학술논문
Analysis of two choir outbreaks acting in concert to characterize long- range transmission risks through SARS-CoV-2, Berlin, Germany, 2020.
Document Type
Article
Author
Reichert, Felix; Stier, Oliver; Hartmann, Anne; Ruscher, Claudia; Brinkmann, Annika; Grossegesse, Marica; Neumann, Markus; Werber, Dirk; Hausner, Marius; Kunze, Mareike; Weiß, Bettina; Michel, Janine; Nitsche, Andreas; an der Heiden, Matthias; Kriegel, Martin; Corman, Victor Max; Jones, Terry Carleton; Drosten, Christian; Brommann, Tobias; Buchholz, Udo
Source
Subject
*COVID-19
*CHOIRS (Musical groups)
*SUPERSPREADING events
*WHOLE genome sequencing
*PARTICULATE matter
*SARS-CoV-2
*
*
*
*
*
Language
ISSN
1932-6203
Abstract
Background: Superspreading events are important drivers of the SARS-CoV-2 pandemic and long-range (LR) transmission is believed to play a major role. We investigated two choir outbreaks with different attack rates (AR) to analyze the contribution of LR transmission and highlight important measures for prevention. Methods: We conducted two retrospective cohort studies and obtained demographic, clinical, laboratory and contact data, performed SARS-CoV-2 serology, whole genome sequencing (WGS), calculated LR transmission probabilities, measured particle emissions of selected choir members, and calculated particle air concentrations and inhalation doses. Results: We included 65 (84%) and 42 (100%) members of choirs 1 and 2, respectively, of whom 58 (89%) and 10 (24%) became cases. WGS confirmed strain identity in both choirs. Both primary cases transmitted presymptomatically. Particle emission rate when singing was 7 times higher compared to talking. In choir 1, the median concentration of primary cases' emitted particles in the room was estimated to be 8 times higher, exposure at least 30 minutes longer and room volume smaller than in choir 2, resulting in markedly different estimated probabilities for LR transmission (mode: 90% vs. 16%, 95% CI: 80–95% vs. 6–36%). According to a risk model, the first transmission in choir 1 occurred likely after 8 minutes of singing. Conclusions: The attack rate of the two choirs differed significantly reflecting the differences in LR transmission risks. The pooled proportion of cases due to LR transmission was substantial (81%; 55/68 cases) and was facilitated by likely highly infectious primary cases, high particle emission rates, and indoor rehearsing for an extended time. Even in large rooms, singing of an infectious person may lead to secondary infections through LR exposure within minutes. In the context of indoor gatherings without mask-wearing and waning or insufficient immunity, these results highlight the ongoing importance of non-pharmaceutical interventions wherever aerosols can accumulate. [ABSTRACT FROM AUTHOR]