End of March 26th "Analysis" by Rev. Jayakrishnatennisepidomiologiananda
[who acts like he knows the sport and the science]
Never in my wildest imagination did I expect us to show only 720 cases by now. Nine days ago, I was praying for 750 in 7 days and expected about 200-300 cases coming in by day for a total of 2000 by now. Somebody (God?) was hearing our prayers so far.
But it is way too early to say anything, even if the Indian curve has certainly flattened a little bit in the last 3 days. From a total ~120 on March 16th to ~475 over 7 days, we kept a 22% rate. At that rate we would've expected about 875 by now (around where Russia, who were closely tracking us a bit behind during the past two weeks has gone now, going past us yesterday). But our increase in the last three days has been at around only 13.5% per day.
But this is not a "final flattening" on the other hand, and it is definitely NOT showing the effect of the lockdown or anything. This is most probably showing the effect of the tracing and quarantining of the foreign-returnees and their contacts. The foreign returnee numbers were going up at essentially the steeply rising "foreign curve" (perhaps an avg of about 25% per day, from our "catchment countries") from March 15 to 22nd. The effect of their ability or inability to infect others would not have gone into our curve till about 3-4 days ago. So the arrival curve started getting affected by a "number of contacts" curve that is added to it, and that curve started rising seriously in the last 4-5 days but it actually dropped the total curve because of a much lower growth rate (may be 10%).
So the drop from 22 percent to 13.5 percent in the last three days is quite probably because the foreign-returnees have only been able to infect an average of 1.3 or so others in 4 days, as opposed to 2+ that is sustaining the 20-40% (but average 25% or so) rate in our catchment countries that have had no trace-test scheme like us. But now, a week later (the delay being due to infection time and a further 2-3 day delay to get the contacts' test results), the second curve is beginning to show its effect. The combination of the foreign arrival's increase rate and the contacts number is now showing a drop from 22% to 13.5%.
Now, if we wait 4-5 days, I think we will start seeing an increase due to the people that the contacts may have infected. A third curve. And many of those are community transmission cases (otherwise they would've been traced and gone into the total sooner). So, if the 13.5 remains for 2-3 more days, we will then start seeing a slightly higher rate. Maybe a lot higher, but probably not more than 20% if I were to guess. Then there is the effect of the lockdown. That is a 4th curve, you may call it, but it really is a drop-in-the-rate factor. That will kick in around the same time or a couple of days later and the rate may drastically drop to 10% or so, if all goes well.
But the 3rd curve could drastically go up if the Pune/Bombay hot spots (or Bhilwara or Kasarkot) out there have caused huge community transmission already. We would need some divine intervention for that not to have happened. Maybe we have got it. I hope so.
So there are some 4 different curves that are adding up at various time lags. But rates of the last couple of curves that started 5-8 days ago and 2-3 days ago are very hard to determine because that is all community transmission. We should be around 2000 (+/- 500) in another week. We need to see the curve after 7 days to get an idea where we are going after that. That will be the "final rate" we will see after a week or so when all effects are added together.
At that point, it will depend on the efficacy of wider testing for community transmission that we do from NOW. If the test numbers go up quickly now, the data will be more reliable and we can catch-and-quarantine a lot more to keep the rate at an overall growth of 15 percent or lower. If we screw up on testing, we will simply get the 25-35 percent growth rate seen elsewhere, after a delay of 15-20 days. Then, that delay will be all we would've accomplished, with all our hard work so far. And it will all go for a naught. We will be totally out of control of the situation even in 10-12 days if we screw up and find a 30% growth rate in 10 to 14 days. In even 10 days, it could all be out of control. I am doing mental simulation here without the "IndiaSim" type simulators that our favorite Dr. Bloviator mentions (which seems to be garbage-in-garbage-out for the context-specific factors that have affected things so far)
That means, if we are testing enough in the right places (no need for huge nationwide tests just yet) and cordoning off areas for contact-tracing, sanitizing etc, we can keep this from starting to grow from a very comfortable 2000 in another week, to a manageable 14K in 3 weeks (a 15% growth rate after a week, that is). On the other hand, if we do not test enough and miss the many community transmission cases that may be hitting the hospitals already, then we will grow after a week at a rate of at least 20% to 25K in three weeks or at a more probable rate of 30% to 80K in 3 weeks with everything out of control. All the strategizing will be of no use at all. Then we are looking at the Bloviator's numbers - some 50K growing to 1 million cases by early May. That is an optimistic figure. Easily 75K deaths in even 5 weeks.
We had better get our act together and test people like our lives depended on it (well it does) this week. Right now!!!