Superspreaders drive the largest outbreaks of hospital onset COVID-19 infections.

Christopher J R Illingworth ,Daniela De Angelis,Iliana Georgana,Theodore Gouliouris,Surendra Parmar,Kayleigh Grainger,Anna Yakovleva,Myra Hosmillo,Matthew Routledge,Nick K Jones,Mark Ferris,Sushmita Sridhar,Lucy Rivett,Tom Dymond,Grant Hall,Tom Fieldman,Theresa Feltwell,Ben Warne,Sally Forrest,Dominic Sparkes,Yasmin Chaudhry,Chris Jackson,Luke W Meredith ,Sarah Caddy ,Martin D Curran ,Effrossyni Gkrania‐Klotsas ,William L Hamilton ,Laura Caller ,Sharon J Peacock ,Fahad Khokhar ,Ian Goodfellow ,Michael P Weekes ,Christopher T Workman ,Aminu S Jahun ,Charlotte J Houldcroft ,Nicholas M Brown ,A.i Popay ,Stephen J Baker ,Malte L Pinckert ,M Estée Török

doi:10.7554/elife.67308

Abstract

SARS-CoV-2 is notable both for its rapid spread, and for the heterogeneity of its patterns of transmission, with multiple published incidences of superspreading behaviour. Here, we applied a novel network reconstruction algorithm to infer patterns of viral transmission occurring between patients and health care workers (HCWs) in the largest clusters of COVID-19 infection identified during the first wave of the epidemic at Cambridge University Hospitals NHS Foundation Trust, UK. Based upon dates of individuals reporting symptoms, recorded individual locations, and viral genome sequence data, we show an uneven pattern of transmission between individuals, with patients being much more likely to be infected by other patients than by HCWs. Further, the data were consistent with a pattern of superspreading, whereby 21% of individuals caused 80% of transmission events. Our study provides a detailed retrospective analysis of nosocomial SARS-CoV-2 transmission, and sheds light on the need for intensive and pervasive infection control procedures.

Highlights

Reducing the spread of SARS-CoV-2 is a crucial priority for controlling and limiting the impact of the COVID-19 pandemic
Our method combines knowledge of SARS-CoV-2 infection dynamics with viral genome sequences and data describing the movements of patients and health care workers (HCWs) within the hospital
Applying our method to these data, we generated maximum likelihood reconstructions of the pattern of transmission events occurring within five infection clusters, each of which was centred on a ward at Cambridge University Hospitals NHS Foundation Trust (CUH)

Summary

Introduction

Reducing the spread of SARS-CoV-2 is a crucial priority for controlling and limiting the impact of the COVID-19 pandemic. Individual variations from this mean can be of vital importance (Lloyd-Smith et al, 2005); a study of SARS-CoV-2 in Hong Kong suggested that 80% of transmission events resulted from only 19% of cases (Adam et al, 2020). Superspreader events are widely reported to play a key role in the spread of the virus in community settings (Shen et al, 2004; Kucharski and Althaus, 2015; Hamner, 2020; Ebrahim and Memish, 2020; Lemieux et al, 2020). As such, testing of asymptomatic health care workers (HCW) has been proposed as a means to reduce viral spread and protect the workforce and patients (Black et al, 2020; Jones et al, 2020)

Methods

Results

Conclusion