Estimating LFT and qPCR test-sensitivity over time since infection from a human challenge study

Abstract

Testing has been central to global policy throughout the SARS-CoV-2 pandemic. Understanding how test sensitivity changes after exposure is crucial for the interpretation of test outcomes and the design of testing-based interventions. Using data from a human challenge study, we derive temporal test sensitivity profiles for lateral flow tests (LFT), quantitative polymerase chain reaction (qPCR) tests and viable virus, measured by focus-forming assay (FFA). The median time to detectability was 2 days (throat swab) and 3 days (nasal) for qPCR and 4 days for LFT (both swabs), and there was strong positive correlation between first LFT positive and first FFA positive for both throat (p=0.00019) and nasal (p=0.00032) swabs, supporting the use of LFTs as a method of detecting infectiousness. Peak LFT sensitivity was 82.4% (67.0%-91.8%) for throat samples, occurring 6 days post-exposure and 93.3% (85.1%-98.0%) for nasal, 7 days post-exposure. These temporal profiles provide quantification of the mean behaviour of these tests and individual-based variability and could inform a framework for investigating future testing-based interventions.

Epidemiological Parameters of SARS-CoV-2 in the UK during the 2023/2024 Winter: A Cohort Study

“…It is possible that some individuals may never have high enough viral loads to test positive by LFD. However, previous estimates suggest that peak LFD sensitivity is comparable to peak PCR sensitivity [13]. LFD sensitivity at the peak was 95% (92%, 98%).…”

“…Test sensitivity for LFD tests has previously been shown to be lower than the gold-standard PCR tests [13,28,42]. However, despite the reduced sensitivity, due to the greatly improved time-to-result and reduced cost, LFD tests can be a powerful tool in the response to the COVID-19 pandemic [43].…”

“…Therefore, this tells us the probability that a test taken a certain number of days after the symptom onset date will return a true positive test result, using LFD test positivity as a proxy for infection. Previous studies [13,[29][30][31][32][33] have shown that peak LFD sensitivity, which occurs around symptom onset, is comparable to peak PCR sensitivity, which is considered the gold standard at detecting infection. Therefore, this will accurately describe the probability of detecting an infection using LFD tests.…”

“…In this paper, data from this survey are used to estimate the duration of LFD positivity, LFD test sensitivity, and the current incubation period (the time from becoming infected to developing symptoms) of SARS-CoV-2. The most recent estimates of these parameters are limited to the pre-Alpha [13], Omicron BA.1 and BA.2 [14], and Omicron BA.5 [8,15] periods, respectively, reducing their utility in current public health policy. These parameters are essential in modelling/designing different interventions or public health messaging, as well as important parts of infectious disease surveillance tools.…”