The quick math of India's second wave.

One silver lining of this massive humanitarian disaster is that there likely won't be a big third wave. But nothing is guaranteed: vaccination remains key.

~21.5% of Indians above the age of 10 (~250m out of ~1150m 10+ Indians out of ~1380m total) were infected by SARS-CoV-2 until 8 January 2021 according to the third round of the Indian Council of Medical Research’s serological survey.

To be conservative (because all seroprevalence data should be taken with a pinch of salt) assume that from 8 Jan to 31 Mar, zero new infections happened. Moreover, herd immunity kicks in at 60-70%, so let us be conservative again and choose 70%.

That leaves us with getting 800m (1.15B*.7) Indians above 10 who need to get immunity. The ICMR data shows we start with 250m ‘covered’ as of say 1 Apr. Moreover for the second wave let us assume April was the acceleration month, May flat, and June deceleration (even though the acceleration clearly started in March if not earlier and hence deceleration could also be longer, but again this will be a conservative assumption)

Since end-March we had around 12.5m reported cases but 250m actual ones as per ICMR, the multiplier was 20x. The positivity rate since then has been very high, so the multiplier should be higher now but let us again make a conservative assumption and stick with 20x. While we are currently recording 400k cases daily approximately, in April we averaged 200k+. Say overall we average ~250k for Q2CY21.

That means ~5m true infections daily (250k*20x). Just to triangulate since let us consider that 1 in 1000 infected Indians die as an order of magnitude (comparing 165k deaths till March end and say 250m “actual cases”). This is not because of our health system obviously, it is in many cases - especially recently - despite it and more due to our lower age, urbanisation, nutrition fulfilment rates - nothing to our credit per se.

But nonetheless it checks up with number of daily deaths and has space for some (say 1.5x-2x but not ~10x or so) underreporting of deaths. Again, all these estimates are rough and there are passionate, politicised debates with regards to this on social media.

Coming back to the main point for this post, say for Apr-Jun again: we average 3m vaccinations per day but due to some earlier infected people being double counted etc here, let us assume a very conservative daily addition of just 1m (because of vaccination) to uniquely counted immune individuals. Although over time, the contribution from vaccines is likely to increase and from infections is likely to decrease for immunity formation purposes.

So that gets us to (5+1=) 6m people getting immunity (not talking of immunity falling off in say 2022 for many indivdiuals, which may require booster or normal shots - and hence vaccination remains key). Over these 90 days, it takes us to around 540m people. 540m and 250m to begin with together takes us close to genuine herd immunity by the beginning of July 2021 though again vaccines are needed to seal the deal.

This back of the envelope analysis shows us that May 2021 is likely going to be the peak of the pain, and by July we could perhaps be back to February numbers. Just like nobody estimated such a massive second wave so suddenly, it is likely to go away equally suddenly. But this time, we cannot be complacent. We must get mass vaccination done as quickly as possible, mistakes so far notwithstanding.

Finally, for every family that loses a loved one - none of the above is remotely comforting. I have no words for even properly addressing that pain.