Characterization of Severe Acute Respiratory Syndrome Coronavirus 2 (SARS-CoV-2) Infection Clusters Based on Integrated Genomic Surveillance, Outbreak Analysis and Contact Tracing in an Urban Setting

Abstract

Background: Tracing of severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) transmission chains is still a major challenge for public health authorities, when incidental contacts are not recalled or are not perceived as potential risk contacts. Viral sequencing can address key questions about SARS-CoV-2 evolution and may support reconstruction of viral transmission networks by integration of molecular epidemiology into classical contact tracing.

Methods: In collaboration with local public health authorities, we set up an integrated system of genomic surveillance in an urban setting, combining a) viral surveillance sequencing, b) genetically based identification of infection clusters in the population, c) integration of public health authority contact tracing data, and d) a user-friendly dashboard application as a central data analysis platform.

Results: Application of the integrated system from August to December 2020 enabled a characterization of viral population structure, analysis of 4 outbreaks at a maximum care hospital, and genetically based identification of 5 putative population infection clusters, all of which were confirmed by contact tracing. The system contributed to the development of improved hospital infection control and prevention measures and enabled the identification of previously unrecognized transmission chains, involving a martial arts gym and establishing a link between the hospital to the local population.

Conclusions: Integrated systems of genomic surveillance could contribute to the monitoring and, potentially, improved management of SARS-CoV-2 transmission in the population.

Keywords: Nanopore sequencing; community transmission; genomic epidemiology; infection chain; rapid sequencing.

Figure 1.

Integrated genomic surveillance in the Düsseldorf area. Population surveillance sequencing enables the characterization of local SARS-CoV-2 population structure, facilitating the discrimination between clonal hospital outbreaks (here: putative outbreak 1) or simultaneously detected but unrelated SARS-CoV-2 hospital ward cases (here: putative outbreak 2). Viral population surveillance data can also enable the de novo identification of infection clusters in the population based on the genetic data. Added value of genomic surveillance is maximized when genetic data are integrated with complementary epidemiological data or approaches, such as contact tracing or hospital outbreak data. Utilization of viral genetic data by diverse stakeholders is facilitated by providing a user-friendly real-time web application (“dashboard”) for analysis and visualization of the generated viral genomes. Abbreviation: SARS-CoV-2, severe acute respiratory syndrome coronavirus 2.

Figure 2.

Local development of SARS-CoV-2 from September to December 2020. A, Newly diagnosed (red line) and sequenced (blue bars; by sample collection week) cases of SARS-CoV-2 by calendar week of 2020 in Düsseldorf. Horizontal bars indicate sample collection times for 4 hospital outbreaks on different wards (A–D) of Düsseldorf University Hospital. B, Sequenced samples by sample origin. C, Clade composition of surveillance samples by sample collection week, using the NextStrain [21] color scheme. D, Substitutions per sequenced surveillance sample and sample collection line; each dot represents one viral genome, blue line: linear fit. Abbreviation: SARS-CoV-2, severe acute respiratory syndrome coronavirus 2.

Figure 3.

A, Phylogenetic tree of the 320 surveillance samples collected during this study; colours are assigned according to the NextStrain [21] clade system. B, Joint phylogenetic tree of 44 samples from 4 hospital outbreaks (Ward A–D) and 320 surveillance samples. For a description of the outbreaks, see main text. Putative population infection clusters are highlighted in yellow (1–5). Gaps in the corresponding shaded areas correspond to related samples not identified by the greedy clustering algorithm (see Methods). Tree visualization based on iTol [22]. C, Minimum spanning tree (calculated with the Python library networkx version 2.5; visualized with Cytoscape version 3.8.2 and Inkscape version 0.92) visualization of the 4 hospital outbreaks, including all identical or near-identical (distance = 0 or distance = 1) from GISAID and the surveillance sequencing cohort. Samples from GISAID are labelled with their country of origin (Lux = Luxemburg; Net = Netherlands; Swi = Switzerland; Eng = England). The large gray circle represents a cluster of identical and near-identical GISAID samples. Solid lines without number indicate distance = 1 and dashed lines indicate distance = 0 between samples. Abbreviation: SARS-CoV-2, severe acute respiratory syndrome coronavirus 2.