key: cord-0999647-yc5lgclk authors: Snell, L. B.; Fisher, C. L.; Taj, U.; Merrick, B.; Alcolea-Medina, A.; Charalampous, T.; Signell, A. W.; Wilson, H. D.; Betancor, G.; Tan Kia Ik, M.; Cunningham, E.; Cliff, P. R.; Pickering, S.; Galao, R. P.; Batra, R.; Neil, S. J. D.; Malim, M. H.; Doores, K. J.; Douthwaite, S. T.; Nebbia, G.; Edgeworth, J. D.; Awan, A. R.; consortium, The COVID-19 Genomics UK title: Combined epidemiological and genomic analysis of nosocomial SARS-CoV-2 transmission identifies community social distancing as the dominant intervention reducing outbreaks date: 2020-11-18 journal: nan DOI: 10.1101/2020.11.17.20232827 sha: 9fbfc1b9f482dd8060b7756bf005e1da826334b2 doc_id: 999647 cord_uid: yc5lgclk Many healthcare facilities report SARS-CoV-2 outbreaks but analysis of transmission during the first wave is complicated by the high prevalence of infection and limited viral genetic diversity. Furthermore, there is limited evidence on the contribution of different vectors for nosocomial infection or on the effectiveness of interventions. Detailed epidemiological and viral nanopore sequence data were analysed from 574 consecutive patients with a PCR positive SARS-CoV-2 test between March 13th and March 31st, when the pandemic first impacted on a large, multisite healthcare institution in London. During this time the first major preventative interventions were introduced including progressive community social distancing (CSD) policies leading to mandatory national lockdown, exclusion of hospital visitors, and introduction of universal surgical facemask-use by healthcare-workers (HCW). Incidence of nosocomial cases, community SARS-CoV-2 cases, and infection in a cohort of 228 HCWs followed the same dynamic course, decreasing subsequent to CSD and prior to introduction of the main hospital-based interventions. We investigated clusters involving nosocomial cases based on overlapping ward-stays during the 14-day incubation period and SARS-CoV-2 genome sequence similarity. Our method placed 80 (89%) of all 90 probable and definite nosocomial cases into 14 clusters containing a median of 4 patients (min 2, max 19) No genetic support was found for the majority of epidemiological clusters (31/44 70%) and genomics revealed multiple contemporaneous outbreaks within single epidemiological clusters. We included a measure of hospital enrichment compared to community cases to increase confidence in our clusters, which were 1-14 fold enriched. Applying genomics, we could provide a robust estimate of the incubation period for nosocomial transmission, with a median lower bound and upper bound of 6 and 9 days respectively. Six (43%) clusters spanned multiple wards, with evidence of cryptic transmission, and community-onset cases could not be identified in more than half the clusters, particularly on the elective hospital site, implicating HCW as vectors of transmission. Taken together these findings suggest that CSD had the dominant impact on reducing nosocomial transmission by reducing HCWs infection. Severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) was first reported in Wuhan, China in 83 December 2019 (1) , since when over 40 million infections and 1 million deaths have been reported worldwide 84 (2) . New cases in the UK peaked during the first wave after implementation of progressive community social 85 distancing (CSD) (3), starting with an announcement on March 16th that, with few exceptions, the whole UK 86 population must stay at home, later being made mandatory on March 23rd. COVID-19 hospital admission then 87 peaked around 8 days later (4). The first few weeks of the pandemic were unique, due to introduction of SARS-CoV-2 into populations without 89 prior immunity. Society and healthcare organisations were adapting to the pandemic for the first time with 90 frequent new infection control policies and behavioural changes while incidence was still accelerating. During 91 the first wave, elective hospital admissions were often postponed (5) . Currently, hospitals around the world 92 either have, or are preparing for, a further influx of COVID-19 patients, while restarting and rescheduling 93 postponed non-COVID-19 work. It is therefore important to identify evidence from these first few weeks to help 94 healthcare organisations efficiently plan for any subsequent increase in cases. Most COVID-19 transmission studies to date have utilised epidemiological analysis alone to identify outbreaks, 96 including those from the UK (6) (7), China (8) , France (9) and South Korea (10). The main limitation with using 97 epidemiology alone is that when point prevalence is high, for instance at 2.2% in London during the first wave 98 in April 2020 (11), this increases the chance two people in epidemiological contact are independent cases. 99 Furthermore, a wide incubation period of between 2 and 14 days (12,13) makes it difficult to define hospital-100 onset cases as hospital-acquired, as even when they arise several days after admission to hospital the 101 transmission event may still have occurred in the community. To improve the confidence in detecting hospital clusters, epidemiological analysis can be supplemented with 103 genomic data obtained by sequencing viral isolates. For COVID-19, thus far only two published studies have 104 used genomic sequencing to analyse hospital transmission (14) (15) . However, application of genomic data 105 to the spread of SARS-CoV-2 is not straightforward, particularly as during early stages of the pandemic, genetic 106 diversity was low with less than 200 mutations registered in the international GISAID database by April 2020 107 (16) . SARS-CoV-2 also has a low mutation rate with only 2-3 mutations per genome per month (17). These 108 factors increase the chance that two people infected with virus sharing identical genetic sequence are not can provide stronger evidence for a COVID-19 transmission cluster than would be possible with either source 111 of data alone. Nosocomial infection is reported to account for around 10-20% of all confirmed cases (18) (19) (6) (20) often 113 in designated outbreaks (6) (7) (8) (9) (10) and with a crude death rate of up to 30% (19) . The most likely 114 routes of transmission are between patients and involving healthcare workers (HCW) , with potential for all 115 combinations of directionality to occur within the same nosocomial cluster. In each case, the transmitting 116 individual may be either symptomatic or asymptomatic. There is also potential for super-spreading events (21). The aim of this study was to develop a methodology for combining epidemiological and genomics analysis of 123 SARS-CoV-2 transmission to help determine the pathways for nosocomial transmission during the first wave 124 of the pandemic. During this time there was a high incidence in the community (11) and high incidence of is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint The copyright holder for this this version posted November 18, 2020. ; https://doi.org/10.1101 https://doi.org/10. /2020 Setting, clinical characteristics and epidemiology of cases in the first wave of the pandemic 129 SARS-CoV-2 PCR testing for all hospital admissions and inpatients with compatible clinical symptoms of new 130 fever or cough was commenced on March 13th in line with national UK recommendations. Between 80 and 131 150 SARS-CoV-2 PCR tests were performed per day until the end of April when testing increased further 132 (Table S1) , with 637 positive tests on 574 individuals up until March 31st. The majority of positive cases were 133 admitted to hospital (483/574, 84%, Table S4 ), consistent with national recommendations to only test those 134 requiring hospital treatment. Cases detected on admission to hospital had a median reported symptom history 135 of 7 days (interquartile range IQR 3-10, available for 385/483, 80%, of cases; Table S4 ). New positive cases increased rapidly to a peak between March 31st and April 8th, before falling steadily 137 through to the end of April ( Figure 1 ). The number of daily new combined probable and definite nosocomial 138 cases initially increased in line with community diagnoses, but peaked a week earlier than hospital SARS-CoV-139 2 admissions on March 23rd to a maximum of 12 new cases per day. Nosocomial cases then rapidly declined 140 to low levels of 0-2 cases a day during April (Figure 1 ) and none in the following 4 months (data not shown). When the symptom onset date for community cases admitted to hospital is considered, the peak of symptom 142 onset for community and nosocomial cases overlap (Supplemental Figure 1 (Table S4, S7) . The cumulative number of HCWs reporting COVID-19 compatible symptoms and having documented 159 seroconversion from a cohort of 228 HCWs is shown in Figure 1. 43/228 (19%) seroconverted to SARS-CoV-160 2 IgG, with 44% (19/43) having done so by their first follow-up (April 10th) and 95% (41/43) by May 1st (Table 161 S2 ). Of note 13/43 (30%) were asymptomatic. Figure 1 presents the predicted period of peak HCW 162 infectectiousness based on a combination of ± 2 days from date of symptom onset (n=30) or seroconversion 163 data where symptoms were not present (n=13) . The rapid rise in HCW infection is predicted to have been 164 between March 16th and 25th overlapping with similar rapid rises in community cases reported elsewhere (3), 165 and both symptom reporting by our community-onset cases and incidence of nosocomial cases. Table S8 ), where each patient in an 178 epidemiological cluster is classified as either belonging to a combined cluster (coloured bar), or having a 179 haplotype that was not in any combined cluster (grey) or without viral sequence (black). In this manner, 31/44 (70%) epidemiological clusters, including some of the largest, did not have genetic 181 support as they did not have even two cases belonging to any combined cluster ( Figure 2c , Table S7 ). 13/44 182 (30%) included at least one case from two or more combined clusters, indicating cases from multiple 183 contemporaneous outbreaks harboured within an epidemiological cluster ( Figure 2c , Table S7 ). The validity of creating these final 14 clusters was supported by comparing their lineage enrichment compared 193 to community sequences reported to COG-UK consortium CLIMB database during the study period (27). 194 Clusters were enriched by between 1 (for the smallest clusters) and 14-fold (median 4-fold) ( Figure 2d , Table 195 S7). In terms of spatial distribution, 7/12 hospital clusters were contained within single wards, 3 clusters were 196 spread on two wards (in 2 cases this involved adjacent wards), and 2 clusters were spread across more than 197 two wards ( Figure 4 ). Unsequenced community-onset cases are unlikely to be originators of clusters We assessed the likelihood that unsequenced community-onset cases may have served as the originator of 200 clusters, where no community-onset or indeterminate cases could be found, by reviewing cases 201 epidemiologically linked with nosocomial clusters. We excluded cases as potential originators if i) they were 202 diagnosed after the first nosocomial case, or ii) available sequence did not meet cluster inclusion criteria (i.e 203 differed from main cluster viral haplotype by >1 SNP), or iii) if they were not community-onset or indeterminate 204 cases (Table S11) is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint The copyright holder for this this version posted November 18, 2020. ; https://doi.org/10. 1101 Representation of healthcare workers in transmission networks 217 HCW were required to self isolate immediately upon developing COVID-19 compatible symptoms and were 218 not routinely offered SARS-CoV-2 RNA testing. For this reason there are only 20 SARS-CoV-2 sequences 219 from HCW in the dataset. 17 did not share genomic similarity with other cases, which excludes them from 220 contributing to transmission within sequenced clusters. The viral haplotypes from 3 HCW shared genetic 221 similarity with cluster cases, 1 of which (case 280) cared for a nosocomial case (case 60) within cluster STH1 222 and shared the same haplotype. Case 280 therefore can be added to cluster STH1, however 280 is unlikely 223 to be the originator of this cluster as they were diagnosed more than 7 days after the first case in this cluster. is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint The copyright holder for this this version posted November 18, 2020. ; https://doi.org/10. 1101 This combined epidemiological and genomic analysis spans 19 days of intense community transmission in 228 London during the first wave of the pandemic, which coincided with the main period of nosocomial 229 transmission, comprising 90 definite or probable nosocomial cases across our healthcare institution. This was 230 also the period when stepwise increases in public health restrictions were introduced, including repeated 231 instructions to stay at home and socially distance by at least 2 metres in public places from 16th March, through 232 to national lockdown on March 23rd. Our transmission analysis during this important period provides evidence informing methodological 234 approaches to nosocomial SARS-CoV-2 outbreak detection. First, both epidemiological and genomic 235 information is essential for accurate linkage of nosocomial cases into putative SARS-CoV-2 clusters: we could 236 not find genomic support for over half of the transmission clusters discovered using epidemiological data alone, 237 and the rest are significantly reduced in size (Fig 3c, that linked nosocomial cases even with identical genomes might be due to chance due to the high prevalence 244 of that sequence in the community. We included a method to assess enrichment compared with the 245 surrounding London population by comparing data from the COG-UK CLIMB database, which showed cluster 246 enrichment ranging from 1 to 14 fold. Importantly, the lowest enrichment was seen for clusters of size 2 ( definitions. Finally, we were also able to robustly estimate median incubation period for nosocomial cases as 254 between 6-9 days, which, to our knowledge, is the first time the COVID-19 incubation period has been 255 estimated using both genetics and epidemiological data. This is consistent with published evidence that most 256 people become symptomatic within 7 days of exposure (13). In addition, our study showed a similar proportion 257 . CC-BY-NC-ND 4.0 International license It is made available under a perpetuity. is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint The copyright holder for this this version posted November 18, 2020. ; https://doi.org/10.1101/2020.11.17.20232827 doi: medRxiv preprint of probable and definite nosocomial cases could be placed into transmission clusters. Thus when a probable 258 nosocomial case is found it is likely to be a genuine nosocomial case and should be treated as such. A number of clusters (GUY1, GUY2, GUY4, GUY5) involved cryptic transmission across multiple wards which 260 is consistent with untested individuals such as staff or asymptomatic patients, but not patient visitors, as 261 missing links in transmission clusters. For instance, clusters GUY1 and GUY2 involved 4 wards, with an 262 identical viral haplotype circulating with high enrichment compared to the community. Whilst these two clusters 263 meet our cluster definition as defined in Methods, they are not merged due to there being no patient overlap 264 in ward stay between the clusters. Given these were highly enriched clusters, in the same building just a few 265 floors apart, and with no community-onset case as the originator of GUY2, the most parsimonious explanation 266 is cryptic transmission between the two. Similarly in GUY1 there is no patient who clearly could have infected 267 case #551 (see Supplementary Figure 1 ). Lastly, GUY4 and GUY5 involve highly enriched haplotypes across 268 neighbouring wards. These situations suggest cryptic transmission, and 'missing links' which may be HCW or 269 asymptomatic, untested patients. With the above methodological approaches defined, we focussed on the initial period of nosocomial SARS-271 CoV-2 transmission to try and identify factors explaining its rapid rise and fall. There was initially a complete 272 temporal overlap between the increase in community cases, healthcare worker symptoms, and incidence of 273 nosocomial cases. We propose that the most likely explanation is that HCWs were an essential contributor to 274 the intensity of nosocomial transmission during this initial period, potentially by both initiating and sustaining 275 transmission. This is supported by a number of findings. First, reported symptoms in HCW with confirmed 276 infection were before or at the time of lockdown on 23rd March, with their peak infectivity predicted to coincide 277 with peak nosocomial transmission ( Figure 1 ). Second, cryptic transmission between wards was clearly 278 demonstrated, suggesting involvement of untested individuals moving between wards, with HCWs being the 279 most likely vector. Third, we were unable to identify community or indeterminate cases who may have served 280 as originators for more than half of clusters, suggesting involvement of untested individuals such as HCW. Indeed many clusters also occurred on the elective sites, where cases were not knowingly admitted. Finally, 282 we did link 1 of only 20 tested HCWs to a transmission cluster, which although not the originator, is consistent 283 with the potential for HCWs to sustain outbreaks. Whilst HCW seem the most likely source of transmission it 284 is however possible that unrecognised patient-to-patient transmission occurred, for instance by asymptomatic 285 patients or in areas where we do not have epidemiological data, such as radiology departments 286 Subsequently, from around March 16th, progressive CSD rapidly reduced community SARS-CoV-2 287 . CC-BY-NC-ND 4.0 International license It is made available under a perpetuity. is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint The copyright holder for this this version posted November 18, 2020. ; https://doi.org/10.1101/2020.11.17.20232827 doi: medRxiv preprint transmission, which would also decrease the incidence of HCWs infected in the community, lowering their 288 potential to contribute to nosocomial transmission. Consistent with this conclusion, nosocomial infection fell 289 rapidly from March 23rd to the end of the study in line with reported community infections, and before any 290 potential impact from surgical mask use or banning of visitors to hospital. It is harder to interpret events after 31st March. It is notable that SARS-CoV-2 patients continued to be 292 admitted, peaking on 8th April, whilst nosocomial cases fell and remained low. One conclusion is that 293 admission of community-onset cases is in itself unlikely to cause large clusters. Partly this may be due to rapid In summary, this study supports routine use of genome sequencing for SARS-CoV-2 outbreak investigation 300 and provides a framework for data interpretation that calls into question continued reliance on using 301 epidemiology alone. A multi-site clinical trial, COG-UK HOCI, is currently investigating whether provision of 302 real-time sequencing for outbreak analysis can help infection control teams interrupt transmission (30). We 303 provide data supporting a conclusion that HCWs infected in the community are a major contributor to 304 nosocomial transmission during the initial stage of the first wave of the pandemic, which was halted by 305 implementation of CSD policies. With community transmission returning, including to HCWs, there will be an 306 opportunity to assess whether nosocomial transmission returns to the same intensity with policies such as 307 universal mask use in place, or whether additional interventions such as regular staff testing or rapid SARS-308 CoV-2 genome sequencing is required to limit nosocomial transmission. This is particularly important given 309 the challenges re-introducing CSD policies and the need to continue with non-COVID-19 clinical activity. is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint The copyright holder for this this version posted November 18, 2020. ; https://doi.org/10.1101/2020.11.17.20232827 doi: medRxiv preprint . CC-BY-NC-ND 4.0 International license It is made available under a perpetuity. is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint The copyright holder for this this version posted November 18, 2020. ; https://doi.org/10.1101/2020.11.17.20232827 doi: medRxiv preprint Anonymisation 343 All patients and samples were anonymised using a unique patient and sample identification number. Ward 344 stays were anonymised using the following schema: locations beginning with 'A' represent ambulatory or 345 outpatient locations numbered A01 to A60; if prefixed with 'ED' represent emergency departments; if starting 346 with 'IC' represent an intensive care unit; each patients usual residence was named 'Home' followed by their 347 anonymised patient ID, and inpatient wards were anonymised using three letter international airport codes. is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint The copyright holder for this this version posted November 18, 2020. ; https://doi.org/10. 1101 was in a transmission cluster with strong epidemiological evidence linking them to 4 patients who all had 370 variants C3096T, G13627T, and C15540T. All three of these variants were found in all of the other patients in In the second phase of cluster construction, genetic information was added by considering the viral haplotype 384 for each patient in the epidemiological cluster. Epidemiological clusters were split into combined 385 epidemiological + genetic clusters where every patient has the same viral haplotype. These combined 386 epidemiological plus genetic clusters are termed "seed" clusters. In the third phase, additional patients were added to seed clusters based on the following criteria: the patient 388 added must overlap with at least one patient in the seed cluster (where overlap is defined as above); the added 389 patient has a viral haplotype that differs only one SNP from the seed cluster viral haplotype. Patients who met 390 criterion 1 (overlap) but whose viral haplotypes differed by 2 SNPs from the seed cluster viral haplotype were 391 added to the cluster if at least one of the "missing" SNP positions was called as an N due to insufficient 392 sequence coverage in consensus sequence generation. For two clusters, (STH4, STH5) phase 3 involved 393 merging two different seed clusters whose viral haplotypes differed from each other by one SNP. There are two clusters that are exceptions to the above rules: GUY3 and GUY4. These two clusters were 395 added upon manual inspection of the combined epidemiological and genetic data. In both cases, two patients 396 whose viral haplotype differed by a single SNP overlapped on the same ward during their incubation period, 397 . CC-BY-NC-ND 4.0 International license It is made available under a perpetuity. is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint The copyright holder for this this version posted November 18, 2020. ; https://doi.org/10.1101/2020.11.17.20232827 doi: medRxiv preprint and a third patient, whose viral haplotype was identical to that of one of the two previously mentioned patients 398 was present on an adjacent ward at the same time, also during their incubation period. These two clusters 399 violate the criterion given above that seed clusters should be on a single ward, but are nonetheless considered 400 as transmission clusters because the two wards (Samaritan and Hedley Atkins) in both cases are spatially 401 adjacent and with frequent sharing of staff across wards, especially at night..These clusters also showed high 402 enrichment compared to community haplotypes (see below.) Calculation of Hospital compared with Community Enrichment for cluster viral haplotypes 404 Enrichment of viral haplotype frequencies amongst nosocomial and community cases were calculated as 405 follows (Table S8 and Fig 2) . For a given haplotype, the hospital frequency (H) was calculated by dividing the 406 number of patients in the dataset by the total number of sequenced patients in our dataset (370). COG-UK 407 consortium data in COV_GLUE was extracted on 29th June 2020 to assess community haplotype frequencies 408 (27) . To assist with SQL queries, a view from several tables was created with the following SQL query: The number of patients with the given haplotype in the community population was calculated by determining 424 the number of sequences in the community population that had all of the SNP variants in the given haplotype, . CC-BY-NC-ND 4.0 International license It is made available under a perpetuity. is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint The copyright holder for this this version posted November 18, 2020. ; https://doi.org/10. 1101 and no more. For this purpose, any homoplasic SNPs that had been removed from consideration when 426 constructing patient viral haplotypes within our study population (see section "Haplotype construction for is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint The copyright holder for this this version posted November 18, 2020. ; https://doi.org/10.1101/2020.11.17.20232827 doi: medRxiv preprint is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint The copyright holder for this this version posted November 18, 2020. ; https://doi.org/10.1101 https://doi.org/10. /2020 is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint The copyright holder for this this version posted November 18, 2020. ; https://doi.org/10.1101 https://doi.org/10. /2020 genetic haplotype (Figure 2b ). The numbers in the ward indicate the number of patients from the given cluster 513 inside that ward during their incubation period (as defined in Figure 2a is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint The copyright holder for this this version posted November 18, 2020. ; https://doi.org/10. 1101 Table S5, Table S8 is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint The copyright holder for this this version posted November 18, 2020. ; https://doi.org/10. 1101 transferred to Ward C (SFO), which sets a lower bound on when they must have been infected (as they were is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint The copyright holder for this this version posted November 18, 2020. ; https://doi.org/10.1101 https://doi.org/10. /2020 is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint The copyright holder for this this version posted November 18, 2020. ; https://doi.org/10.1101 https://doi.org/10. /2020 624 SUPPLEMENTAL is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint The copyright holder for this this version posted November 18, 2020. ; https://doi.org/10.1101 https://doi.org/10. /2020 is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint The copyright holder for this this version posted November 18, 2020. ; https://doi.org/10.1101/2020.11.17.20232827 doi: medRxiv preprint is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint . CC-BY-NC-ND 4.0 International license It is made available under a perpetuity. is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint The copyright holder for this this version posted November 18, 2020. ; https://doi.org/10. 1101 Leadership, supervision, samples, logistics and metadata curation: is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint The copyright holder for this this version posted November 18, 2020. is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint The copyright holder for this this version posted November 18, 2020. ; https://doi.org/10. 1101 is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint The copyright holder for this this version posted November 18, 2020. ; https://doi.org/10.1101/2020.11.17.20232827 doi: medRxiv preprint . CC-BY-NC-ND 4.0 International license It is made available under a perpetuity. is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint The copyright holder for this this version posted November 18, 2020. ; https://doi.org/10.1101/2020.11.17.20232827 doi: medRxiv preprint . CC-BY-NC-ND 4.0 International license It is made available under a perpetuity. is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint The copyright holder for this this version posted November 18, 2020. ; https://doi.org/10. 1101 Northumbria University, 62 Northumbria University, 63 South Tees Hospitals NHS Foundation Trust, 64 1058 . CC-BY-NC-ND 4.0 International license It is made available under a perpetuity. is the author/funder, who has granted medRxiv a license to display the preprint in (which was not certified by peer review) preprint The copyright holder for this this version posted November 18, 2020. ; https://doi.org/10. 1101 A pneumonia outbreak associated with a new coronavirus of probable bat origin COVID-19) in the UK. GOV NHS roadmap to safely bring back routine operations Nosocomial transmission of COVID-19: a retrospective study of 66 hospital-657 acquired cases in a London teaching hospital The potential transmission of SARS-CoV-2 from patients with negative RT-PCR swab 661 tests to others: two related clusters of COVID-19 outbreak Whole-genome sequencing to track SARS-CoV-2 transmission in nosocomial outbreaks Emergence of genomic diversity and recurrent mutations in SARS-CoV-2 Temporal signal and the phylodynamic threshold of SARS-CoV-2 Clinical Characteristics of 138 Hospitalized Patients With Nosocomial COVID-19 infection: examining the risk of mortality. The COPE-Nosocomial 687 Study (COVID in Older PEople) Scientific Advisory Group for Emergencies Clustering and superspreading potential of SARS-CoV-2 infections in Hong Kong Disinfection of environments in healthcare and non-healthcare settings potentially contaminated with 694 Transmission of SARS-CoV-2: implications for infection prevention precautions Pandemic peak SARS-CoV-2 infection and seroconversion rates in London 700 frontline health-care workers SARS-CoV-2 seroprevalence and asymptomatic viral carriage in healthcare workers: a 702 cross-sectional study Surveillance definitions for COVID-19 MAJORA: Continuous integration supporting decentralised sequencing for SARS-706 Cold Spring Harbor Laboratory Leadership, supervision, funding acquisition, samples, logistics, and metadata curation Leadership, supervision, metadata curation, samples, logistics, sequencing and analysis Professor Matthew W Loose PhD 54, and Joanne Watkins MSc 33 Metadata curation, sequencing, analysis, leadership, supervision and software and analysis tools Dr Matthew Bull PhD33 , and Dr Sam Nicholls PhD 15 Leadership, supervision, visualisation, sequencing, analysis and software and analysis tools: 781 Professor David M Aanensen PhD 1 Sequencing, analysis, samples, logistics, metadata curation, and visualisation Metadata curation, sequencing, analysis, visualisation, software and analysis tools Sequencing, analysis, visualisation, metadata curation, and software and analysis tools Maksimovic FD sport science 36 Dr Inigo Martincorena 1 ,Dr Tamyo Mbisa 7, Kathryn McCluggage MSC 973 36 Dr Gaia Nebbia PhD Ruis 3 ,Dr Christine M Sambles PhD PhD 21 ,Karla Spellman FD 36 ,Thomas D Stanton BSc 19 Dr Edith E Vamos PhD 4,Dr Tetyana Vasylyeva24 Dr Shirelle Burton-Fanning MD 66, Dr Vicki Chalker 7 Dr Anastasia Kolyva PhD 51 Dr Nicholas W Machin MSc 2 , 47, Mailis Maes M.phil3 Dr Divya Shah PhD40, Nicola Sheriff BSc 67, Dr Graciela Sluga 1016 Dr Theocharis Tsoleridis PhD55 Dr Rebecca Williams BMBS 31, Dr Iona Willingham56 Dr Sarah Wyllie 70 , and Jamie Young BSc 3. 1021 1022 Software and analysis tools Dr Will Rowe PhD 15, and Dr Igor Siveroni PhD96. 1024 1025 Visualisation Wellcome Sanger Institute, 2 Public Health England, 3 University of Cambridge, 4 Health Data Research 1029 Public Health Agency 1031 Departments of Infectious Diseases and Microbiology 11 Division of Virology The Francis Crick Institute, 13 Cambridge Institute for Therapeutic Immunology and Infectious Disease 17 Queen Elizabeth Hospital, 18 Heartlands Hospital 27 Big Data Institute Basingstoke Hospital, 30 Centre for Genomic Pathogen Surveillance Hampshire Hospitals NHS Foundation Trust, 32 University of Southampton, 33 Public Health Wales 1044 NHS Trust, 34 Cardiff University, 35 Betsi Cadwaladr University Health Board, 36 Cardiff and Vale 1045 Great Ormond Street NHS Foundation Trust 44 Imperial College Healthcare NHS 1048 45 NIHR Health Protection Research Unit in HCAI and AMR University of Exeter, 50 Royal Devon and Exeter NHS Foundation Trust Northern Lincolnshire & Goole 1055 NHS Foundation Trust, 58 Clinical Microbiology Hub for Biotechnology in the Built Environment, Northumbria University, 61 NU-OMICS Trust, 66 Newcastle Hospitals NHS Foundation Trust, 67 County Durham and Darlington NHS 1060 Foundation Trust, 68 Centre for Enzyme Innovation University Hospitals Coventry and Warwickshire, 73 Warwick Medical School and Institute of Precision 1063 74 Genomics Innovation Unit, Guy's and St Foundation Trust, 75 Centre for Clinical Infection & Diagnostics Research 76 Department of Infectious Diseases, King's College London, 77 Guy's and St Hospitals NHS Foundation Trust, 78 Centre for Clinical Infection and Diagnostics Research Department of Infectious Diseases, Guy's and St Thomas' NHS Foundation Trust