In March and April 2020, more than 900 COVID-19 cases were diagnosed at Cook County Jail in Chicago, across all 10 of its housing divisions and 13 buildings. Investigators from CDC, the Chicago Department of Public Health, Cermak Health Services and the Cook County Sheriff's Office asked a network question: who was in the same place at the same time as whom — and where did infections cluster more than chance would predict?
Their answer pointed in two directions. Among detained people, isolating those with COVID-19 appeared to work. Among staff, infections clustered along shared shifts, suggesting spread between them.
Networks linking staff members and detained persons to the jail's housing divisions; people with COVID-19 in blue (staff) and red (detained). Credit: CDC, Morbidity and Mortality Weekly Report.
Why jails
Correctional and detention facilities have been hit disproportionately by COVID-19 because people share space and staff and detained people move around inside them. Network analysis — the study of the elements of a network and the connections between them — has helped trace tuberculosis and shown how social networks spread sexually transmitted infections. This was the first known use of it to describe a COVID-19 outbreak in a U.S. correctional or detention facility.
Who was counted
The study covered March 1 to April 30, 2020, and 5,884 people: 2,041 staff members of the Sheriff's Office with shift records, and detained people with at least one bed assignment.
| Detained people | Staff members | |
|---|---|---|
| Had COVID-19 | 472 (12.3%) | 198 (9.7%) |
| Counted as a case | a positive RT-PCR test | a reported positive test (confirmed) or COVID-19 symptoms (probable) |
| Tested | those with symptoms or who were close contacts; 1,080 were tested | not systematically; required to report symptoms or positive results |
| A "connection" | a bed or cell assignment in the same place at the same time | a shift in the same place at the same time |
During the outbreak the jail stopped visits (March 15), suspended programs (March 23), converted cells to single occupancy, and required masks for staff (April 2) and detained people (April 13).
What the networks showed
- Infections clustered among staff. Of 103,701 shift connections between staff members, 1.4 percent linked two people with COVID-19 — significantly more than the 0.9 percent chance alone would produce. In March, staff members with COVID-19 had more interactions on average than those without (377 against 321), a difference not seen in April.
- Busier divisions, more sick staff. The more staff connections a division had, the higher the share of its staff with COVID-19 (a strong correlation, r = 0.86). For detained people there was no significant relationship.
- Isolation worked for detained people. Across all 1,214,462 connections among detained people, links between two people with COVID-19 were more common than expected (3.2% against 1.5%) — but those counts include people in medical isolation and at off-site locations. Leaving those out, links between infected detained people were significantly less common than expected (1.0% against 1.1%).
Person-to-person networks of staff members and detained persons. Credit: CDC, Morbidity and Mortality Weekly Report.
What to do with it
The findings support finding and isolating people with COVID-19 quickly and limiting movement between divisions. For staff, they point to keeping groups of staff together, keeping shift locations consistent and enforcing masks. A network approach can show where transmission runs higher than expected, and so where to act: fewer staff gatherings, closed breakrooms, no contact sports.
Share of staff and of detained persons with COVID-19 against each division's number of connections. Credit: CDC, Morbidity and Mortality Weekly Report.
Limits
- The networks capture only people in the same place at the same time, not other routes such as contaminated surfaces.
- Shift data covered only Sheriff's Office staff assigned to housing divisions — not transportation or kitchen staff, volunteers or contractors.
- Staff were not systematically tested, so their cases and connections may be undercounted; staff also go home to their households and communities after every shift.
- Detained people and staff were analyzed separately, though contact between the two groups likely matters, and exposures outside the jail or from staff without symptoms were not assessed.
Sources
Based on "Network Characteristics and Visualization of COVID-19 Outbreak in a Large Detention Facility in the United States — Cook County, Illinois, 2020," by Uzay Kırbıyık, Alison M. Binder and colleagues (CDC; Chicago Department of Public Health; Cermak Health Services; Cook County Health; Cook County Sheriff's Office; University of Illinois at Chicago), Morbidity and Mortality Weekly Report, Centers for Disease Control and Prevention; a work of the United States government in the public domain.
许可协议： CC0 1.0（公有领域） · 改编自 www.cdc.gov
1
0
0
0
留言
