Co-authored with Zeph Landau, a research scientist in computer science at UC Berkeley
How do we get from a country currently under necessary but crippling social distancing measures to one where schools, restaurants and non-essential businesses are open, people are going to work, and the economy is thriving?
Almost all proposed solutions require being able to identify and isolate contagious people quickly. And that means huge amounts of testing.
The U.S. is testing about 200,000 people a day -- way fewer than the millions a day which most experts say will be needed. We'll need to be regularly testing anywhere groups of people gather: schools, workspaces, apartment buildings, prisons and more.
Most of these future tests will come up negative, and that fact can be exploited using an old and ingenious idea called pooled testing.
Pooled testing works by grouping samples together and then administering a single test to each group. The test is designed to be sensitive enough to come up positive whenever at least one sample is positive. Instead of testing each sample individually, you test each group, then only those groups that test positive undergo a second round of testing of each individual sample.
When testing a population where a positive test is rare, most of the groups will turn out negative which saves a large number of tests. This testing scheme was first proposed in 1943 for testing WWII soldiers for syphilis. It has been put to good use in many different applications over the years including blood bank testing and testing for the avian flu.
How much savings in resources does this produce? A lot. For testing the general public, the amount of savings will depend on the size of the group test we can implement. In a population where few people are infected, the pool size can be larger. Let’s say we want to test 1 million people. If we can implement a group test of size 10, we'll only need roughly 100,000 tests (10% of the original cost). If we can increase the group test to 100 per group, we'll need roughly 10,000 tests (1% of the original cost).
Israel’s Hadassah Medical Center is doing pooled testing (group test size 8) right now ( https://www.medrxiv.org/content/10.1101/2020.04.17.20069062v1 ) and their work gives a clear roadmap for developing good group tests along with logistical lab design to implement it in large automated testing labs. However, if we want to accomplish this quickly there are 3 essential roles for state governments to play:
1. To establish pooling as a clear priority and be actively involved coordinating the work.
2. To enable direct consumer-level testing on a large scale without involvement of the traditional healthcare system.
3. To use funding (public or private) to support establishment, setup and validation of pooled assays on a large scale.
Pooled testing does not solve all our testing problems: we'll still have to gather samples at an astronomical scale, and it won't reduce the demand for swabs. However, it will be an important piece of the puzzle.
We know we'll need a huge amount of testing to get through this. Pooled testing will help and provide critical information at a fraction of the cost and effort, ramping up testing capacity quickly and efficiently while using the available resources.