Test tubes

Pooled testing to help Australia tackle the second wave of COVID-19

3 August 2020
From our ‘Thinking outside the box’ series
Pooled testing is a cost-effective method of testing in which a single test is run on a group of samples. Dr Saman Eskandarzadeh and Professor Ben Fahimnia explain how it could help Australia tackle a second wave of COVID-19 whilst reducing the need for tough restrictions.

The threat of a second wave of COVID-19 looms on the horizon, with case numbers jumping in Victoria since the Australian Government’s easing of restrictions. Prime Minister Scott Morrison says outbreaks should be anticipated and that the government plans for localised restrictions to contain the spread of the virus. “Pooled testing” is a cost-effective method for keeping track of the virus by screening targeted/local populations for the virus – shows research at Sydney University.

Prompt detection of the Sars-COV-2 virus is vital to the government’s approach to track and contain the COVID-19 outbreak. Currently, testing occurs only in symptomatic individuals or close contacts of those with confirmed disease. But an infected individual can be shedding virus for days prior the development of any symptoms[1], and some may not get any symptoms at all[2].

Extensive testing of the local population within a “hotspot” can ensure that infected people are identified and quarantined before they infect others. However, it is also beyond Australia’s budget and testing capacity to perform so many tests. A solution to this problem is pooled testing.

Pooled testing, as the name suggests, is the pooling together of a group of samples to run only one test[3]. For example, every week all households in one street could collect their own samples of saliva[4], mail those samples to a testing centre or deposit them in a testing bin (methods already in use for screening programs such as Bowel Cancer screening). The lab then combines together all the saliva samples from one household, or one street, or even one suburb, into a single sample for testing. In this way the number of tests required can be reduced by a factor of 10 or more.

Pooled testing makes sense where the disease prevalence (number of cases in a population at a given time) is low. If the disease prevalence is high, then most of the grouped samples will be positive, which then requires individual testing of all the individuals in that group – rendering pooled testing a useless exercise to reduce test numbers.

Since January 22 this year, Australia has performed over two million tests, with an average of 0. 4% tests returning positive[5]. The incidence (number of new cases in a population) of COVID-19 in Australia since May is 0.04-0.58 cases per 100000, and the prevalence is less than 0.1%. This makes pooled testing perfect for the Australian scenario.

There are three common pooled testing methods: Simple grouping (S method), Square Array grouping (SA method), and Household-based Square Array (HSA method).

In the S method[6], individual samples are combined together and a single test is run on the grouped sample. If the grouped sample tests positive, then each sample is individually tested. For our population of 25 million, if the group size is 10, and the prevalence of infected cases in the population is one percent, then 2,750,000 tests per week are needed. Though this number is a great deal less than 25 million, it is still beyond Australia’s testing capacity.

In the SA method, a grouped sample (with 10 individuals) is further combined with other groups, making the group size 100. Individual samples are arranged in a square array of 10 by 10. A test is run on combined samples of each column group and each row group. As there are 10 columns and 10 rows, 20 tests are run. An individual sample is considered positive if its column group test or its row group test is positive. For the Australian population, again, we need 5,000,000 tests per week. The difference with the SA method is that each sample is tested twice, and this decreases the probability of a false negative.

The HSA method[7] is similar to the SA method, except that each sample is a pooled sample of an entire household. A household sample is considered positive if both its column group and its row group test is positive. In this case, if a sample is tested positive then a whole household needs to be quarantined as the test does not differentiate between members of a household. However, as it is quite likely that all members of the household are infected, under current guidelines the entire household would most likely be quarantined regardless. This method reduces the required number of tests to 2,000,000 per week, assuming a household size of 2.5 people.

With each of these three testing methods, the group sizes can be increased to any number to accommodate any population size and any budget. One criticism is that the number of false negatives may be too high, but given that the tests are repeated every week, even if some infected individuals are not detected in one week, they will most likely be detected in the following week. Thus, pooled testing makes testing of a whole population a feasible goal.

Pooled testing is a method that could enable Australia to test its entire population for COVID-19 every week, eliminating the threat of asymptomatic spread and super-spreaders, and allowing “targeted” quarantine. With the adoption of testing methods like saliva testing (currently not available in Australia), pooled testing allows the rapid lifting of restrictions while avoiding a second wave of infection.

[1] https://www.cdc.gov/mmwr/volumes/69/wr/mm6914e1.htm

[2] https://www.acpjournals.org/doi/10.7326/M20-3012

[3] https://healthpolicy.usc.edu/research/getting-americans-back-to-work-and-school-with-pooled-testing/

[4] https://www.abc.net.au/news/health/2020-07-01/new-saliva-test-for-covid-19-how-accurate-is-it/12406912

[5] https://www.health.gov.au/news/health-alerts/novel-coronavirus-2019-ncov-health-alert/coronavirus-covid-19-current-situation-and-case-numbers

[6] https://healthpolicy.usc.edu/wp-content/uploads/2020/05/USC_Schaeffer_PooledTesting_WhitePaper_FINAL-1.pdf

[7] https://docs.google.com/document/d/1joxMjHdWWo9XLFqfTdNXPQRAfeMjHYEyvVljqNCaKyE/edit