Factors Associated With SARS-CoV-2 Transmission in Settings of High COVID-19 Vaccination Coverage: A Case-Control Study

Résumé fourni par la source

Despite the availability of effective vaccines, community exposures continue to drive the coronavirus disease 2019 (COVID-19) pandemic across the United States. Though vaccination coverage is high in some regions, uptake has lagged elsewhere (1). Characterization of risk factors remains critically important for informing policies on nonpharmaceutical interventions that target the highest-risk activities and reduce severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) transmission while preserving economic and societal structures. While prior case-control studies have identified diverse risk factors for SARS-CoV-2 transmission (2–4), there is a critical need to understand how increasing vaccination coverage affects the risk of SARS-CoV-2 transmission, particularly with the increased spread of the delta (B.1.617.2) variant (5, 6). We determined risk factors for SARS-CoV-2 transmission in the City and County of San Francisco, California, from April to June 2021, when the proportion of fully vaccinated residents increased from 27.4% to 62.6% (7). Distribution of Demographic Characteristics and SARS-CoV-2 Transmission Risk Factors Among COVID-19 Cases and Controls in the City and County of San Francisco, April 7–June 8, 2021 Abbreviations: CI, confidence interval; COVID-19, coronavirus disease 2019; LGBTQ, lesbian, gay, bisexual, transgender, or queer; OR, odds ratio; SARS-CoV-2, severe acute respiratory syndrome coronavirus 2; SF, San Francisco; SRO, single room occupancy. a Adjusted for age and Latinx ethnicity. b The OR for age was not adjusted for age. c The OR for Latinx ethnicity was not adjusted for Latinx ethnicity. d We defined the SF Bay Area as the 9-county area including Alameda, Contra Costa, Marin, Napa, San Francisco, San Mateo, Santa Clara, Solano, and Sonoma counties. e The rest of California, elsewhere in the United States, or internationally. Our objective was to identify factors associated with transmission of SARS-CoV-2 in San Francisco between April 7, 2021, and June 8, 2021. We used an unmatched case-control design, recruiting individuals ≥18 years of age. We recruited cases from individuals reported to the San Francisco Department of Public Health who tested positive for SARS-CoV-2 infection by reverse-transcriptase polymerase chain reaction and controls from those who tested negative. Given the narrow time window, we did not match for date of test result. Ineligibility criteria included residing outside of San Francisco County, living in a congregate setting (e.g., a long-term care facility or homeless shelter), or being unable to confirm one’s birthdate or test date. Cases were also ineligible if they were unaware of their status or had not yet been interviewed by the case investigation team. Participants were contacted within 3 weeks of their test dates. Both cases and controls who were fully or partially vaccinated at the time of the testing were also ineligible because of their different risk for infection. Moreover, we only included individuals who were unvaccinated at the time of testing in the final analysis. All eligible study participants completed a telephone interviewer-led, closed-ended questionnaire focused on exposures and activities during the 2-week period prior to their test date. Interviews assessed demographic factors; household exposures (numbers of persons and bedrooms in the home); occupational exposures (job category, workplace setting, and mask use at work); frequency of outdoor and indoor dining and visits to bars; and community and travel exposures. For travel exposures, we defined the San Francisco Bay Area as the 9-county area including Alameda, Contra Costa, Marin, Napa, San Francisco, San Mateo, Santa Clara, Solano, and Sonoma counties. Interviews were conducted in English, Spanish, Mandarin, or Tagalog. We constructed a logistic regression model to assess differences in community exposures between cases and controls. Because the number of participants in the control group was low (n = 27), we used Firth’s penalized maximum likelihood regression method, which yields more accurate point estimates and confidence intervals for sparse data than standard maximum likelihood methods (8, 9). All results were adjusted for age quartile and self-reported Latinx ethnicity. Statistical analyses were conducted using R software (10). This study was approved by the University of California, San Francisco, Human Research Protection Program. Informed consent was obtained from all study participants. Among 258,026 records, 1,202 cases (SARS-CoV-2–positive) and 193,137 controls (SARS-CoV-2–negative) were sampled. Of these, 957 cases and 1,358 controls were reached by telephone, and of those reached by phone, 184 cases and 44 controls were screened eligible, consented, and were interviewed (see Web Figure 1, available at https://doi.org/10.1093/aje/kwac045). Fifty-one cases and 17 controls enrolled early in the study were excluded because of vaccination at the time of testing (eligibility criteria were subsequently changed), leading to a final analytical sample of 133 cases and 27 controls. Overall, cases were younger than controls (38.3% of cases were aged 18–29 years vs. 18.5% of controls), more likely to be female (49.6% vs. 37.0%), and more likely to identify as Latinx (37.6% vs. 14.8%) (Table 1). Controls were more likely than cases to identify as lesbian, gay, bisexual, transgender, or queer (18.5% controls vs. 6.8% cases) and White (40.7% vs. 15.8%). The odds of testing positive were higher in Latinx participants than in White participants (adjusted OR (aOR) = 6.04, 95% confidence interval (CI): 1.89, 22.52 (adjusted only for age)). Housing status was not associated with testing positive. Regarding workplace exposures, 74.4% of cases and 77.8% of controls were employed; 24.8% of cases and 22.2% of controls were unemployed, retired, or students (see Web Table 1). Participants who traveled outside of the 9-county San Francisco Bay Area (within the rest of California, elsewhere within the United States, or internationally) during the 2 weeks prior to testing, versus no travel or travel only within the Bay Area, had higher adjusted odds of testing positive (aOR = 7.07, 95% CI: 2.05, 37.41). Participants who traveled and had an overnight stay, versus no travel, had higher adjusted odds of testing positive (aOR = 3.09, 95% CI: 1.05, 10.96). In this setting of high COVID-19 vaccination coverage, before the delta (B.1.617.2) SARS-CoV-2 variant was widespread, the odds of testing positive were significantly greater among persons who had recently traveled outside of the San Francisco Bay Area or had an overnight stay during such travel. At the end of study data collection, the 9-county San Francisco Bay Area population was highly vaccinated, with the proportion of the entire population having completed a full series ranging from 41.6% in Solano County to 66.9% in Marin County (7). While travel has been identified as a risk factor for SARS-CoV-2 infection in other studies (2, 4, 11), our findings illustrate the importance of travel as a risk factor when local vaccination coverage is high. San Francisco has been successful in implementing population-level interventions and has experienced low COVID-19 case rates compared with other US cities (7); nonetheless, these findings underscore how local efforts may be ineffective in preventing imported cases. We did not find an increased transmission risk associated with restaurant or bar dining (indoors or outdoors) or working in congregate environments, such as stores, offices, or schools. While transmission in these settings might have been reduced because of high local vaccination coverage, we interpret these negative findings with caution, given the small sample size, especially as prior studies have identified these factors as being associated with transmission. Consistent with the findings of other studies, we found an increased risk of SARS-CoV-2 transmission among Lat

Ce résumé expose les affirmations des auteurs. BNTIC ne l’interprète pas comme une validation indépendante des résultats.