TestEd Research Preliminary Results
Initiated in January 2021, the TestEd study sought to provide cheap, high sensitivity testing, to the student and staff population at the University of Edinburgh. In here we describe the preliminary findings revealing that different sensitivity is found in different levels of pooling: November 2022
Initiated in January 2021, the TestEd study sought to provide cheap, high sensitivity testing, to the student and staff population at the University of Edinburgh. During the COVID-19 pandemic many countries implemented testing programmes to detect SARS-CoV-2 in symptomatic individuals (such as the testing provided by NHS lighthouse labs in the UK). However, a disadvantage of such an approach is that it does not identify pre-symptomatic or asymptomatic individuals who are at risk of spreading the virus within the community, education, or work environments.1,2 Underscoring this, studies have estimated that anywhere from 20-70% of individuals infected with SARS-CoV-2 appear to remain asymptomatic, but capable of spreading disease, during infection.1,3–6
Arguments against broader testing programmes include the inherent cost and logistical challenges, coupled with low uptake for intrusive testing. In order to overcome such issues, TestEd setup at scale an asymptomatic saliva-based testing programme to provide testing at The University of Edinburgh. To maximise uptake, we developed an easy to use, non-invasive, saliva-based PCR test for SARS-CoV-2 infection, analysing 164,603 samples over a 23-month period to identify 868 SARS-CoV-2 positive cases (Figure 1).
Figure 1. Total number of tests carried out by the TestEd study, number of positive cases identified, and test positivity rate (incidence %) from January 2021 – November 2022.
(A) The number of samples tested and positive samples identified each week by the TestEd project. (B) The number of samples tested and test positivity rate (incidence %) each week during the course of the TestEd study. Number of tests = 164,603. Number of positive tests = 868.
One possibility to reduce the cost of testing, which is essential if desiring to screen a large pre-symptomatic/asymptomatic population, is to employ sample pooling. By pooling multiple samples together and running a single test for that pool of samples, the overall number of tests required can be reduced. For example, if 8 samples are combined to form a pool and then the test that is carried out on the pool yields a negative result, all the samples in the pool can be assumed to be negative with 7 less tests carried out compared to if the samples had been tested individually. In the case that such a pooled test, tests positive then the samples that make up the pool can be analysed individually to identify the positive sample(s). However, for such a pooling strategy to be effective the proportion of samples that test positive when analysed individually, and also test positive when diluted in a pool of negative samples needs to be known. This is the test sensitivity for that pooling method. To determine the cost-benefit of sample pooling, the TestEd study analysed samples individually, in pools of 8, 24, and 72, finding that pools of 8 reduced sensitivity to 88.4%, but concomitantly reduced costs 5.8 times (Figure 2).
Figure 2. SARS-CoV-2 positive samples identified by TestEd study from saliva of asymptomatic participants demonstrate reduction in sensitivity when pooling samples 1:8, 1:24, and 1:72.
(A) Schematic showing protocol if samples in a pool test negative (blue) or positive (red). (B) Percentage of TestEd positive samples detected when run in pools of 1:8, 1:24, and 1:72 with other samples. The percentage of samples in which virus genes were detected by RT-qPCR (sensitivity) is given at the bottom of each bar and the percentage of samples in which no virus genes were detected at the top of each bar. Each sample was run in technical quadruplicate using a multiplex assay to detect SARS-CoV-2 N gene (N1, N2, or N3 assay), E gene, S gene and human control gene, ACTB. Samples are binned according to the number of replicates of N and E genes combined detected for each sample as shown by the graph legend, for a given sample a maximum of 4 N gene and 4 E gene amplifications can be detected. When more than 1 positive sample was present in a pool it was excluded from the analysis. 1:8 pool n = 268, 1:24 pool n = 229, 1:72 pool n = 144.
Plotting the weekly number of tests carried out and number of positive samples identified over the course of the TestEd study shows that the number of positive cases fluctuated in waves that matched those seen in the whole of Scotland over this period (Figure 3A-C) [coronavirus.data.gov.uk/details/cases]. To identify SARS-CoV-2 variants causing infection in the TestEd study, samples were analysed using an S gene target failure assay or by whole-genome sequencing. As expected, the variants causing infection in TestEd participants changed across the study duration and closely match UK-wide data [covariants.org]: with Alpha (B.1.1.7) responsible for a wave of infections in January – April 2021; Delta (B.1.617.2) causing infections in June 2021 – January 2022; the emergence of the Omicron (BA.1) variant in December 2021; and the appearance of Omicron (BA.2) in late January 2022 (Figure 3D and E).
Figure 3. The weekly number of SARS-CoV-2 positive samples detected by the TestEd study correlate with UK wide increases in incidence and waves of variants of concern.
(A) Weekly number of tests for SARS-CoV-2 carried out across the study with an overlay of the number of positive samples identified in that week. Total number of tests = 127,265, total number of positive samples = 447 (11th January 2021 – 3rd April 2022). (B) Weekly test positivity rate (% cases positive) for SARS-CoV-2 tests carried out across the TestEd study compared to test positivity rate for the whole of Scotland. (C) Weekly number of TestEd positive samples compared to weekly number of SARS-CoV-2 positive cases reported in Scotland. (B-C) Scotland data from coronavirus.data.gov.uk. Week ending date shown on x axis. (D) and (E) Whole genome sequencing of TestEd SARS-CoV-2 positive samples identifies the variant causing infection. In (D) the number of SARS-CoV-2 positive samples identified as a given variant in a 2-week period are shown. In (E) the percentage of positive samples identified as a given variant in a 2-week period are shown. Before 02 Sep 2021 variants are assigned based on variant prevalent in UK at the time. After this date, variants are assigned based on whole genome sequencing or S gene target failure. n = 404, duplicate samples (given by the same participant within 60 days) were removed for this analysis.
While not a primary aim of the TestEd study, participants who tested positive with TestEd were asked to confirm their test result by either an NHS PCR test or lateral flow test when possible. The majority of participants (93%) who were identified as positive for SARS-CoV-2 by TestEd and that contacted the study after testing positive, were able to confirm their TestEd result (Figure 4). A minority of participants that tested positive were not able to confirm their positive result, however, of these, half reported experiencing symptoms of COVID-19. Those participants who tested positive by TestEd but subsequently tested negative by either NHS PCR test or lateral flow test and did not report experiencing symptoms of COVID-19 may represent false positives. However, it is likely that many of these cases are true positives but differences in when the tests were carried out or the type of test used account for discrepancies in results.
Figure 4. The majority of SARS-CoV-2 positive samples identified by the TestEd study were confirmed by a subsequent NHS PCR or lateral flow test and went onto develop symptoms.
(A) Percentage of samples which were identified as SARS-CoV-2 positive by TestEd and for which the participant contacted the study to confirm their TestEd result with an independent NHS PCR test or lateral flow test (LFT). The table shows the breakdown of samples identified as being positive for SARS-CoV-2 by the TestEd study and whether participants contacted the project manager. Participants who did not contact the study after being given a positive result were excluded when calculating the percentage of positive samples that were confirmed by an independent test for SARS-CoV-2. Positive samples that were classed as duplicate (submitted within 60 days of a positive TestEd result) or unassigned samples (no associated participant barcode) were also excluded.
The findings of the TestEd study are currently being prepared for publication and will include further data about test development, the effects of SARS-CoV-2 variants on test sensitivity, and cost analysis of the different pooling strategies.
- Johansson, M. A. et al. SARS-CoV-2 Transmission from People without COVID-19 Symptoms. JAMA Netw. Open 4, 1–8 (2021).
- Furukawa, N. W., Brooks, J. T. & Sobel, J. Evidence Supporting Transmission of Severe Acute Respiratory Syndrome Coronavirus 2 while Presymptomatic or Asymptomatic. Emerg. Infect. Dis. 26, E1–E6 (2020).
- Long, Q. X. et al. Clinical and immunological assessment of asymptomatic SARS-CoV-2 infections. Nat. Med. 26, 1200–1204 (2020).
- Oran, D. P. & Topol, E. J. Prevalence of asymptomatic SARS-CoV-2 infection. A narrative review. Ann. Intern. Med. 173, 362–368 (2020).
- Buitrago-Garcia, D. et al. Occurrence and transmission potential of asymptomatic and presymptomatic SARSCoV-2 infections: A living systematic review and meta-analysis. PLoS Med. 17, 1–25 (2020).
- Davies, N. G. et al. Age-dependent effects in the transmission and control of COVID-19 epidemics. Nat. Med. 26, 1205–1211 (2020).
- Hu, J. et al. Increased immune escape of the new SARS-CoV-2 variant of concern Omicron. Cell. Mol. Immunol. 19, 293–295 (2022).