More Than Two-Thirds of COVID19 Deaths in India From 10 Most Populated Cities – Analysis

The population share of the ten most populated cities in India is only 5% of the total population but two-thirds of Covid19 deaths in India are being reported from these 10 cities. This skew was expected, however, I wanted to do this analysis and see if any patterns can be found. For those of you who are short on time, I have highlighted the key points so you can just scroll through quickly. You are more than welcome to read further on how I came up with these and some more data and graphs to back these up.

Please note that I don’t have a background in epidemiology and all inferences are based on data and trends.

Observations at the city level

The spread of the corona virus globally has been initiated around cities and India is no different. Instead of looking at the data at the state level, we should focus on cities. Penetration to rural areas is still quite low so discussing overall population numbers does not make sense (As per the world bank 2018 data, 66% of India is still rural)

1. Mumbai and Delhi account for almost one-third of India’s confirmed cases (32%).
2. Mumbai alone reported 23% of COVID19 deaths in India.
3. For Mumbai, the most impacted city in India, the mortality rate of 3.8%. It’s not far off from the country’s average (3.3%). Delhi, at 1.1% has one of the least mortality rates in the country. It will be interesting to marry this data with the age buckets.
4. Kolkata, at 10.6% has a very high mortality rate (almost 3 times India’s average). Ahmedabad, Pune and Baroda also have high mortality rates (twice as that of India’s average).
5. Bengaluru has an extremely low number of cases reported. This could mean that there are less number of infected cases or the wrong people were tested.
6. Bengaluru, Surat, Jaipur, Lucknow, Bhopal, Vadodara, Ghaziabad – In these cities, more than half the infected cases have recovered.

Observations at the state level

1. West Bengal, with 1 death in every 10 reported cases has the highest mortality percentage among all the Indian states.
2. Tamil Nadu and Delhi, with 1 Covid19 death every 100 reported cases have the most favourable mortality percentage (among states/ union territories with more than 1000 reported cases)
3. The recovery rate for the four most impacted states (Maharashtra, Gujarat, Delhi and Tamil Nadu) are all lower than the national average (30.6%). Maharashtra at 18.8% is a key reason for bringing the national average down. If we exclude these 4 states, the recovery rate for the rest of the country is 44%!
4. Kerala is showing the way in recovery to the rest of the country with a 96% recovery rate and a 0.8% mortality rate. Telangana (65%) and Rajasthan (58%) also have very high recovery rates. Is there anything we can learn from these states?

Are we doing enough testing?

In short, no. Currently, India is doing 1.17 tests every 1,000 people.

The good news? The number is getting better.

(May 12 UPDATE: 86,191 tests were done on May 11. The Union minister added that the government supplied 76.42 lakh N95 masks and 40.18 lakh PPE kits in the country.)

In the comparison below, please bear in mind that data for Iceland and Estonia is skewed because their population is substantially lesser than the other countries in the list. You can add other countries to the data below using this link.

Another interesting comparison is below. 25.5 tests are revealing one confirmed case in India. Just to make understanding this easier – lower this number, more worry-some it is. The same number for India on March 13th was 86.7.

You can add other countries to the data below using this link.

Below is the same data for the states in India. Out of the states with more than 1,000 confirmed cases, the 3 most impacted states have a substantially higher tested to confirmed ratio. Tamil Nadu, Uttar Pradesh, Rajasthan and Andhra Pradesh seem to have done a better job at containing the spread of the virus.

Things to think about

Large cities (especially Mumbai/ Delhi) are the hotspots and it might make sense from an epidemiology angle to continue keeping the borders closed to and from the top 10 cities till we actually see the flattening of the curve. This means that migrant workers will be impacted even more.

Midway will be for the state governments to prove to the central government/ national Covid19 task force that they have enough isolation areas ready and a dedicated task force in place. The central government should come up with a playbook for all the states to accept migrant workers. I was almost sure it’s already there but after reading about the 70 migrant workers who jumped quarantine in Bihar, I’m not sure any more.

After 22 migrant workers, who returned from hotspot Surat to Jharkhand, were tested Covid19 positive, the SDO said the return of migrant workers from other states has added a new dimension to the viral outbreak.

Another important aspect to worry about is the rise of asymptomatic cases. 75% of Delhi’s Covid-19 patients asymptomatic or have mild symptoms. The good news? Out of the 6,923 confirmed Covid-19 cases in Delhi, only 1,476 are/ were admitted at hospitals. The rest or about 75% have either recovered or are getting treatment at their homes and Covid-19 centres.

Last, but not the least, the biggest question – Was it a good decision to release the lockdown on May 4th? Time will tell.

If there is anything you can infer from this data or if anything I have written doesn’t make sense, please comment below!

Notes:

1. COVID Data Source: covid19india.org (Data till Sunday, May 10th 9 AM IST)
2. Population data is from 2011, the last official census. (Next one is scheduled for 2021)
3. Bengaluru include Bengaluru urban and rural. Kolkata includes Howrah
4. If anyone wants to use this data and do some further analysis, feel free to reach out to me
5. There have been some news about under-reporting of deaths in Delhi – this might be the case for some other states as well. This means the total number of deaths and the mortality rates will be higher than what’s currently reported