You can calculate Covid-19 Survival Rate

As an insurance person / actuary, I have been repeatedly asked what the survival rate is from Covid-19.

It occurred to me that the historical death rate for the New York City area can be approximated. This article can show you how to verify it for yourself. It is based on the idea that a) The CDC has been testing for anti-bodies in New York and can provide a valid estimate of the number of people who had the disease (this is different than the number of people who have confirmed the disease) and b) publicly available death data from Covid-19 in the New York City area is accurate. If you believe these two things, I will show you how to approximate the historical survival rate.

How many had Covid-19?

The CDC estimated this! You can obtain these numbers by following the link below. Make sure you select the site = New York City Metro Area and the Round = 2:

You should see that the CDC believes these percentages by age:

(For example, the CDC thinks that 29.1% of the people age 50-64 had the disease!!!! ... unless this is some sort of typo on their website, but it has been there for weeks now. They recently came out with Round 3, which has lower numbers, but I stuck to round 2 so I could match historical rates of the disease to historical deaths.)

How many died from Covid-19?

Next find out how many people have died in New York by age. I got this from the "statistica website"

I pulled this data as of 8/12/2020 per 100k.

If you divide this by 100k you get:

This is the chance of death from Covid-19 by age for those in the New York City area during this time period.

How many died as a percentage who had the diseases?

We want to know how many died who had the disease, not how many died who lived in New York. To do this, match up the age groups from the two studies in an approximate way, 0-18 matched to 0-17, 19-49 matched to 18-44, 50-64 matched to 45-64 and 75+.

Then adjust upward the ratio of death by recognizing that only a portion of the people had the disease. This is to represent the chances of death if you had the disease (not if you just lived in New York.) For example, to compute 0-18, take 0.000750% and divide by 21.5%. This puts the chance of death for those who got the disease in that age bracket at 0.000750% / 21.5% = 0.0035%

If you keep following this process for each age bracket, you get the percentage who died who contracted the disease:

Survival Rate

To turn this into survival rates, subtract the values from above from 1. (For example 1-10.740% becomes 89.26%)

#covid