Abstract
SARS-CoV-2 is the etiological agent responsible for the COVID-19 pandemic. It is estimated that only 10 aerosol-borne virus particles are sufficient to establish a secondary infection with SARS-CoV-2. However, the dispersal pattern of SARS-CoV-2 is highly variable and only 10– 20% of cases are responsible for up 80% of secondary infections. The heterogeneous nature of SARS-CoV-2 transmission suggests that super-spreader events play an important role in viral transmission. Super-spreader events occur when a single person is responsible for an unusually high number of secondary infections due to a combination of biological, environmental, and/or behavioral factors. While super-spreader events have been identified as a significant factor driving SARS-CoV-2 transmission, epidemiologic studies have consistently shown that education settings do not play a major role in community transmission. However, an outbreak of SARS-CoV-2 was recently reported among 186 children (aged 10-17) and adults (aged 18 +) after attending an overnight summer camp in Texas in June 2021. To understand the transmission dynamics of the outbreak, RNA was isolated from 36 nasopharyngeal swabs collected from patients that attended the camp and 19 control patients with no known connection to the outbreak. Genome sequencing on the Oxford Nanopore platform was performed using the ARTIC approaches for library preparation and bioinformatic analysis. SARS-CoV-2 amplicons were produced from all RNA samples and >70% of the viral genome was successfully reconstructed with >10X coverage for 46 samples. Phylogenetic methods were used to estimate the transmission history and suggested that the outbreak was the result of a single introduction. We also found evidence for secondary transmission from campers to the community. Together, these findings demonstrate that super-spreader events may occur during large gatherings of children.
Main Text
The rapid spread of Severe Acute Respiratory Syndrome Coronavirus 2 (SARS-CoV-2), the virus driving the Coronavirus disease 2019 (COVID-19) pandemic, is partly due to the highly infectious nature of the virus. Approximately 10 aerosol-borne virus particles are sufficient to establish a novel infection (1). However, the dispersal pattern (K) of SARS-CoV-2 is highly variable. Approximately 60–75% of infected patients do not transmit SARS-CoV-2 infection, and only 15% of cases are responsible for up to 80% of secondary infections (2). The heterogenoeus dispersal dynamics of SARS-CoV-2 suggest that super-spreader events play a major role in transmission dynamics.
Super-spreader events result in the transmission of SARS-CoV-2 from an infected individual to an unusually high number of persons due to some combination of environmental, biological and/or behavioral factors. Such events have been reported frequently, especially involving mass gatherings (reviewed (3)). Given the well-documented role of super-spreader events in the transmission of SARS-CoV-2, there is considerable concern over the risk of super-spreader events occurring in educational settings among children, their caregivers, and staff. Studies in Europe (4–11), North America (12–15), and Oman (16) have repeatedly shown that educational settings are not major drivers of SARS-CoV-2 transmission in the community, despite the dense populations and prolonged exposure of children and adolescents in classrooms. Even when infected, children usually develop milder disease (17) and emit fewer virus particles than adults (18). Outbreaks among schools report low secondary attack rates (4,19) with child-to-child transmission, which is significantly lower than adult-to-adult and adult-to-child transmission (19–22). However, several studies have noted an important distinction between transmission among young children (age 6-12) and adolescents (23,24) with reports of larger and more frequent outbreaks among secondary schools (25). Rather, household transmission appears to be a more important driver of SARS-CoV-2 spread (14,22,26).
While educational settings have not been a major source of SARS-CoV-2 transmission, large outbreaks among children and adolescents have been reported in several summer camps throughout the United States of America (27–29), with primary attack rates as high as 46% (28) in Georgia and even 76% in Wisconsin (29). However, some reports have demonstrated success preventing and event halting the transmission of SARS-CoV-2 with multi-layered mitigation approaches, including but not limited to vaccination, pre-arrival quarantine, pre-post arrival testing, masking, and social distancing (30–32). However, studies investigating SARS-CoV-2 outbreaks in summer camps have relied on diagnostic and epidemiological methods alone and have lacked phylogenetic approaches that could provide insights into the patterns of viral transmission.
Such an outbreak occurred at an overnight summer camp in Texas during June 2021. Four hundred and fifty-one individuals attended including 364 youths (ages 10-17) and 87 adults (age 18 or older). At the conclusion of the camp, attendees began to show symptoms and tested positive for SARS-CoV-2 by RT-PCR. An investigation of the outbreak led to the identification of 186 SARS-CoV-2 positive cases (33). The vaccination rate among attendees was low (19% fully vaccinated and 6% partially vaccinated), pre-arrival testing was not required for camp attendance, and no post-arrival testing was conducted. The primary attack rate within the camp was 41% (48% among unvaccinated attendees and 20% among vaccinated), which is consistent with outbreaks reported in overnight summer camps in Georgia (28).
Notably, the Texas outbreak occurred at the beginning of the Delta variant wave in the US, while cases were still low in the community but rapidly increasing. To better understand the pattern of viral transmission, we obtained 55 SARS-CoV-2 positive nasopharyngeal swabs collected by the University of Texas Medical Branch, Galveston, TX from patients who attended the summer camp (36) as well as from unrelated community members who tested positive during the same time period (17). RNA was extracted from the swabs as previously described (34) and used for deep sequencing with the well-defined ARTIC nanopore approach and associated bioinformatic pipeline (34). SARS-CoV-2 reads were detected in all samples, and 70% genome coverage greater than 10X was achieved for 43 samples (31 camp attendees and 13 Galveston County residents). The results were aligned with 3 additional genomes that were isolated from an Arlington, TX family that tested positive following contact with asymptomatic attendees after the camp concluded. The Arlington genomes were sequenced using similar methods (35). All genomes are available on GSAID EPI_ISL_12486186-213 (Table 1).
Sequence Metadata. All relevant metadata are provided including sample names, dates the samples were collected, type of patients (camper or community member) and the GSAID accession number.
The evolutionary history of the genomes was inferred using a maximum likelihood approach implemented with IQ-Tree (36) (Figure 1A). The phylogenetic tree demonstrated that all the SARS-CoV-2 genomes collected from camp attendees shared a single common ancestor, including the genomes collected from the family from Arlington, TX that became sick after contact with camp attendees. This suggests that the outbreak was initiated from a single infected individual. Genomes from campers that shared suspected risk factors, including bus or cabin assignment, did not cluster together. Similarly, the genomes collected from siblings did not appear to cluster together, suggesting that transmission occurred while the youths were at the camp as opposed to when they returned home. However, these negative results should be interpreted with caution because there was insufficient diversity to resolve all branches within the phylogeny. The lack of diversity also prevented generation of a transmission tree, which may have provided additional epidemiologically relevant insights. This was unsurprising, given the short duration of the outbreak (5 days) relative to the generation time (delta variant mean=4.7 days, 95% CI=4.1-5.6 days)(37). Interestingly, several genomes from patients that did not attend the camp clustered among the campers, suggesting that transmission occurred from campers to the community.
Evolutionary history of genomes collected from camp attendees. The genomes collected from camp attendees with at least 70% genomic coverage were aligned. (A) The evolutionary history was inferred with IQ-Tree using the Maximum likelihood method based on the general time reversible model. A discrete gamma distribution was used to model variation among sites and allow for invariable sites. The consensus tree following 1000 UF bootstraps is shown. Genomes collected from camp attendees are shown in red and genomes collected from community members are shown in black. Bootstrap values greater than 70 are provided. (B) All complete genomes (n=4085) available on GSAID collected in Harris and Galveston Counties between May and July were combined with the genomes sequenced in this study. The evolutionary history was inferred using the Nextstrain platform, which using Augar to preform bioinformatic analysis including aligning and filtering the genomes, generating a phylogeny with IQ-Tree, and removing polytomies, inferring node dates, and pruning branches with TimeTree. The phylogeny was visualized with Auspice and Figtree. The branches of the phylogeny were colored to indicate clade membership. The cluster containing genomes collected from campers is indicated by the black circle and expanded (C). Genomes isolated from campers and community members are depicted in red and black, respectively. Genomes obtained from GSAID are depicted in blue.
To further investigate the role of the camp outbreak in community transmission, all complete genomes published on GISAID collected in Galveston County and Harris County, Texas (4085 genomes in total) were compared to the genomes collected from the campers. Phylogenetic analysis using the Nextstrain platform (38) identified 29 genomes that clustered among the camper’s genomes that were collected after the camp concluded between June 28, 2021 and July 30, 2021, suggesting that community transmission originating from the camp outbreak continued at least until the end of July (Figure 1B-C).
It is essential to understand the mechanisms that support super-spreader events, such as the Texas summer camp outbreak of 2021, and the risk factors associated with them. Taken together, our study demonstrates that the Texas camp outbreak was likely the result of a single introduction that spread in the camp environment and eventually into the community, creating a chain of transmission that persisted until at least July 30, 2021. This study also highlights the risks associated with overnight summer camps that do not employ adequate prevention strategies, such vaccination, pre- and post-arrival testing, etc. Furthermore, this study is the first of our knowledge to combine epidemiological, genomic, and phylogenetic approaches to investigate an outbreak at an overnight camp. It illustrates the importance of multidisciplinary collaborations between public health specialists and evolutionary virologist in responding to this and future pandemics.
Data Availability
All data produced in the present study are available upon reasonable request to the authors.
Acknowledgments
The authors of this manuscript gratefully acknowledge the following funding sources: ‘Data collection grant’ and ‘COVID-19 funding’ from IHII to ALR; NIH grant R21AI151725 from to ALR; NIH grant R24 AI120942 to SW; The Sealy and Smith Foundation to SW and UTMB; CDC Contract 200-2021-11195 to ALR and Kempner Fellowship to DMS; FAPESP Fellowship 2019/27803 to RRGM, the West African Center for Emerging Infectious Diseases to SW, BM, AM, RM, NS.
References
