Is the U.S. Doing Enough Testing? No, We Need Much More

By Randall Bolten, data scientist, professor at UC Berkeley Extension, author of Painting with Numbers

Special to The Digest

No, we’re not doing enough testing – not if we look at the rest of the world.  We’ve all heard that testing is a measure of how effectively a country is responding to the COVID-19 pandemic.  It not only helps us identify cases for treatment, it also enables contact tracing to identify other cases and control the spread of the pandemic.  But how do we measure testing?  And which countries are doing well?  We’ll see that the U.S. does not score high.

It’s been said that the U.S. has performed more testing than the rest of the world combined.  Not so – while the U.S. had reported 4.7 million tests through 4/23, that’s only 19% of the 24.7 million tests reported worldwide.  And that worldwide total excludes both China, which provides no test information to the WHO, and Taiwan, which has been hailed for its strong response to the pandemic, but which has no relationship with the WHO.  And when you look at how hard the U.S. has been hit by the pandemic, the picture looks even worse:  that 19% of the world’s testing looks even smaller when compared to the fact that 33%, or nearly one-third, of the world’s reported COVID-19 cases are in the U.S.

It’s widely understood that:

(1) a significant amount of testing is driven by simply the number of cases, and

(2) additional testing beyond that is strategic and preventive. 

Comparing total cases to tests performed will give us a sense of how well a nation addresses both objectives.  To level the playing field, we must look at both metrics on a per-capita basis. The following graph compares 45 countries, which collectively have reported 90% of the world’s COVID-19 cases:

The bubble graph format enables us to add a third dimension of information: the country’s population.  Also, the “wealthy west” datapoints – western Europe, U.S., Canada, Australia, and New Zealand – are colored differently.  The wealth and sophisticated healthcare systems of that group of countries should put them at the forefront of using testing as a preventive measure, especially since many of them have been hit so hard by COVID-19. 

The upward-sloping red line added to the graph reflects the notion that some amount of testing is driven simply by caseload.  Countries right on that line are doing the minimum testing necessary to manage the existing cases.  Countries that sit well above that line are using testing strategically, to identify hard-to-spot cases and to manage the future spread of the pandemic.  Of the countries with more than 1,000 cases per 1MM population, only a few – Iran, U.K., France, Belgium, and Spain – sit right up against that red line.  Not only is each one of those countries among the countries hardest hit by the pandemic, they tend to have a significantly higher case fatality rate (CFR) – that is, total deaths divided by total cases – than the other countries shown (with the tragic exception of Italy). 

This conversation begs the question of how much testing is needed before the U.S. lockdown can be completely lifted.  Estimates range from 10 million per week – just to get the healthcare workforce secure – to as many as 30 million per day.  Even if those estimates are excessively pessimistic, at our current rate of just over 200,000 tests per day – we are a long way from where we need to be.  The only safe and reasonable approach is to lift restrictions gradually, as testing capacity increases and other measures are put in place. 

Any way you slice it, getting to adequate levels of testing will take a concerted, collective effort throughout the U.S. economy and at all levels of government.  Among other things, effective testing requires smooth operation throughout the “supply chain” – it’s not only a question of adequate testing supplies, it’s a question of people to administer the tests, distribution to deliver tests to labs, and lab capacity to turn lab results around promptly – after all, taking ten days to get a coronavirus test result back to a person is a little like taking six months to process a pregnancy test. 

All this data suggests that countries are addressing the testing challenge in wildly different ways.  The data also suggests that the countries having the hardest time with the pandemic tend to be the ones doing the least amount of testing.  More effective, better organized testing must be a top priority at all levels in the U.S. 

[A FINAL NOTE: Four of the datapoints – Iceland, Luxembourg, San Marino, and the UAE – don’t appear on the graph because they all reported more tests than 30,000 tests per MM – the highest value on the vertical axis scale.  Iceland deserves special mention because as of 4/23 it had performed 45,000 tests; with a population of 341,000, that’s an astonishing 132,000 tests per MM – by far the highest rate in the world.  Also, while Iceland’s 5,243 cases per MM is one of the highest per-capita figures in the world – which isn’t surprising when you’re testing so many people – it has reported only 10 fatalities, for a CFR of only 0.6%.]

Randall Bolten is a data scientist, professor at UC Berkeley Extension, a longtime Silicon Valley CFO, and author of the book Painting with Numbers