A systematic outbreak investigation of SARS-CoV-2 transmission clusters in a tertiary academic care center

Abstract

Background: We sought to decipher transmission pathways in healthcare-associated infections with severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) within our hospital by epidemiological work-up and complementary whole genome sequencing (WGS). We report the findings of the four largest epidemiologic clusters of SARS-CoV-2 transmission occurring during the second wave of the pandemic from 11/2020 to 12/2020.

Methods: At the University Hospital Basel, Switzerland, systematic outbreak investigation is initiated at detection of any nosocomial case of SARS-CoV-2 infection, as confirmed by polymerase chain reaction, occurring more than five days after admission. Clusters of nosocomial infections, defined as the detection of at least two positive patients and/or healthcare workers (HCWs) within one week with an epidemiological link, were further investigated by WGS on respective strains.

Results: The four epidemiologic clusters included 40 patients and 60 HCWs. Sequencing data was available for 70% of all involved cases (28 patients and 42 HCWs), confirmed epidemiologically suspected in house transmission in 33 cases (47.1% of sequenced cases) and excluded transmission in the remaining 37 cases (52.9%). Among cases with identical strains, epidemiologic work-up suggested transmission mainly through a ward-based exposure (24/33, 72.7%), more commonly affecting HCWs (16/24, 66.7%) than patients (8/24, 33.3%), followed by transmission between patients (6/33, 18.2%), and among HCWs and patients (3/33, 9.1%, respectively two HCWs and one patient).

Conclusions: Phylogenetic analyses revealed important insights into transmission pathways supporting less than 50% of epidemiologically suspected SARS-CoV-2 transmissions. The remainder of cases most likely reflect community-acquired infection randomly detected by outbreak investigation. Notably, most transmissions occurred between HCWs, possibly indicating lower perception of the risk of infection during contacts among HCWs.

Keywords: COVID-19; Epidemiologic cluster; Nosocomial outbreaks; Outbreak investigation; SARS-CoV-2 cluster; Whole genome sequencing.

Fig. 1

Ward floor plan of Clusters a–d with colored squares representing SARS-CoV-2 positive patients. Numbers represent individual patients, which may have been hospitalized in different rooms within a ward. The virus-strains of patients represented by a blue square are identical, whereas patients with unique virus-strains are indicated by a green square. Patients with no sequencing data available are represented by grey squares. There were two patients in cluster a in adjacent rooms with the same virus-strain, but different to the dominant strain within the cluster (*). The index-patient of cluster b (patient number 4) was hospitalized in both wards. Cluster d occurred on a ward containing a dedicated intermediate care unit (IMC). In this figure includes 39 rather than 40 patients as one patient-contact occurred on another unit

Fig. 2

NGS-results of the different clusters in a SNP-tree (single-nucleotide polymorphism tree). The blue dots represent samples from HCWs, whereas patient samples are labelled with green dots. NC_045512.2 (SARS-CoV-2 isolate Wuhan-Hu-1) was used as the reference genome. a Cluster A: One patient and three HCWs are missing because the lineage could not be determined due to too low coverage (however, they are unique virus strains or have no genetic association to all other strains within this cluster). b Cluster B: Three patients and one HCW are missing because the quality control failed (e.g. low coverage, high percentage of minority variants), they could not be included in the SNP based analysis (however, they have several SNPs difference to all other strains within this cluster) c Cluster C: Two patients and six HCWs are missing because the quality control failed (e.g. low coverage, high percentage of minority variants), they could not be included in the SNP based analysis (one additional HCW in the cluster, otherwise different strains). d Cluster D: Five patients and 14 HCWs are missing because the quality control failed (e.g. low coverage, high percentage of minority variants), they could not be included in the SNP based analysis