India’s R value coming down: Sitabhra Sinha

Sitabhra Sinha, lead researcher at the Institute of Mathematical Sciences, Chennai, tells Anubhuti Vishnoi that while the decrease in the Covid ‘R’ count in the first wave was linked to the national lockdown, there are many factors at play in the current phase which has lowered the R count. He rules out any systematic attempts to fudge numbers. TN and Rajasthan continue to be a worry. Excerpts:

On India’s R value decline
India’s R value is now close to 1 which technically means that our active cases are no longer increasing at the national level. This is primarily due to improvement in the situation in states leading the tally so far, in terms of active cases. Until last week, Maharashtra was at the highest but the number of active cases here have been declining for a fairly long time and this rate of decline has increased further last week. So, it is not surprising that the national R level will also reflect that.


Karnataka is still leading in active cases as of this week but it has also shown improvement as its R value was 1.25 in the first week of May and is down to 1.04 in the last few days. Both these have contributed to India’s R value decreasing. Several other states which were biggest contributors are also below 1 –– UP is at 0.79, Gujarat has been less than 1 since May 6. Chhattisgarh has been a bit erratic but last week, it finally went below 1. A similar trend was noted in Madhya Pradesh too. Haryana went below 1 on May 9, Bihar on May 8, Delhi has been there since April 30, Telangana from May 3.

On states of concern
Some states have bucked the trend. Two states are showing an increase. Rajasthan was close to 1 on May 6 and has now risen to 1.07. If it continues to grow, it could soon be among the top-affected states. The biggest worry is Tamil Nadu. It was at an R of 1.12 in the first week of May again and close to 1.3 in the second week of May. This is also reflected in the city-wise data. Chennai is at 1.7 — the highest among metros and rising from 1.08 in the first week of May.

On spread in rural areas
The R value is calculated for the state as a whole and it is almost dominated by urban areas because the highest number of active cases are reported there. When I look at district-wise data, the calculation is difficult as there are too many fluctuations given the smaller numbers of cases. For instance, I did an analysis for Odisha districts and it was easier to assess the R value of Cuttack with its 14,000 active cases. However, at Kandhamal district, we are noticing a textbook example of exponential growth but it was not possible to get an R value for it, given the small numbers.

There are testing constraints in rural districts that prevent estimation of growth rate. One day I may detect a high R value and the next day it suddenly flattens out. This is most likely because testing capacity is saturated in the district and hence there is no increase seen in active cases.

On alleged underreporting
We are not looking at Case Fatality Rate (CFR) closely due to the difficulties in statistical estimation at the district-level as mentioned above and given that testing data is collated at the state level, not at the district level. Typically, CFR should be roughly a constant fraction of total active cases but currently there is a lot of noise around it. It is hard to pin point whether the decline in R is because of underreporting of mortality data or some other cause which may not be deliberate at all. What one can try to answer is that whether systematic attempts are there at deception, or to bring down the R value by fudging testing numbers. I don’t think there has been any attempt at that. It seems quite unlikely. For example, in several instances where a state’s R value comes down, we noticed that there is a simultaneous slight increase in testing or the testing numbers are constant.

First wave to Second wave
The first wave settled by September end. Then, there was a relative lull up to about January- February, with the exception of an outbreak in Kerala. India remained quiet till February, after which the second wave spread very quickly. Several states had an R level higher than 1 in just 2-3 weeks. While it is difficult to pinpoint reasons for the spread and decline of the virus as there are so many factors at play, the decrease in R count in the first wave was linked to the lockdown. So, from an R of 1.92 around March 14-April 5, India was at 1.3 by mid-April amid the 2020 lockdown. This time around, it is difficult to pinpoint the exact reason for the R value decline as yet because there is no single national lockdown preventing people from travelling.