Retrospective screening of routine respiratory samples revealed undetected community transmission and missed intervention opportunities for SARS-CoV-2 in the United Kingdom

Abstract

In the early phases of the SARS coronavirus type 2 (SARS-CoV-2) pandemic, testing focused on individuals fitting a strict case definition involving a limited set of symptoms together with an identified epidemiological risk, such as contact with an infected individual or travel to a high-risk area. To assess whether this impaired our ability to detect and control early introductions of the virus into the UK, we PCR-tested archival specimens collected on admission to a large UK teaching hospital who retrospectively were identified as having a clinical presentation compatible with COVID-19. In addition, we screened available archival specimens submitted for respiratory virus diagnosis, and dating back to early January 2020, for the presence of SARS-CoV-2 RNA. Our data provides evidence for widespread community circulation of SARS-CoV-2 in early February 2020 and into March that was undetected at the time due to restrictive case definitions informing testing policy. Genome sequence data showed that many of these early cases were infected with a distinct lineage of the virus. Sequences obtained from the first officially recorded case in Nottinghamshire - a traveller returning from Daegu, South Korea - also clustered with these early UK sequences suggesting acquisition of the virus occurred in the UK and not Daegu. Analysis of a larger sample of sequences obtained in the Nottinghamshire area revealed multiple viral introductions, mainly in late February and through March. These data highlight the importance of timely and extensive community testing to prevent future widespread transmission of the virus.

Keywords: COVID-19; Community Transmission; Molecular Epidemiology; Pooled Screening; SARS-CoV-2; Whole-Genome Sequencing.

Fig. 1.

Number of new SARS-CoV-2 cases detected in Nottingham University Hospitals patients per day, from the rollout of localised testing on 12 March until 2 June 2020.

Fig. 2.

Proportion of SARS-CoV-2 lineages detected between 21 February and 2 June 2020 in Nottingham, compared with the prevalence of the same lineages in the rest of the UK and elsewhere in the world during the same time period (a). The total count of these lineages is also shown as a percentage of the total count of the same lineages detected in the rest of the UK and world (b).

Fig. 3.

Frequency of SARS-CoV-2 lineages detected in Nottingham between 21 February and 2 June 2020, per week, as total count (a) and as a proportion of the total number (b).

Fig. 4.

Phylogenetic relationships of SARS-CoV-2 sequences from lineage B based on their entire genome (29412nt). This is an extraction of a sub-tree that contains most of the lineage B sequences detected in Nottingham, including those identified during retrospective PCR testing. The earliest known sequence is shown in green, other sequences identified in patients during the retrospective analysis are shown in purple, sequences obtained from the first confirmed local case (a traveller returning from South Korea) are shown in blue and sequences obtained from other local patients identified after introduction of wide-screen testing are shown in red. For clarity some branches have been collapsed. Where a branch has been collapsed the geographical locations of where the sequences were obtained are shown at the tip, and the number of individual sequences represented is shown in parentheses. The tree was rooted on a Wuhan sequence sampled on 2020-01-05. Reference sequences are indicated by their GISAID accession numbers. Branch lengths are drawn to a scale of nucleotide substitutions per site. Numbers above individual branches indicate SH-aLRT bootstrap support. The complete, unedited lineage B phylogenetic tree is presented in Fig. S4.