Recovery rate

In one of my previous posts, I had pegged the recovery rate for Covid 19 as close to 99% based on the serological study conducted by ICMR. Of course no one took me seriously.

So let’s look at the recovery rate based on the numbers we know – ie the numbers that you’re tracking on a daily basis in the newspapers and worldometer. Thanks to Prashanth Jnanendra for planting the seed for this post.

Based on current reports, the current recovery rate is over 50%. Pretty good as is, but as misleading as it can get.

Worldometer pegs the current recovery rate as 95%. Much better than 50%, but still not accurate as its based on any case which has had an outcome.

Imagine there are 1000 infected every day and 1 person dies every day. And it takes 14 days on average for someone to recover. So on Day 15, there would be 15,000 cases but only 999 recoveries from Day 1. So the recovery rate is 999 divided by 15,000, or 6.6%. On Day 21, there would be 21,000 cases but only the first 7 days would have recovered. So voila, the recovery rate is 33.3%!

Improvement because of better medical facilities? Impact of lockdown? Wrong. It’s just simple maths.

Back to Day 1. You need to check how many from Day 1 have recovered. You can’t add on the guys who are still recovering to calculate the recovery rate. So for day 1, the recovery rate is 99.9%.

We don’t have the data on recoveries by day. But if you look at it, I’m pretty sure it’ll come close to the ICMR number of 99%.

Back to this graph. If you look at the slope of the recoveries, it’s following the slope of the infections, except lagging behind by 2 to 3 weeks accounting for the recovery period. It’s like a one day cricket match. The oppositions run rate is following the first innings, but just lagging behind a few overs. And that’s good news.