COVID-19, Doubling of Cases In India – Analysis 2

Here is a followup article on my previous post of doubling of cases in India along with a new forecast for the next 14 days.

In my previous post about doubling of COVID-19 cases in India I had presented a rather optimistic picture of the situation. Though the situation has not deteriorated 2 weeks later, the situation has not gotten any better, if anything it has gotten slightly worse with respect to the number of cases in India. However, as noted in my previous fatality rate article, you need not worry too much, at least not statistically. However, the fact remains, that COVID-19 is a dangerous virus for the aged and anyone with underline health concerns. And this is an undisputed fact that the 368,944 people who have lost their lives to this infections virus till date, would have been alive today if not for COVID-19. So when you hear people compare the death rate of COVID-19 to that of road accidents, do not let their illogical comparisons influence your understanding of this virus. Road accidents are not infectious for start.

Followup on 2 Weeks Back Predictions

I first published the above table for doubling of cases in India 2 weeks back. At that time I had just started to do my own analysis of the infectious virus COVID-19 in India. I made some predictions at the time and now in the updated table above, in yellow, I have shown the variance of my predictions compared to the actual.

My forecast was pretty on spot for the first 6 days post my analysis, but for the remainder of the 6 days the forecast turned out to be rather off. The reason is pretty obvious to me; the government has started to systemically lift the lock down in India keeping in mind the economy and the lively hood of people.

Whether the government gave in to the critics of the lock down or did their own analysis to take this decision is only known to them. Either ways, results are out and it is clear the number of cases are on the rise. However, with the recent developments, what can we predict for the next 2 weeks?

Next 14 Days Prediction for Doubling of COVID-19 Cases in India

The situation has changed, and therefore I have made some changes in how I forecast the doubling of cases in India for the next 2 weeks. Given the lock down is mostly going to be eased, I will have to now factor in the R Naught of COVID-19 to predict the doubling of cases in India for the next 14 days.

There are 3 theories to consider now –

1. R Naught (R0) Theory. R Naught indicates how many persons does 1 infected person infect. The WHO R0 for COVID-19 is 2.5. However, instead of taking a R0 of 2.5 only and making one prediction, I need to factor in a few situations. Therefore, I have created a forecast range based on a conservative R Naught and a slightly more aggressive R0. I have taken the conservative R Naught (R0) as 1.8 and the the aggressive R0 as 2.5. I am being conservative and slightly aggressive with this number because while the lock down in India is being lifted, there are still good policies for containment zones and mass gathering. For example, restaurants, malls, movie theaters, concerts, etc will remain closed. At the same time, execution of these policies plays a big role in the actual number.
2. It takes anywhere between 7-14 days for an infected person to show symptoms of the virus. Instead of taking 7 or 14 days, I am taking the mid value of 10 days for my analysis.
3. I am making an assumption that the day someone is diagnosed COVID-19 positive, there is no delay (in days) in putting the person in quarantine and hence his individual spread of COVID-19 would fall to zero. Therefore, the New Cases reported daily plays an important role in this prediction as compared to the Total Cases as of that date.

Now that these assumptions and considerations are stated, lets move forward with the two extremes of this forecast range for the next 14 days.

With R Naught of 1.8

Considering the above points, below is the new forecast table for doubling of COVID-19 cases in India. Note, the number of total cases until 17th May 2020 have already doubled as of 31st May 2020. Therefore, the below table starts from 18th May 2020.

Looking at the forecast for doubling of cases in India for COVID-19, this Key Performance Indicator of days to double does not seem to deteriorate over the next 14 days under the conservative model with a R Naught of 1.8. The days to double seems to stay at 14 days consistently despite India hitting record high numbers on a daily basis. If you refer the chat trend chart at the top of this article, 14 days is the second best India has hit until now.

With a R-Naught of 1.8, in the next 14 days the total number of COVID-19 cases would jump from ~182,000 to nearly 394,000. This might sound bad, but with the Days of doubling staying at 14, it might just give India’s healthcare system enough bandwidth to absorb the new cases while still able to treat the older patients.

On the flip side, lets say India does not maintain a R Naught of 1.8 due to poor execution of maintaining the lock down. What would the situation be?

With R Naught of 2.5

Here is how the forecast of covid-19 cases looks like over the next 14 days if the new lock down is not able to contain the spread.

This slightly aggressive model puts an upper limit to the extent of the spread of COVID-19. Under this model the lowest number of days to double would be 11 days. While the total number of cases at the end of the next 14 days would be around 364,000 as compared to today’s 182,000 total cases.

What worries me a little with this model is the number of new cases per day touching 43,000.

Conclusion

On a conservative side we will see the daily new cases count go up to 22,000 and on the extreme side it might go up to 43,000 new cases. It currently stands at 8,380. As for the doubling of the new cases, I now doubt we are going to cross the 14 day barrier anytime soon.

If you ask me, I would put my money for the next 14 days to look like somewhere between the two extremes called out below. Here is a quick comparison of the two models –

Which side of the above extremes will India be in the next 14 days will come down on the government polices and the ability to execute them. And obviously the government is not the only one playing a role here, the citizens of this country will play a bigger role by being sensible in their choices. After all, it is now all about “Atmanirbhar“! I personally feel it has always been about Atmanirbhar; meaning self dependent.

Is Days to Double The Correct KPI In The Future

In short, No.

Days to Double would probably become redundant as a Key Performance Indicator once the Total number of cases cross a certain threshold.

As the total number of cases increase dramatically and the number of recoveries from COVID-19 increase, Days to Double would naturally start to increase dramatically and break the current 14 day barrier. At that point, the Days to Double would not be a good indicator of initiatives undertaken by India to control the spread of this virus. This event will also mark the beginning of the benefits gained from Herd Immunity.

With over 1.7 Million cases, we can already see this happening in the USA. In the below chart you can very clearly see the sharp correlation days to double has with the Total number of cases in the USA. Today, the Days to Double in the USA stands at a little less than 80 days as compared to nearly 10 Days at the end of March.

Note – you can download my analysis excel file by clicking on “add to cart”. Obviously, it is free but it will require you to register with the blog.

Citation –

Analysis – Done by Anant Agnihotri, published on https://mymixedthoughtsblog.com/covid-19-doubling-india-analysis-2/

Raw data – Max Roser, Hannah Ritchie, Esteban Ortiz-Ospina and Joe Hasell (2020) – “Coronavirus Pandemic (COVID-19)”. Published online at OurWorldInData.org. Retrieved from: ‘https://ourworldindata.org/coronavirus’ [Online Resource]

-Anant Agnihotri