COVID-19: Can We Test Our Way Out Of This?
by Raywat Deonandan, PhD
Epidemiologist & Associate Professor
University of Ottawa
(I add my credentials to these COVID-19 blog posts in case they get shared. I want readers to know that my opinion is supposedly an educated and informed one)
Slovakia did an amazing thing this month. They tested two thirds of their population over a couple of weekends to find close to 58,000 active cases. The UK was so inspired by what they saw, that they did a similar thing in Liverpool, as well.
That got my spouse (a family doctor) and I talking this morning about whether mass testing could be a tolerable way out of the COVID mess for parts of Canada. So this blog post is a real-time thought experiment in which I take a stab at describing how such a strategy might work in some Canadian provinces.
First, why do we need this? Well, as we speak, parts of Canada, in particular Western Canada, are metaphorically on fire. Hospital capacity is being strained. Hallway medicine is a real thing. Death rates are climbing. And predictably, there are calls for another lockdown to slow the spread.
I described some of the options for slowing transmission in this post. They all come with extraordinary cost –in terms of business activity, loss of income, loss of employment, mental health concerns, and even the exacerbation of other health concerns, like delayed cancer screenings.
Now there is a national call in some circles for ZeroCovid, meaning the elimination strategy. If pursued, it could come at the price of staggering economic pain, with the payoff of a relatively open economy as we wait for the light of the vaccine.
The great confounder is that it’s difficult to predict when the vaccine will save us. Will it be a year or two? Or will it be a single season? While it now seems that we in Ontario will be receiving some doses by the March of next year, my crystal ball remains fuzzy as to when all of this will be over.
The lesser confounder is public will. What can the public tolerate? A painful hard lockdown, as experienced by Australia, might not be tenable in parts of Canada where anti-mask rallies are now the norm. Even a series of shorter “circuit-breaker” lockdowns will not be met with much enthusiasm.
Hence the need to explore some other options.
Why Mass Testing?
Imagine if you could quickly identify who is infected and infectious right now, and could quarantine them and just them. Everyone else could go about their lives. But that minority of people who are in fact the problem could be safely excised from society until they can recover?
Based on experiences in places like Slovakia and Liverpool, the fraction of the total population who is likely to be presently infectious is small…. very much under 10% and likely much smaller. That means it’s a very small portion of people who actually need to be restricted. If we can take them out of the equation, most of our problem is solved.
In such a scenario, everything could be kept open, and hospitals would have to endure the single wave of identified cases that go bad. The reproduction number (the average number of new cases produced by an old case) would drop to zero within one disease cycle.
If done perfectly, this could stop the epidemic in its tracks.
How Would It Look?
In this thought experiment, I am imagining a small, manageable but contained population. The provinces of Manitoba and Saskatchewan come to mind. The each have barely over a million people. They are both also geographically huge provinces, which would seem like a disadvantage for this strategy. But in their cases, land travel to and from adjacent provinces are via a handful of known highways that can be easily closed or monitored; same with the small number of airports and train stations. The US border is already mostly closed.
In my fever dream, these provinces each acquire about four million rapid antigen tests, the kind made cheaply by South Korea, and that render a result in under an hour.
As you might recall from this earlier post, many of these cheaper tests have poor sensitivity and/or specificity, rendering their positive and negative predictive values suspect. In this specific instance, the Asian rapid antigen test has an elevated rate of false negatives, and a low rate of false positives.
So they are not ideal for genuine diagnostic testing, like you would do in a hospital. For that, you would still need to do the expensive (and slow) PCR swab test.
Michael Minna has also been advocating for a testing-based solution, but pushing new at-home testing technology as the solution. I think he’s right.
So how do you control for the high likelihood of false positives and/or false negatives? This is how:
Step 1: acquire a buttload of these tests (4 million for the province of Manitoba, say)
Step 2: create a database for test results, linked to social insurance numbers or some other verifiable electronic identification system
Step 3: mail or courier them out to all residences in the province, and to shelters and prisons and military bases. Send two tests to each person.
Step 4: compel everyone to stay home for ONE WEEKEND, during which they take their tests. They are instructed to take the tests one day apart. Each test renders a result in under an hour.
Step 5: after each test, the individual will log their result onto an online portal, using their social insurance number
Step 6: anyone testing negative TWICE is issued a pass (electronically tagged to their SIN) that allows them to venture forth into society. Anyone testing positive TWICE is compelled to self-quarantine for 2 weeks. There will be follow-up from health officials, who now know about this case.
Step 7. Anyone with two different results from their test are visited by a mobile testing unit who apply a standard PCR test as the final determinant of infection status. The option also remains for anyone testing positive twice, as they will likely only be a few thousand.
Step 8: Anyone *refusing* to test or to enter their data will also be compelled to self-quarantine for 2 weeks.
Step 9: Repeat the process 3-4 weeks later to ensure that you got everyone.
Step 10: monitor the borders for infectious travelers, and double down on community surveillance to be able to detect new incident cases.
And there it is. Within one month, you have a COVID-free province.
The test quality is the first big problem. This is why I recommend serial (or orthogonal) testing. This diminishes the chances of an erroneous reading. False positives aren’t a problem; those people quarantine anyway. It’s the false negatives that pose a threat. By repeating the program 3-4 weeks later, we reduce the chance of false negatives re-seeding the community.
Public compliance is the biggest problem. Can we trust people to (a) self test, (b) enter their information accurately, and (c) self-quarantine if they need to? I don’t know that we can. So, as always, the success of any COVID program depends on the ability of the citizens to marshal their desires.
Of course, there is always the risk of missing people. The homeless, the nomadic, those who might be “off the grid”, and very remote communities might be particularly problematic.
I don’t know if this would work. But the fact that Slovakia has tried it, as has Wuhan and Liverpool, and the fact that luminaries such as Michael Minna are suggesting similar tactics tells me that this is worth thinking about.
The cost is comparatively miniscule (the price of 4 million tests, two weekends of no economic activity, and the price of human resources needed to monitor and enforce the process). The potential benefit is a completely open society and economy in a month…. without the need for school or business closures.
What do you think? Have I missed something obvious?