Covid positivity: a heavy load, does it explode?6 min read . Updated: 04 Sep 2020, 07:32 AM IST
If positivity rate is rising, the virus may be spreading faster than the case count suggests
The largest daily spike in the world, the largest daily spike ever, uncounted cases might be 6 or 10 or 20 or 200 (!) times official counts, an economy in serious reverse gear … it’s hard to keep abreast of all the news about all that this pandemic is doing to India. I keep my nose to the data as much as I can. But it’s hard. So many questions, worries, puzzles. One of those is about testing for the virus. Through these pandemic months, we’ve heard plenty about testing. Specifically, that India needs to be testing more. But today, India has tested over 43 million people —close to the population of Spain — and we are testing over a million people a day.
Those are large numbers. In fact, going by the number of tests conducted, we are third in the world — only China (90 million, though there are questions about this number) and the US (83 million) have tested more people for the covid infection. The US is now testing about 670,000 people daily, down from a peak of about 900,000 a month ago. That’s substantially less than India’s count of a million, and no other country comes anywhere close to our daily testing tally.
Yet when we look at the numbers like this, we have to also factor in India’s population. We’re testing more than most other countries, but we also have more people than most other countries. There are nearly 1.4 billion Indians, four times the number of Americans and second, of course, only to the Chinese. So while we have indeed tested a large number of Indians, that 43 million count doesn’t seem quite as large any more when we compare it to our population. Of every million Indians, we have tested over 31,000, or 3.1%. That is ,we have tested one of every 30 Indians.
Around the world, those are not bad numbers. But of the world’s 10 most populous countries, China, the US, Brazil and Russia are all conducting far more tests per million population than India is. China is at 63,000, Brazil at 67,000, the US and Russia at about 250,000. (Indonesia, Pakistan, Nigeria, Bangladesh and Mexico are the others in that top-10, all with testing rates far below India’s, though their covid outbreaks are nowhere near as severe as India’s). But even with our testing rate so low, India has the world’s third-highest count of cases, and is on track to overtake Brazil to second-highest in a couple of weeks. What will our case numbers be if we ramp testing up to the level of Brazil, let alone the US and Russia? How many Indians will we find who are infected? In fact, how many Indians are infected?
Such questions are why many people who pay attention to these things think that India has far more infections than the official figure, which is now nearing 4 million. How many more? In a conversation last weekend with Karan Thapar, the Harvard public health expert Ashish Jha suggested it could be ten or twenty times as many; others have speculated the factor could be between 40 and 200 (!). 200 times the official count would mean 800 million Indians have been infected, or nearly 60% of us. Is that possible? Plausible? I honestly don’t know. Yet it is sobering to think that there are informed estimates that range that high.
Consider too that what we perhaps should be looking at more closely is not really how many tests we perform, but what the tests tell us. Of course, this is important because positive results identify infected persons who can then be quarantined or treated as necessary. But we should also ask, what fraction of the tests return a positive result? Or, what is the positivity rate of the testing?
We must ask, because the positivity rate is vital to understanding the pandemic, and in two ways.
The first is that it tells us whether that country (or city, or region, whatever) is testing adequately, given how large the outbreak of the pandemic is there. It’s testing that gives us an idea of how widespread the outbreak is, and that in turn lets us plan how best to control its further spread. Without doubt, countries in which the virus has spread more widely need to be testing even more widely. This is why the variation we find in countries’ positivity rate is telling. Leave aside countries — usually small ones — which have relatively low case counts. What about those that the virus has slapped hard?
South Korea has tested nearly 2 million people and has had about 20,000 infections: a positivity rate of 1%. Serbia has conducted nearly a million tests and found 31,000 infections: a little over 3%. El Salvador? 315,000 tests, 26,000 infections, 8.25%. Moldova? 210,000, 37,000, 17.6%. Ecuador? 330,000, 114,000, 34.5%. And India? 43 million, 3.7 million, 8.6%.
Some of these are still small countries, but all have had significant encounters with coronavirus: case counts in five and six digits tell that story. But of these, the positivity rate tells us that South Korea and Serbia have probably done a reasonable job of testing—and then, of controlling the spread of the virus. In fact, epidemiologists use 5% as a benchmark: bring the positivity rate below that, and the pandemic is probably under control.
Which brings us to the second way the positivity rate is vital. When looked at together with figures about those infected, it helps us understand how the virus is spreading in that country (or city, etc). In South Korea and Serbia, it takes several dozen tests to find a single new case. This suggests that those countries have managed to slow the spread of the virus to a trickle. But in Ecuador, every third test is positive; in Moldova, every sixth; in India, every 11th. In countries like these that have high positivity rates, it’s so easy to find new infections that their covid reality is a large pool of people whose infections are still to be detected. The official case count is probably just a fraction of the true number of infections. Put another way, a sure indicator of how far the virus has spread in a country is the ease of finding a new infection, seen purely in terms of the number of tests that search needs. More alarming still, if the positivity rate is rising, then it’s likely that the virus is spreading faster than the case count graph suggests.
This is the context in which to think about India’s numbers and our response to the virus.
For example, in mid-August a news report published by The Print on 20 August told us that “Maharashtra’s positivity rate increased to 19.91% from 15% in June". Now the same report said that testing in Maharashtra had increased seven-fold —that’s right—since June. Still, there was this significant increase in the positivity rate too. The inference: the virus is spreading faster in Maharashtra than we can monitor it, outstripping even the pace of testing. Another report published by Hindustan Times on 7 August discussed Mumbai’s positivity rate, which came “down to 20.83%, from 21.68% in July." Mark the decrease, fair enough, but the issue here is that at 20+%, that positivity rate is troublingly high. Like with Ecuador and Moldova, it tells a story of a virus that’s spreading quickly through the city.
As ever with statistics, there are multiple ways of looking at the data. Especially during a pandemic that has thrown economies around the world into disarray — famously, a 23% contraction in India — guys like me can’t help examining the numbers for reassurance, possible trouble, or more.
We are adding record numbers of new cases nearly every day; but in percentage terms, that daily growth rate has been declining for months. Yet our high positivity rate is the smoking gun here. What will it do in the months ahead?
I’m reminded of Langston Hughes’ famous poem “Harlem", in which he asks:
What happens to a dream deferred?
Does it dry up, like a raisin in the sun?
Maybe it just sags like a heavy load.
Or does it explode?
Once a computer scientist, Dilip D’Souza now lives in Mumbai and writes for his dinners. His Twitter handle is @DeathEndsFun