India’s surging ‘R’ number
The TLDR: The number of Covid cases are surging, and both the government and experts are warning of a second wave. Now all attention is focused on that dreaded ‘R’ or reproduction number—which often determines whether or not we go back into a lockdown. But what is this number, and why does it matter?
First, the numbers
The big tally: India added 46,951 new cases yesterday, bringing our total number of active infections to 11,646,081. We have added 131,750 over the space of just 72 hours! Maharashtra—which logged 24,645 new cases—accounts for over 60%. And six states account for a whopping 84.5%. These include Maharashtra, Punjab, Kerala, Karnataka, Gujarat and Madhya Pradesh. Among the cities, the big worry is Mumbai which reported the highest daily spike of 3,775—and 316 buildings in the city have now been sealed. Pune with 2,978 cases and Nagpur (2,747) are not far behind.
A second wave? While the word is often used to describe any sudden spike, ‘wave’ has a specific meaning in epidemiology. The WHO executive director Mike Ryan defines it so: “There’s a period of time in which there’s very low or no activity, and then the disease returns in a large way.” So the great flu pandemic of 1918 had three distinct waves and looked like this in the UK:
India’s pattern: is fairly similar. We were adding a stratospheric 97,800 cases in mid September. The numbers went steadily down until we hit a low of 9,110 on February 9. Soon after, we started to climb again—hitting 14,264 on February 21, then 17,407 cases on March 4 and finally 46,951 cases on Monday.
Point to note: Maharashtra is especially worrying. The reason: In the last wave, it was only adding around 23,000 cases a day. Now that number is 30K-plus.
What’s with that R number?
The R-nought: R in itself just refers to the reproduction rate of a virus. If one person has Covid, say, how many people will she infect? One version of this number is R0 (Or R-nought). This assumes that everyone in a given population is equally vulnerable to infection—and there are no attempts to stop its spread, i.e. there is no social distancing or vaccinations etc. So the R-nought for this coronavirus is 3, while the number for measles is 15. Typically, experts arrive at this number after the fact—and early in the pandemic. They work back from the data available:
“[D]isease modellers look at current and previous numbers of cases and deaths, make some assumptions to find infection numbers that could have explained the trend and then derive R from these.”
Here’s a chart showing R0 numbers for different diseases:
The real Rt: This is the effective rate of spread, and calculated as the virus spreads. This requires figuring out how many people may already have immunity—either because they’ve already survived the disease or been vaccinated. Often, when health authorities speak of the R number, this is the one they are referencing. And if you have lots of social distancing or widespread vaccinations, there can be a wide variance between Rt and R0.
Point to note: When it comes to Rt, any number above 1—where one person only infects one other person or less—is dangerous. And it becomes a key factor in deciding whether or not to impose a lockdown. Below 1, the virus eventually sputters out, running out of enough people to infect. Above 1, it starts to gather momentum, and the cases grow exponentially. Here’s BBC’s chart on how different Rt numbers produce different case loads.
And what’s our Rt number?
As of March 22, it stands at 1.32—the highest since April last year, when there were less than 27,000 cases. And it dipped below 1 in November, and stayed there until late February—which is when the second wave started to build. The big point to note: “The top 16 states with the highest number of active Covid-19 cases, with the exception of Kerala, have Rt over 1.” As of March 19, among the states, Chhattisgarh had the highest Rt number of 1.44—while Maharashtra was 1.29. At the time, the India number was 1.19.
So it’s Maharashtra not India, really!
In a sense, yes. The Rt number offers a big picture that can often be misleading.
One: It can’t capture how one region or even a city or locality within that region single-handedly drives that overall figure. So our 1.39 number more accurately reflects the situation in one state rather than the country. As The Print notes, “A comparison between the R values of Maharashtra with that of the rest of the country shows that India’s R value roughly mirrors the peaks and troughs of the Maharashtra R graph.”
Two: It’s a ‘lagging indicator’ since it relies on cases reported today—which only capture infections that occurred anywhere between 1-3 weeks ago. It can’t capture how the disease is spreading right now. As one expert notes: “If you have your Rt estimate lagging by at least ten days, possibly two weeks, then it’s not going to be that useful as a real-time decision-making tool”—especially when it comes to decisions like imposing a lockdown.
Three: Over the course of one year of Covid, experts have learned that this virus doesn’t spread in a uniform way. One person or event can be a super-spreader, while others don’t pass on the virus at all. This is why they are now looking at a new number called ‘K’—which captures ‘dispersion’. During the 1918 flu epidemic, the virus had a K value of around 1—which meant around 40% of those infected didn’t pass on the disease. But for this coronavirus, that number is 0.1—i.e. 70% of those infected don’t infect others.
So it becomes hard to predict the progression of a disease. As The Atlantic notes,
“This highly skewed, imbalanced distribution means that an early run of bad luck with a few super-spreading events, or clusters, can produce dramatically different outcomes even for otherwise similar countries.”
And it may explain what we’re seeing right now in Maharashtra—as opposed to other parts of the country.
And how does the K number help us?
It offers a different way to tackle the pandemic—or an urgent situation like Maharashtra. Here are two possible strategies:
One: Minimise the risk of super-spreader events—all of which have the same characteristics: prolonged contact, poor ventilation, a highly infectious person, and crowding. So rather than a total lockdown, we may look at shutting down theatres, local trains etc.
Two: Change how we do contact tracing. Right now, when a person tests positive, we look ‘forward’ to figure out who else they infected. But in a highly variable disease like Covid, that number may actually be small in most cases. The trail will inevitably peter out. So some experts advocate ‘backward’ tracing—figure who infected the person who tested positive. This makes it much more likely that we will identify clusters and superspreading events and people.
The bottomline: Today marks the one year anniversary of the janata lockdown—when we shut the entire country down to stop the pandemic in its tracks. Clearly, that strategy has not worked. Nor has opening everything, and throwing caution to the winds. Maybe it is time to deploy a new set of more finely-focused measures that embrace the reality of this virus.
Reading list
NDTV has the latest on our R number. The Print has a good set of graphs on R numbers for India and Maharashtra. Nature does an excellent job of explaining the R number and its limitations. The Atlantic and The Conversation dive into the K number and why it matters.