Share this page!

Murad Banaji Profile picture
Twitter logo
31 Dec 20, 10 tweets, 4 min read
Mumbai #COVID19 update. Things have improved - daily recorded cases and deaths are at about a quarter of second wave peak values. But there is still a steady stream of new infections. Mumbai's data forces us to ask: could reinfection be more common than we think? 1/10 Image
The recent picture:
- the second wave receded steadily, apart from a small post-Diwali blip
- daily cases, daily deaths and test positivity have all been declining
- but the epidemic is still not dying away. Image
Each day there are about 600 new cases. If we (optimistically) assume 1 in 10 infections are detected, and each infection is "active" for 10 days, that's 600X10X10 = 60,000 active infections = about 0.5% of the city. Low compared to May or September, but still surprisingly high.
Where are these infections occurring? We know that Mumbai has had two linked epidemics: in the slums and nonslum areas. The first wave hit the slums badly. The second had more nonslum infections. Current infections seem to be slightly more from nonslum areas. ImageImage
Even so, at the moment the slums are generating about 40% of new infections. There are uncertainties in this estimate, but however we look at it, a few thousand slum dwellers are being infected daily. This is surprising because...
...Mumbai's slums have already seen very high levels of infection. Different approaches (using seroprevalence data, fatality data, modelling, etc.) all suggest that over 60% in the city as a whole have had COVID, probably over 65%. In the slums it is 75% or higher.
Modelling suggests the continuing slum epidemic can be explained via
a very high 'R0' in the slums
+ the nonslum epidemic leading to some transmission from nonslum to slum dwellers
+ (possibly) a growth in the susceptible population e.g., through migration or loss of immunity
What proportion of current infections are reinfections? We don't know. Many of the slum infections occurred in April and May, 7 to 8 months ago. Loss of immunity wouldn't have to be very rapid or very common to play a part in feeding the epidemic.
nationalgeographic.com/science/2020/1…
Conclusion: nothing definitive, but Mumbai's continuing slum epidemic hints at some reinfections. If so, this certainly complicates all arguments around herd immunity. (Replug of this herd immunity explainer.)
science.thewire.in/health/india-c…
Some background and technical details of the calculations are in earlier threads and in docs linked from here: maths.mdx.ac.uk/research/model…

A more complete one-stop write-up with all the data-driven and modelling work will follow. (At some point.) 10/10

• • •

Missing some Tweet in this thread? You can try to force a refresh
　

Keep Current with Murad Banaji

Murad Banaji Profile picture

Stay in touch and get notified when new unrolls are available from this author!

Read all threads

This Thread may be Removed Anytime!

PDF

Twitter may remove this content at anytime! Save it as PDF for later use!

Try unrolling a thread yourself!

how to unroll video
  1. Follow @ThreadReaderApp to mention us!

  2. From a Twitter thread mention us with a keyword "unroll"
@threadreaderapp unroll

Practice here first or read more on our help page!

More from @muradbanaji

Murad Banaji Profile picture
6 Feb
For this piece, I tried to gather together some thoughts about the serosurvey data coming out of India. There is a great deal of this data, and the messages are important but not always clear. Longish #thread. 1/n
scroll.in/article/986097…
First, to see the bogus narratives you can construct when you ignore serosurvey data, you just have to look at Chapter 1 of the recent "Economic Survey". (This thread took just one example, but there are many.)
Key message from the serosurveys? Extremely variable surveillance of infections *and probably deaths*. In some places, a decent proportion of infections are picked up; in others a tiny fraction. Some areas have seen a huge number of infections, but almost no recorded deaths.
Read 16 tweets
Murad Banaji Profile picture
5 Feb
My piece in The Wire on the third national serosurvey. The headline (~21% prevalence) is probably not a major underestimate. Apparently the antibody test used was less vulnerable to missing old infections than the one used in the second serosurvey. 1/6
science.thewire.in/health/third-n…
The increase in prevalence from 2nd to 3rd survey is roughly consistent with the increase in cases over this period.

The breakdown of prevalence suggests that disease was moving towards rural areas even as daily cases peaked and declined nationally (September). 2/6
Weaker rural surveillance of infections and deaths could explain a moderate drop in detection, and a more noticeable drop in the naive infection fatality rate (recorded deaths over estimated infections) between 2nd and 3rd national serosurveys. 3/6
Read 6 tweets
Murad Banaji Profile picture
1 Feb
Chapter 1 of the Economic Survey 2020-21 (indiabudget.gov.in/economicsurvey/): a glowing report on the govt's handling of the COVID crisis, and a case study in what happens when science is abandoned for propaganda. Page after page of spin and dishonesty, but here's just one example... 1/5
...These two pictures show "actual vs. expected" cases and deaths for different states. Bihar does well (green) and Chhattisgarh does badly (red). Why? Because Bihar has fewer recorded COVID cases and deaths than "expected", and Chhattisgarh more than "expected". But... 2/5
I'd earlier examined data from Bihar and Chhattisgarh quite carefully for a piece in The Wire Science - measured cases, recorded deaths and, crucially, serosurvey data. The basic conclusions of this analysis were simple. In the surveyed districts... 3/5
science.thewire.in/health/bihar-c…
Read 5 tweets
Murad Banaji Profile picture
19 Oct 20
How many have had #COVID19 in Delhi? What % of infections have been detected? What is the fatality rate?

A (longish) #thread on Delhi's epidemic, with some analysis of its three serosurveys + other data. Details in a technical document linked at the end. (1/11)
First, Delhi's current surge (which may be winding down) is real - not just about better detection. But the actual surge in infections has been considerably smaller than in June - detection has increased a lot, making it seem larger. (Similar story to Mumbai - more later.) 2/
Prevalence estimate: by mid-August between 37% and 49% of Delhi people had had COVID. By mid-September: between 43% and 60%. The wide range reflects many uncertainties. When the October serosurvey results are out, we'll know more. (bloombergquint.com/coronavirus-ou…) 3/
Read 11 tweets
Murad Banaji Profile picture
11 Aug 20
Some general thoughts (in random order) on COVID-19 death undercounting in India following the recent article in The Hindu (thehindu.com/opinion/op-ed/…) by @giridar100 and collaborator, and the response from @oommen (orfonline.org/expert-speak/f…). 1/
The documented instances of COVID-19 death undercounting are too many and too major in scale to be treated as aberrations. Many have followed similar patterns - for example omitting deaths from "comorbidities" and suspected deaths.
The "urban areas have high MCCD coverage" argument is something of a red herring. Delhi, with its high MCCD coverage, saw *huge* undercounting before subjected to pressure. Mumbai too added in lots of missed COVID-19 deaths in June (~1700).
Read 11 tweets
Murad Banaji Profile picture
31 Jul 20
I took a look at Mumbai's #COVID19 data by age and came to some worrying conclusions. I'd encourage people celebrating seemingly low IFR values to look more closely. I used: Spain's IFR data; and Mumbai's age pyramid, age-structured fatality data, and seroprevalence data. 1/6
Naive IFR from Mumbai's seroprevalence and fatality data is ~0.12% (not, I think, 0.05-0.1% as is widely quoted).

An "expected IFR" for Mumbai, using Mumbai's age pyramid and Spain's age-dependent IFR is ~0.22%.

These two values hide a more complicated story. 2/6
IFR in some age groups, particularly 40-60, appears to be quite a lot higher (about 60% higher) in Mumbai than in Spain, even *assuming no death undercounting*. If there is death undercounting in this age group, then IFR goes up further. This should be a cause for concern. 3/6
Read 6 tweets

Did Thread Reader help you today?

Support us! We are indie developers!

This site is made by just two indie developers on a laptop doing marketing, support and development! Read more about the story.

Become a Premium Member ($3/month or $30/year) and get exclusive features!

Become Premium

Too expensive? Make a small donation by buying us coffee ($5) or help with server cost ($10)

Donate via Paypal Become our Patreon

Thank you for your support!

Follow Us on Twitter!

Like this author's thread?

Get emails when @muradbanaji has new unrolls!

Check out other threads from @muradbanaji!

Read all threads by @muradbanaji