Categories
Covid-times

India Covid deaths weekly projection – 23 Aug update

My big question every week (since May) is, when will India cross 1 lakh total reported Covid deaths? Total cumulative Covid death toll as of yesterday stands at almost 56k (actual figure could be as high as twice this value, for various reasons documented here).

On an average 965 people reportedly died per day this week (compared to 931 last week). That’s five fully loaded A320 airplanes crashing and killing every passenger every day.

One could try forecasting the future deaths by simply using the existing growth rate trend for weekly Covid deaths.

Yes there are many ups and downs as we see above. In fact in earlier weeks, I struggled to spot any trend. But the graph has now started to look somewhat predictable – the weekly deaths seem to be growing at somewhere around 10-15% week-on-week rate. Don’t they?

So I ran three week-on-week growth rate scenarios to see how the cumulative total would look like, in the future.

We are at almost 56k deaths (as of yesterday – 22 AUG). We will cross 1 lakh in another 4 to 5 weeks (most like by 20-26 Sep 2020).

Alright, let’s now try a slightly more nuanced (albeit indirect) approach to project cumulative deaths. Here we will first look at the week-on-week growth rate of total weekly cases (instead of weekly deaths).

4.5 lakh total positive cases were detected this week, which is only 3% higher than the total cases detected the week before (4.4 lakh).

I think this week is an exception. Something is off about “3%” growth rate – unless India has started to peak. Yes we know, some cities are on the verge of peaking.

But India is a big country and for many parts, the growth has started much late, so 3% overall national growth rate in unlikely to repeat (at least this year). Anyway, we shall get more clarity on this in the coming weeks but for now let me assume the w-o-w cases growth rate will bounce back.

For my projection, I will assume a range from 15 to 25% (X). That’s my assumption #1.

Now in general, people who die of Covid in a given week, are either tested positive the same week, or the week before. Do we have some idea of what %age (Y) of such cases die? We do actually.

As you see above, death %age (Y) has been consistently around 1.6% since past few weeks.

6,750+ people died this week (16-22 Aug) which is 1.5% of half of total cases from this week + half of total cases from last week.

In other words, Y for this week is 1.5%. For the future, let’s assume a range from 1.3% to 1.5%? That’s my assumption #2.

Let’s forecast now…

Let me consider 3 scenarios:

  • X=15%, Y=1.5% (baseline)
  • X=15%, Y=1.3% (optimistic: lesser %ge of deaths)
  • X=25%, Y=1.5% (worse: expecting faster growth in cases)

With the above assumptions, below chart represents the future cumulative death count:

The result is more or less the same as that from the first direct method I used.

India will cross 1 lakh total deaths by Sep-end.

Now, 1 lakh total deaths for India is basically equivalent to 72 deaths per million of the total population (currently we are a little over 40).

To what extent would the death toll figures keep going up – before it flattens / peaks?

If we look at other countries, death toll for many started to flatten out only after anywhere between 400 to 600 per million of their population died!! Scary, I know!

Y axis = no. of days (all the countries are arranged in a way that starting point of 10 deaths per million is common to all)

If we assume that for India, the death toll flattens out even at say 200 deaths per million, that would be equivalent to ~3 lakh total deaths!

It’s difficult to imagine why India would see any less deaths than that. The only populous countries across the globe where death toll flattened at much lower levels (like say Japan and China) happened when they somehow didn’t let the total deaths cross even 5k (Japan for example didn’t even let it cross 1k). We clearly couldn’t control things to that extent in India (most countries haven’t). So now let’s just be hopeful that the total death cap estimate that I am guessing is on the conservative end – otherwise, we could lose even upto 5 lakh people (or 362 deaths per million)!

That’s it for this post. I’ll get back with updated projections next Sunday (30 Aug). Stay safe.

Categories
Covid-times

India Covid deaths weekly projection – 16 Aug update

Every week (since May this year), I look at the existing trend of a) growth in reported Covid deaths and b) growth in the positive cases for India. And then I run few scenarios to forecast when India will cross 1 lakh total reported Covid deaths.

Earlier I was also interested in looking at when the avg daily deaths in a week would cross the 1,000 mark. Back in first week of June, the avg daily death figure was less than 200. I stopped doing that estimate once we reached 700 avg daily deaths.

My video from first week of June 2020

In the video above (will start playing from 06:20), you will see that my “most randomly theoretical conservative” timeline projection for when India would start seeing 1,000 avg daily deaths was the 3rd Sunday of August. That’s today, is it?

Almost 1,000 died yesterday (15 Aug 2020) and at least on two occasions this week, the total daily toll was over 1k.

It’s a good thing that the growth has taken the conservative path (of all my assumptions). Anyway so where do we head from here?

Total cumulative Covid death toll as of yesterday stands at almost 50k (actual figure could be around 1.5 times this value, for various reasons documented here). Remember that a month ago, we were just half of this number (with around 26k dead). So basically total death toll is sort of doubling in a month.

On an average 931 people reportedly died per day this week (compared to 858 last week). That’s more than four fully loaded A320 airplanes crashing and killing every passenger every day.

One could try forecasting the future cumulative deaths by simply using the existing growth rate trend for weekly Covid deaths. There’s just one problem.

The growth trend is not so obvious.

For example as you can see in the above chart, total weekly deaths this week (6500+) was 9% higher than total weekly deaths the week prior to that (6000+). But the weekly growth last week was a much higher 17% and the week before that, an abysmally low 1%.

If one *must* project future Covid deaths just from the above chart, what growth-rate range would you consider? Let me consider three scenarios: week on week deaths grows by 8%, 12% and 16%. Below is how the projection for total cumulative deaths looks like if we extrapolate.

We are at 50k deaths today – we will cross 1 lakh in another 4 to 6 weeks.

Alright, let’s now try a more nuanced (albeit indirect) approach to project cumulative deaths.

Let’s first look at the week-on-week growth rate of total weekly cases (instead of weekly deaths).

4.4 lakh total positive cases were detected this week, which is 11% higher than the total cases detected the week before (3.9 lakh).

Based on the pattern so far, I expect the total weekly cases in the coming weeks to grow by anywhere between 15% to 25% (X). That’s my assumption #1.

Now in general, people who die of Covid in a given week, are either tested positive the same week, or the week before. Do we have some idea of what %age (Y) of such cases die? We do actually.

As you see above, death %age (Y) has been consistently falling since many weeks now and seems to have flattened at 1.6%.

6,500+ people died this week (09-15 Aug) which is 1.6% of half of total cases from this week + half of total cases from last week.

In other words, Y for this week is 1.6% (it was the same last week too, and even the week before).

Let’s assume a range for death %age (Y) from 1.4% to 1.8%? That’s my assumption #2.

Let’s forecast now…

Let me consider 3 scenarios:

  • X=20%, Y=1.6% (baseline)
  • X=15%, Y=1.4% (optimistic: expecting even lesser %ge of deaths)
  • X=25%, Y=1.8% (worse: expecting faster growth in cases AND even higher percentage of deaths)

With the above assumptions, below is how the future cumulative death count looks like:

The projection is pretty much the same as the first direct method I used.

India will cross 1 lakh total deaths by middle of September to Sep-end.

Now, 1 lakh total deaths for India is basically equivalent to 72 deaths per million of the total population.

To what extent would the death toll figures keep going up?

If we look at other countries, death toll for many started to flatten out only after anywhere between 400 to 600 per million of their population died!! Scary, I know!

Y axis = no. of days (all the countries are arranged in a way that starting point of 10 deaths per million is common to all)

If we assume that for India, the death toll flattens out even at say 200 deaths per million, that would be equivalent to ~3 lakh total deaths!

It’s difficult to imagine why India would see any less deaths than that. The only populous countries across the globe where death toll flattened at much lower levels (like say Japan and China) happened when they somehow didn’t let the total deaths cross even 5k (Japan for example didn’t even let it cross 1k). We clearly couldn’t control things to that extent in India (most countries haven’t). So now let’s just be hopeful that the total death cap estimate that I am guessing is on the conservative end – otherwise, we could lose even upto 5 lakh people (or 362 deaths per million)!

That’s it for this post. I’ll get back with updated projections next Sunday (23 Aug). Stay safe.

Categories
Covid-times

India Covid deaths weekly projection – 09 Aug update

Every week (since May this year), I look at the existing trend of a) growth in reported Covid deaths and b) growth in the positive cases for India. And then I run few scenarios to forecast when India will cross 1 lakh total reported Covid deaths. Total death toll as of yesterday (08 Aug 2020) stands at 42k+ (actual figure could easily be two times this value, for various reasons documented here).

On an average 858 people reportedly died per day this week (compared to 736 last week).

One could try forecasting the future cumulative deaths by simply using the existing growth rate trend for Covid deaths. There’s just one problem.

There is no existing growth rate trend for Covid deaths.

For example as you can see in the above chart, total weekly deaths this week (6000+) was a little over 15% higher than total weekly deaths the week prior to that (5100+). But the same thing last week was less than 5%!!

If one *must* project future Covid deaths just from the above chart, what growth-rate range would you consider? I would go by a range of anywhere from 5% to 15% week-on-week growth in total weekly deaths. If we do that, then below is how the projection for total cumulative deaths looks like.

So as per this simplistic projection, total death count will cross 1 lakh by anywhere between August-end to Sep-end (again, just to remind you, as of now the total Covid death toll for India stands at around 42k).

Alright, let’s now try a more nuanced (albeit indirect) approach to project cumulative deaths.

Let’s first look at the week-on-week growth rate of total weekly cases (instead of weekly deaths).

3.9 lakh total positive cases were detected this week, which is 9% higher than the total cases detected the week before (3.6 lakh).

Although for this week and the week before that, weekly total cases have grown by less than 20% week-on-week, in general this has hovered between 20 to 30% – as you can see in the chart above.

So based on the pattern so far, I expect the total weekly cases in the coming weeks to grow by anywhere between 20% to 30% (X). That’s my assumption #1.

Now in general, people who die of Covid in a given week, are either tested positive the same week, or the week before. Do we have some idea of what %age (Y) of such cases die? We do actually.

As you see above, death %age (Y) has been falling since its peak of 3.4% in the 07-13 June week.

6,000+ people died this week (02-08 Aug) which is 1.6% of half of total cases from this week + half of total cases from last week.

In other words, Y for this week is 1.6% (it was the same last week too).

Let’s assume that for the coming weeks, death %age (Y) could get as low as 1% and if something random happens, could say go up to a max of 2%. That’s my assumption #2.

Let’s forecast now…

Let me consider 3 scenarios (which are actually same as last week):

  • X=20%, Y=1.5% (baseline)
  • X=20%, Y=1% (optimistic: expecting even lesser %ge of deaths)
  • X=30%, Y-2% (worse: expecting faster growth in cases AND even higher percentage of deaths)

With the above assumptions, below is how the future cumulative death count looks like:

The projection is pretty much the same as the first direct method I used. We will easily cross 50k before Independence day and then…

India will cross 1 lakh total deaths by end of Aug to latest by end of September.

Now, 1 lakh total deaths for India is basically equivalent to 72 deaths per million of the total population (in comparison, the present death toll of 42k+ is equivalent to ~31 deaths per million).

To what extent would the death toll figures keep going up?

If we look at other countries, death toll for many started to flatten out only after anywhere between 400 to 600 per million of their population died!! Scary, I know!

Y axis = no. of days (all the countries are arranged in a way that starting point of 10 deaths per million is common to all)

If we assume that for India, the death toll flattens out even at say 200 deaths per million, that would be equivalent to ~3 lakh total deaths!

It’s difficult to imagine why India would see any less deaths than that. The only populous countries across the globe where death toll flattened at much lower levels (like say Japan and China) happened when they somehow didn’t let the total deaths cross even 5k (Japan for example didn’t even let it cross 1k). We clearly couldn’t control things to that extent in India (most countries haven’t). So now let’s just be hopeful that the total death cap estimate that I am guessing is on the conservative end – otherwise, we could lose even upto 5 lakh people (or 362 deaths per million)!

That’s it for this post. I’ll get back with updated projections next Sunday (16 Aug). Stay safe.

Categories
Covid-times

India Covid deaths weekly projection – 02 Aug update

Every week (since May this year), I look at the existing trend of a) reported Covid deaths and b) the positive cases for India. And then I run few scenarios to forecast when India will cross 1 lakh total reported Covid deaths. Total death toll as of yesterday stands at 36k+ (actual figure could easily be two times this value, for various reasons documented here). On an average 736 people reportedly died per day this week (compared to 726 last week).

One could try forecasting the future cumulative deaths by using the existing growth rate of deaths. There’s just one problem.

There is no existing growth rate trend for death figures (see the red line below – it’s all over the place).

This week, total deaths figure grew by just a little over 1%, while last week it had grown by over 20%!

So how do we project the future deaths now?

Notice the blue line in the same chart above? Even when it fluctuates, it still has some pattern. This blue line is the trend for growth rate in positive cases (let’s call it X).

3.6 lakh total positive cases were detected this week, which is 20% higher than the total cases detected the week before (almost 3 lakh). In other words, X for this week (26Jul-01 Aug) is 20%

Based on the pattern so far, I expect the total cases in the coming weeks to grow by anywhere between 20% to 30% (X). That’s my assumption #1.

Now in general, people who die of Covid in a given week, are either tested positive the same week, or the week before. Do we have some idea of what %age (Y) of such cases die? We do actually.

As you see above, death %age (Y) has been falling since its peak of 3.4% in the 07-13 June week. 5100+ people died this week (26 Jul-01 Aug) which is 1.6% of half of total cases from this week + half of total cases from last week. In other words, Y for this week is 1.6%.

Let’s assume that for the coming weeks, death %age (Y) could get as low as 1% and if something random happens, could say go up to a max of 2%. That’s my assumption #2.

Let’s forecast now…

Let me consider 3 scenarios:

  1. X=20%, Y=1.5% (baseline)
  2. X=20%, Y=1% (optimistic: expecting even lesser %ge of deaths)
  3. X=30%, Y-2% (worse: expecting faster growth in cases AND even higher percentage of deaths)

With the above assumptions, below is how the future total death count looks like:

As you can see above, we will cross 1 lakh deaths sometime in September, whichever scenario we consider.

Now, 1 lakh total deaths for India is basically equivalent to 72 deaths per million of the total population (in comparison, the present death toll of 36k+ is equivalent to 25 deaths per million).

To what extent would the death toll figures keep going up?

If we look at other countries, death toll for many started to flatten out only after anywhere between 400 to 600 per million of their population died!! Scary, I know!

Y axis = no. of days (all the countries are arranged in a way that starting point of 10 deaths per million is common to all)

If we assume that for India, the death toll flattens out even at say 200 deaths per million, that would be equivalent to ~3 lakh total deaths!

It’s difficult to imagine why India would see any less deaths than that. The only populous countries across the globe where death toll flattened at much lower levels (like say Japan and China) happened when they somehow didn’t let the total deaths cross even 5k (Japan for example didn’t even let it cross 1k). We clearly couldn’t control things to that extent in India (most countries haven’t). So now let’s just be hopeful that the total death cap estimate that I am guessing is on the conservative end – otherwise, we could lose even upto 5 lakh people (or 362 deaths per million)!

That’s it for this post. I’ll get back with updated projections next Sunday (09 Aug). Stay safe.