Testing Everyone: COVID-19 in 26 Detroit Nursing Homes, Spring 2020

Checking nursing home residents for symptoms was not enough to find the coronavirus. When health workers in Detroit tested every resident of 26 skilled nursing facilities (SNFs) in spring 2020, 44% turned out to be infected. Within 21 days of diagnosis, 37% of them were hospitalized and 24% died.

Credit: CDC, Morbidity and Mortality Weekly Report*.*

Test everyone, then test again

The report, by the Detroit Health Department, CDC and local hospital and university partners, describes point prevalence surveys: testing every resident of a facility at once, not only residents with symptoms. Twelve facilities had a second round, along with on-site infection prevention and control (IPC) support.

"Tested" counts residents not already known to be infected.

Across all 26 facilities, 2,773 residents were tested at some point through May 8; 1,207 tested positive, 446 were hospitalized and 287 died. The share of residents infected ranged from 16% to 66% between facilities. In the 12 that were tested twice, the share of newly identified cases fell from 35% to 18%.

Residents' first positive tests, hospital admissions and deaths, by date. Credit: CDC, Morbidity and Mortality Weekly Report*.*

What the testing made possible

• Finding silent cases. Repeated surveys identified COVID-19 early, including in residents with no symptoms.
• Separating residents. Results guided cohorting — grouping infected residents and the staff caring for them, apart from everyone else — and the creation of COVID-19 care units.
• Targeting help. IPC assessment and consultation after testing helped facilities tackle the suspected causes of ongoing spread:
• • incomplete cohorting of residents and staff;
  • the virus being reintroduced — by admitting residents whose COVID-19 status was unknown, or residents who left for routine outpatient care such as hemodialysis;
  • too little space to give residents of unknown status private rooms.
• Better care. Earlier diagnosis can mean closer monitoring, such as more frequent vital signs, and faster transfer to hospital when needed.

New cases were still being found in the second round, but 21-day hospitalization and death rates fell over the period. Testing plus IPC support, the authors conclude, might have reduced transmission.

What should follow a positive result

The report says testing should lead to prompt, specific action:

1. Transmission-based precautions for residents' care, and keeping staff who test positive away from work.
2. Strict cohorting of residents and staff.
3. Active clinical monitoring of confirmed cases.
4. Safe transfers to and from outside facilities.
5. Ending precautions, if a test-based strategy is used.

After a confirmed case, CDC now recommends retesting all residents and staff who previously tested negative — for example every 3–7 days — until no new cases are found, with health departments stepping up IPC support alongside.

Limits

Staff, a likely source of spread, were not tested on the same days as residents, and their results were not included. The first survey came about a month after the virus arrived in most facilities, so some positives may have been residents who had recovered but still tested positive, and 14 days between surveys may have been too long. Testing methods differed between rounds, and second-round data covered only 12 of the 26 facilities.

Sources

Based on Guillermo V. Sanchez, Caitlin Biedron, Lauren R. Fink, et al., "Initial and Repeated Point Prevalence Surveys to Inform SARS-CoV-2 Infection Prevention in 26 Skilled Nursing Facilities — Detroit, Michigan, March–May 2020," Morbidity and Mortality Weekly Report, volume 69, Centers for Disease Control and Prevention, posted as an Early Release on July 1, 2020; a work of the United States government in the public domain.