Outbreak of SARS-CoV-2 B.1.617.2 (Delta) Variant Infections Among Incarcerated Persons in a Federal Prison — Texas, July–August 2021

Investigation and Response

On July 12, 2021, 18 persons incarcerated in a federal prison in Texas reported COVID-19–like symptoms to BOP health services staff members. All 18 received positive SARS-CoV-2 test results using the Abbott BinaxNOW COVID-19 Ag Card (rapid antigen) test; 11 were fully vaccinated. Three of these persons had reported to the on-site clinic on July 8 with symptoms including coryza, cough, headache, myalgia, or rhinorrhea but did not receive SARS-CoV-2 testing at that time.^§ The 18 persons with positive test results lived in two interconnected units (unit A and unit B) that operated as a single cohort and housed 233 persons in 2- to 10-person cells without doors. Standard COVID-19 prevention protocols that were in place among incarcerated persons included mandatory masking in common areas, cohorting of housing units for daily activities, and head-to-toe sleeping arrangements. Among staff members, prevention protocols included mandatory masking and mandatory daily COVID-19 symptom screening and temperature checks (5).^¶ Before the outbreak, incarcerated persons moved freely between units A and B and were together for meals, recreation, and work; they did not have contact with incarcerated persons housed in other units. After initial identification of COVID-19 cases, unit A was designated as a quarantine unit for persons with negative test results, and unit B was designated as a medical isolation unit for COVID-19 patients. Staff members assigned to units A and B rotated between these two units and to other units on the basis of daily staffing needs.

During July 12–August 14, 2021, BOP staff members offered same-day SARS-CoV-2 rapid antigen testing to all 233 persons in units A and B reporting symptoms or known exposures; the entire quarantined cohort received testing from BOP during July 12–13 and again on July 14, July 19, July 22, August 2, and August 10 with a combination of rapid antigen and RT-PCR tests.** SARS-CoV-2 testing among staff members was voluntary and was performed off-site by staff members’ health care providers. A subset of 70 incarcerated persons in units A and B consented to a secondary investigation for which they reported symptom data through a questionnaire and provided nasal midturbinate swabs daily for up to 20 days after symptom onset. Specimens were tested by RT-PCR.^†† Viral culture was attempted for RT-PCR–positive specimens from a nonrandom subset of participants.^§§ Genomic sequencing was attempted for one RT-PCR–positive specimen from each participant, when possible.

COVID-19 vaccination was voluntary for BOP staff and incarcerated persons. In 2020, BOP worked with CDC to develop a vaccine prioritization plan in which all staff members were offered vaccination first, followed by incarcerated persons. Among incarcerated persons, those aged ≥65 years and those with underlying medical conditions associated with severe COVID-19 were the first to receive a COVID-19 vaccine. In this prison, the Pfizer-BioNTech vaccine was the first available, with first doses administered to incarcerated persons in January 2021.^¶¶ Staff vaccination coverage in this report includes only doses administered as part of the BOP occupational health program. BOP was unable to determine the number of staff members who were vaccinated through other providers.

Information on vaccination, demographic characteristics, and underlying medical conditions was extracted from BOP electronic medical records for all 233 persons living in units A and B. Demographic characteristics, underlying medical conditions, and COVID-19–associated hospitalizations and deaths were compared by vaccination status and, among vaccinated persons, by vaccine product received. Attack rates were compared by demographic and medical characteristics, vaccination status and vaccine product, and time since vaccination. Descriptive statistics were calculated. Differences between groups were assessed using chi-square or Fisher’s exact tests. P-values <0.05 were considered statistically significant, adjusted for multiple comparisons using the Bonferroni correction method. Statistical analyses were performed using SAS (version 9.4; SAS Institute). This activity was reviewed and approved by the BOP Research Review Board and CDC and conducted consistent with applicable federal law and CDC policy.***

Among 233 incarcerated persons, 185 (79%) of whom were fully vaccinated, 172 (74%) received positive SARS-CoV-2 test results during July 12–August 14 (Supplementary Figure, https://stacks.cdc.gov/view/cdc/109901). Among a subset of 70 symptomatic persons providing swabs for serial testing, no significant difference was found in the median interval between reported symptom onset and last positive RT-PCR result in vaccinated versus unvaccinated persons (9 versus 11 days, respectively; p = 0.37) (Figure). Virus was cultured from one or more specimens from five of 12 (42%) unvaccinated and 14 of 37 (38%) fully vaccinated persons for whom viral culture was attempted. Genomic sequencing confirmed the AY.3 sublineage of the Delta variant in 58 specimens from 58 persons.

Vaccination coverage was 79% among incarcerated persons in units A and B. Among fully vaccinated persons, 93 of 122 (76%) Pfizer-BioNTech recipients and 0 of 50 (0%) Moderna recipients had been vaccinated ≥4 months before the outbreak (p<0.001). A larger proportion of Pfizer-BioNTech recipients had diabetes (p = 0.02) or hypertension (p<0.001) than Moderna or Janssen COVID-19 vaccine recipients, and a higher proportion of Pfizer-BioNTech and Janssen recipients had a history of smoking (p<0.001) than Moderna recipients (Table 1).

Attack rates were higher among unvaccinated persons (39 of 42; 93%) than among fully vaccinated persons (129 of 185; 70%) (p = 0. 002) and among persons vaccinated ≥4 months before the outbreak (83 of 93; 89%) than among those vaccinated 2 weeks to 2 months before the outbreak (19 of 31; 61%) (p<0.001) (Table 2).

Among both persons with and without a previous SARS-CoV-2 infection, the attack rate was lower among fully vaccinated versus unvaccinated persons (1 of 21 [5%] versus 4 of 7 [57%], p = 0.008; 128 of 164 [78%] versus 35 of 35 [100%], p<0.001) (Supplementary Table, https://stacks.cdc.gov/view/cdc/109901). Among fully vaccinated persons without a previous SARS-CoV-2 infection, the attack rate was higher among Pfizer-BioNTech recipients (99 of 117; 85%) than among Moderna recipients (19 of 35; 54%) (p<0.001).

Among 172 infected persons, four (2%) were hospitalized for COVID-19, including three (8%) of 39 unvaccinated patients, and one (1%) of 129 fully vaccinated patients (p = 0.04). One (3%) of the unvaccinated hospitalized patients required endotracheal intubation and mechanical ventilation and died in the hospital (Table 1).^†††

Nine of 275 (3%) staff members, four of whom worked in units A or B, reported a positive SARS-CoV-2 test result during the outbreak and were restricted from work per BOP policy. BOP administered COVID-19 vaccine to 37% of staff members in the prison.

Discussion

This study demonstrates the potential for SARS-CoV-2 Delta variant outbreaks in congregate settings including correctional and detention facilities, even among resident populations with high vaccination coverage. In this outbreak involving almost three fourths of the incarcerated population in the affected housing units, fewer hospitalizations and deaths occurred among vaccinated than unvaccinated persons, highlighting vaccination as an important strategy to reduce serious COVID-19–related illness and death in congregate settings. In addition, the high number of infections in vaccinated persons, comparable duration of positive RT-PCR test results after symptom onset regardless of vaccination status, and presence of infectious virus in specimens from both unvaccinated and vaccinated infected persons underscore the importance of implementing and maintaining multiple COVID-19 prevention strategies in settings where physical distancing is challenging, even when vaccination coverage is high. Prevention strategies that were in place during this outbreak, including promptly separating infected and exposed persons and cohorting housing units for daily activities, might have prevented the outbreak from spreading to other areas of the prison.

Three of the four hospitalizations and the only death occurred in unvaccinated persons. These findings are consistent with a previous study in which vaccination with a COVID-19 mRNA vaccine (Pfizer-BioNTech or Moderna) reduced the risk for hospitalization associated with Delta variant infection (6). These findings reinforce the critical importance of vaccination in reducing risk for severe illness and death from SARS-CoV-2 Delta variant infections, particularly in congregate settings.

Natural infection with SARS-CoV-2 confers some degree of immunity, although the duration of protection is unknown (7). In this outbreak, the lowest attack rate occurred among fully vaccinated persons with previous infection, highlighting the importance of vaccination, even among persons with previous infection. In addition, attack rates in persons without previous infection were higher among Pfizer-BioNTech recipients than among Moderna recipients. In a recent study, the Moderna vaccine was found to be more effective at preventing COVID-19–related hospitalizations among U.S. adults without immunocompromising conditions (6). In this outbreak, attack rates were also higher in persons who were vaccinated ≥4 months before the outbreak compared with persons vaccinated more recently. Because all persons vaccinated ≥4 months before the outbreak received the Pfizer-BioNTech vaccine, determining the independent impact of vaccine product versus time since vaccination was not possible. Additional research is warranted to assess the duration of vaccine-induced and natural immunity, as well as the duration of infectious virus shedding by vaccinated and unvaccinated infected persons.

BOP records indicate that nearly two thirds of staff members in this prison were unvaccinated, and at least nine were infected during this outbreak. In addition, during the 2 weeks before the outbreak, community transmission was high.^§§§ SARS-CoV-2 can be introduced into correctional facility populations and back into the community through daily entry and exit of staff members and interfacility transfers of incarcerated persons, and the identification of a single viral lineage among all sequenced specimens in this outbreak suggests a single introduction of the virus into the prison (8). Bidirectional connections between correctional facilities and communities highlight the importance of high vaccination coverage among both staff members and incarcerated persons, early diagnostic testing, routine screening testing when community transmission is high, maintaining consistent assignments of staff members for each housing unit, and excluding staff members from work when they are symptomatic or have COVID-19 (5,9).

The findings in this report are subject to at least five limitations. First, although rapid antigen testing can identify cases quickly, its limited sensitivity for detecting infections in asymptomatic patients can underestimate attack rates (10). Second, transmission might have preceded initial identification of cases, resulting in an underestimation of total cases. Third, it is uncertain whether lower attack rates by vaccine product were caused by differences in waning vaccine-induced immunity, varying levels of protection among vaccine products, or differences in exposure level among persons who received different vaccine products. Fourth, testing was not mandatory for BOP staff members, limiting the ability to confirm the total numbers of COVID-19 cases. Finally, RT-PCR–positive specimens were not selected randomly for viral culture and thus are not representative of all vaccinated and unvaccinated participants.

During a COVID-19 outbreak in a federal prison involving the highly transmissible SARS-CoV-2 Delta variant, transmission was high among vaccinated and unvaccinated persons. Although hospitalizations, deaths, and attack rates were higher among unvaccinated than vaccinated persons, the duration of positive serial test results was similar between these two groups, and infectious virus was cultured from both vaccinated and unvaccinated participants. Widespread vaccination among incarcerated persons and staff members in coordination with other prevention strategies, including early diagnostic testing for all persons with any COVID-19–like symptoms, screening testing, medical isolation, quarantine, masking, and physical distancing where possible, remain critical to limiting SARS-CoV-2 transmission and COVID-19–related illness and death in congregate settings, including correctional and detention facilities (5).