Rapid Transmission of Severe Acute Respiratory Syndrome Coronavirus 2 in Detention Facility, Louisiana, USA, May–June, 2020

Megan Wallace; Allison E. James; Rachel Silver; Mitsuki Koh; Farrell A. Tobolowsky; Sean Simonson; Jeremy A. W. Gold; Rena Fukunaga; Henry Njuguna; Keith Bordelon; Jonathan Wortham; Melissa Coughlin; Jennifer L. Harcourt; Azaibi Tamin; Brett Whitaker; Natalie J. Thornburg; Ying Tao; Krista Queen; Anna Uehara; Clinton R. Paden; Jing Zhang; Suxiang Tong; Danielle Haydel; Ha Tran; Kaylee Kim; Kiva A. Fisher; Mariel Marlow; Jacqueline E. Tate; Reena H. Doshi; Theresa Sokol; Kathryn G. Curran


Emerging Infectious Diseases. 2021;27(2):421-429. 

In This Article



Facility X is a medium-security local jail that houses up to 800 detained persons. Before the COVID-19 pandemic, the facility operated at nearly 100% capacity. On May 7, the facility was at ≈85% capacity because of a reduction in occupancy in response to COVID-19. Detained persons from 6 dormitories (A–F) were enrolled in this prospective cohort investigation. Five dormitories (A–E) had detained persons with laboratory-confirmed COVID-19 cases; dormitory F, which housed a detained person with COVID-19 symptoms and negative SARS-CoV-2 test results, was enrolled because of proximity to dormitories A, B, and D. All detained persons with suspected and confirmed COVID-19 were moved to medical isolation, and persons within the dormitories were quarantined as a cohort.

Testing Strategy and Cohorting by Test Result

Nasopharyngeal swab specimens were collected for initial SARS-CoV-2 testing on day 0 for all consenting persons residing in dormitories A–F (Figure 1). Persons who had positive results by real-time reverse transcription PCR (rRT-PCR) were moved to the designated SARS-CoV-2–positive dormitories upon facility receipt of results (<24 hours after specimen collection). Serial testing was offered on day 4 to detained persons who tested negative for SARS-CoV-2 on day 0, and again on day 14 for persons who tested negative on day 4. To assess persistence of viral shedding, detained persons testing positive on day 0 or day 4 were offered testing 14–15 days and 19–27 days after their first positive test result.

Figure 1.

Rapid transmission of SARS-CoV-2 in detention facility, Louisiana, USA, May–June 2020. Enrollment and follow-up at each timepoint for detained persons (n = 143) in dormitories A–E and F. The sequence of testing for all enrolled dormitories is shown, along with the number of persons who were positive or negative for SARS-CoV-2 by real-time reverse transcription PCR and percentage of total at each timepoint. Red boxes indicate SARS-CoV-2 positive, and blue boxes indicate SARS-CoV-2 negative. *The first positive test result for SARS-CoV-2 among persons detained occurred on the following dates in each dormitory: April 7 in A, April 9 in B and C, April 17 in D, and April 23 in E. Introduction in dormitory F likely occurred between May 11 and May 29. †One inconclusive result was considered negative; ‡One inconclusive result was considered positive. §16 persons were tested on May 26 only, 14 on May 27 only, and 2 on May 26 and June 3. ¶10 persons were tested on May 28 only, 1 on May 29 only, 1 on June 3 only, and 6 on May 28 and June 3. SARS-CoV-2, severe acute respiratory syndrome coronavirus 2.

In dormitory F, where all detained persons tested negative for SARS-CoV-2 on day 0, a serial testing strategy was not used. Rather, a second survey and repeat test was conducted on day 18.

Dormitory Survey and Symptoms, Concurrent Conditions, and Behavioral Risk Assessment

The investigation team administered a structured dormitory survey among facility staff to assess physical layout, capacity, activities, and practices. During day 0 testing, detained persons completed a self-administered, paper-based questionnaire of demographics, symptoms in the preceding 2 months and 2 weeks, facility exposures, and preventive measures. On the day of each subsequent test, detained persons received an abbreviated self-administered, paper-based questionnaire of symptoms experienced since the last testing day. The team verbally verified responses with detained persons and assisted as necessary. Medical history data were abstracted from facility medical records. Data were deidentified and entered into a secure database (Research Electronic Data Capture, version 8.8.0; Vanderbilt University, https://redcap.vanderbilt.edu).

Laboratory Testing

Nasopharyngeal swab specimens collected for the investigation during May 7–June 3 were immediately placed on dry ice and sent by courier to the Louisiana Office of Public Health Laboratory for SARS-CoV-2 testing by using the CDC 2019-Novel Coronavirus (2019-nCoV) Real-Time rRT-PCR Diagnostic Panel. Cycle threshold (Ct) values for 2 viral nucleocapsid protein genes (N1 and N2) were obtained for each specimen; Ct values <40 cycles for both N1 and N2 were considered positive for SARS-CoV-2.[14] All samples that were positive at the Louisiana Office of Public Health Laboratory were refrozen and shipped to CDC for viral culture by using Vero-CCL-81 cells.[15] Positive viral culture for SARS-CoV-2 replication-competent virus was confirmed in cells that showed a cytopathic effect by using rRT-PCR.

Nucleic acid was extracted from 41 rRT-PCR–positive specimens or isolates and subjected to Oxford Nanopore MinION Sequencing (https://nanoporetech.com) according to published protocols;[16] consensus sequences were generated by using Minimap version 2.17 (https://github.com/lh3/minimap2) and Samtools version 1.9 (https://www.htslib.org). Representative full-genome sequences were downloaded on August 28, 2020, from GISAID (https://www.gisaid.org), and phylogenetic relations were inferred by using maximum-likelihood analyses implemented in TreeTime (https://evol.bio.lmu.de/_statgen/software/treetime) and the Nextstrain pipeline.[17] Sequences were submitted to GenBank and GISAID.


We performed descriptive analyses for the population demographics (age, sex, race/ethnicity), underlying medical conditions (respiratory disease, diabetes, hypertension, other cardiovascular disease, other condition), obesity (body mass index >30 kg/m2), tobacco use, and dormitory characteristics (capacity at start of the investigation, toilets/sinks, showers per person). Overall cumulative incidence and dormitory cumulative incidence for each test day were calculated.

We calculated descriptive statistics for Ct values and culture results, stratified by symptom status. The rRT-PCR analyses used the Ct value reported for the N1 genetic target because N1 and N2 approximate each another.[18] Persons were categorized as presymptomatic, symptomatic, postsymptomatic, or asymptomatic on the basis of symptoms at sample collection. Any CDC-listed coronavirus symptom with a reported onset date on or after March 29, 2020, the illness onset date of the first reported COVID-19 case in the facility, was included in analyses.[19] Persons were classified as symptomatic if they reported ≥1 present or ongoing symptom. If 2 courses of illness were distinguishable from the symptom data, in which multiple symptoms were reported to occur with symptom onsets ≥14 days apart and the first course of illness (earlier dated symptoms) was reported to have resolved, only the symptoms reported closer to the date of testing were used for classification. Postsymptomatic persons were those who reported symptoms that had resolved before the first positive test result or before the start of the investigation (day 0) for those who were tested and remained negative during the investigation. Persons reporting symptoms whose surveys were missing current symptom status were considered symptomatic if the onset date was ≤10 days the start of the investigation. Presymptomatic persons reported ≥1 symptom with onset after their first positive test result and had no previously reported symptoms. Asymptomatic persons reported no symptoms throughout the investigation. Persons were classified as having an unknown symptom status if any symptom data were missing and no symptoms were reported. Ct value and culture results were graphed by days from symptom onset and original dormitory.

To compare individual symptoms, facility exposures (bunk sleeping location, travel out of dormitory, exposure to someone visibly ill), and preventive measures (handwashing, mask use) by SARS-CoV-2 test result, we performed bivariate analyses by using Fisher exact tests for proportions. Analyses were completed by using R statistical software version 4.0.0 (The R Foundation, https://www.r-project.org) and SAS 9.4 software version 6.2.92 (SAS Institute Inc., https://www.sas.com).


This activity was determined to meet the requirements of public health surveillance as defined in 45 CFR 46.102(l).[2] All persons provided voluntary oral consent for testing and to complete questionnaires.