SARS-CoV-2 has caused outbreaks among vaccinated and unvaccinated people in large public gatherings (Brown et al., 2021) and congregate settings, such as prisons (Hagan et al., 2021), homeless shelters (Nanduri et al., 2021), and long-term care facilities (Pray et al., 2021). On August 29, 2021, the U.S. government began Operation Allies Welcome, an interagency effort to resettle eligible persons from Afghanistan to the United States. Evacuees were housed in temporary congregate housing at eight U.S. military bases and one hotel facility (Safe Haven sites). We report on COVID-19 outbreak investigations and responses in five sites during November 2021–February 2022. This activity was reviewed by CDC and was conducted consistent with applicable federal law and CDC policy.¹

Evacuees received immigration and medical processing, including COVID-19 and other vaccinations, before being resettled by resettlement agencies. Services were provided by interagency U.S. government staff, military service members, NGO staff, and contract staff.

CDC and public health teams investigated and responded to SARS-CoV-2 outbreaks, and outbreak data were collected through case investigation interviews. These data, site census data, and staffing rosters were used to calculate attack rates for evacuees, military, and civilian staff. Quarantine and isolation, serial testing, and other site-specific data were collected. Two sites identified SARS-CoV-2 variants through whole genome sequencing (WGS). Case demographic data were described across sites and used to construct epidemic curves (Figure 1). Site-specific mitigation and CDC response strategies were defined.

Across five sites (Table 1, Figure 1), there were 939 evacuees who tested positive for SARS-CoV-2, and among the four sites reporting staff data (sites 1, 2, 4, 5), 771 staff members tested positive. Staff in sites 1, 2, and 4 were categorized as civilian or military, but staff in site 5 were categorized differently. Attack rates in sites 1, 2, and 4 were highest among military (107/1000), followed by civilian staff (43/1000), and lowest among evacuees, all of whom were required to remain on-site (31/1000).

At site 1, there were five evacuee cases (attack rate 2/1000), four military cases (attack rate 9/1000), and 23 civilian staff cases (attack rate 153/1000) during November 1, 2021−December 18, 2021. The Delta variant was identified in four cases through sequencing (WGS).

At site 2, there were 67 evacuee cases (attack rate 28/1000), 112 military cases (attack rate 146/1000), and 14 civilian staff cases (attack rate 60/1000) during November 1, 2021−January 16, 2022. The Omicron variant was identified through WGS. All evacuees entering the medical clinic were tested for SARS-CoV-2. All staff were required to wear surgical masks and encouraged to wear N95 masks indoors while interacting with evacuees.

At sites 1 and 2, a 3-day staff serial testing program was implemented and CDC-recommended quarantine and isolation guidelines at the time were followed for cases and close contacts (Quarantine and Isolation: Centers for Disease Control and Prevention, 2021).

At site 3, there were 355 evacuee cases (attack rate 41/1000) (staff data not available) during November 1, 202−February 9, 2022. Cases were identified through daily 10% evacuee population convenience sample testing, testing of those entering the clinic, and serial testing during quarantine. Evacuees who tested positive for SARS-CoV-2 were isolated with other cases or their families on an isolation floor in separate barracks. Exposed barracks were put under “quarantine in place” for 10 days and were tested every 3 days. Any evacuee who tested positive under quarantine was moved to isolation and the quarantine period restarted for the rest of the cohort.

At site 4, there were 46 evacuee cases (attack rate 18/1000), 71 military cases (135/1000), and 17 civilian staff cases (125/1000) during November 1, 202−January 26, 2022. A serial testing program was conducted for evacuees and civilian staff; however, military were only tested if symptomatic. CDC-recommended quarantine and isolation guidelines were followed for cases and close contacts; congregate setting guidelines were followed starting December 29, 2021 (Quarantine and Isolation: Centers for Disease Control and Prevention, 2021; Quarantine and Isolation Specific Settings Archived: Centers for Disease Control and Prevention, 2021).

At site 5, there were 464 evacuee cases (attack rate 31/1000), 101 contract medical staff cases (testing began December 5, 2021; attack rate 140/1000), and 430 other civilian staff and military cases (testing began December 26, 2021; attack rate 178/1000). Evacuees who tested positive were isolated in separate buildings. Close contacts were tested on entry and exit from quarantine. Staff followed state guidelines for individual-based, elective testing, with subsequent isolation and quarantine according to CDC guidelines.

This report is subject to limitations. First, data were collected across five different states, with different partners, testing strategies, and application of quarantine and isolation strategies. Numbers of cases were likely underreported due to gaps in testing, asymptomatic status, stigma, and fear of resettlement delays. For site 5, the data could not be categorized into the same “role” designations as sites 1 through 4. We were unable to include data on timing of vaccination. Some evacuees received the COVID-19 vaccination before arriving in the U.S. Also, we did not have the data to include person-time in the calculation of the attack rates to account for amount of time people worked at Safe Havens, thus potentially underestimating the attack rates among staff. Finally, we did not have the data to include weekly number of persons at risk in the attack rate calculations, thus limiting our ability capture the dynamic nature of Safe Haven populations across time.

Even in vaccinated populations, COVID-19 outbreaks can occur. Overall data indicated staff, (free to leave sites), compared to evacuees (restricted to sites), had higher attack rates. Institutional policies for congregate settings, such as vaccination and boosting, wearing well-fitting masks while indoors, serial testing staff and evacuees with rapid antigen tests, applying isolation and close contact quarantine, remain critical strategies to slow transmission of SARS-CoV-2. Application of these strategies served to mitigate transmission in communities as evacuees were resettled and staff returned home.