second, a virological test (e.g. RT-PCR) may produce a false negative result unrelated to its limit of detection-for instance due to an improperly collected sample. Let se be the test sensitivity, in the particular sense of the probability of correctly diagnosing an individual as positive when that person's viral load should, in principle, have provided a sufficiently high RNA concentration to be detectable.
Let f 0 be the total infectiousness removed with no testing policy, i.e. due to symptom-driven self isolation. Let f test (se) be the fraction of total infectiousness removed with a chosen testing policy, inclusive of symptom-driven self isolation, as well as the test sensitivity se introduced above. The quantity f test may be computed for any screening program, including virological testing or symptom screening.
Both f 0 and f test (se) can be estimated rapidly via Monte Carlo by drawing trajectories and applying a population screening regimen to them in which a fraction 1 se positive tests are discarded uniformly at random. In the main text, we found that estimating these values using 10, 000 randomly drawn trajectories was sufficient to produce stable estimates.
Under the assumption of statistical independence between an individual's participation or refusal, viral load or olfactory dysfunction status, and se, we can approximate the reproductive number as which simply expresses a weighted combination of removed infectiousness via screening regimen participation and no test. Intuitively, note that if there is complete refusal to participate ( = 0) or an entirely ineffective test (f test (se) = f 0 ), then R ⇡ R 0 , as expected.