Table 1 Surveillance strategies evaluated for detection of COVID-19 outbreaks in a LTCF. Strategies differ in how swabs and tests are apportioned to patients and staff. Arrows (→) indicate order of priority for testing cascades. Test, RT-PCR test; swab, nasopharyngeal swab; symptoms, COVID-like symptoms; admission, arrival of new patient to the LTCF

Single indication

Administer tests to any individuals indicated for testing, up to the daily testing capacity. If the number of individuals indicated exceeds the number of tests available, select randomly among them.

Symptoms (severe) [reference strategy]

No

Symptoms (any)

No

Admission

No

Random

Each day, randomly administer tests to individuals in a particular demographic group.

Random (patients)

Yes

Random (HCWs)

Yes

Random (all: patients, HCWs and ancillary staff)

Yes

Cascade

A combination of indications and random testing. First, use indications to administer tests according to a given order of priority. Then, if any tests remain, distribute them randomly among patients not otherwise indicated for testing.

Symptoms (severe) → Symptoms (mild) → Random (patients)

Yes

Symptoms (severe) → Symptoms (mild) → Admission → Random (patients)

Yes

Symptoms (severe) → Admission → Random (patients)

Yes

Symptoms (severe) → Admission → Symptoms (mild) → Random (patients)

Yes

Group testing

Classic two-stage sample pooling, modified to account for clinical urgency of severe COVID-19. First, administer individual tests to any patients or staff presenting with severe symptoms. Then, if at least one test remains, pool clinical specimens together and run one test across this group sample. If the test result is positive, individually re-swab and re-test all included individuals to identify cases. The maximum number of samples per group test was varied from 2 to 64.

Symptoms (any)

No

Admission

No

Random (patients) (always maximizes number of specimens per group test)

No

Random (HCWs) (always maximizes number of specimens per group test)

No