Mumbai’s sero-prevelance study - detailed report with technical details : tcs.tifr.res.in/~sandeepj/avai… @CT_Bergstrom @mlipsitch @cmyeaton @nataliexdean @apoorva_nyc @velumania 1/
2/Study period: 14 days in July 20.
Areas selected:Three wards (R-North, M-West and F-North) in Mumbai were chosen based on: (a) City and Suburban areas (b) East, West and North areas and (c) representative of areas with low to high prevalence based on reported cases on 02/06/20
Areas selected:Three wards (R-North, M-West and F-North) in Mumbai were chosen based on: (a) City and Suburban areas (b) East, West and North areas and (c) representative of areas with low to high prevalence based on reported cases on 02/06/20
3/Sampling design: Sampling from households in impoverished tenements or buildings, which were separated by at least 3 households/buildings ensured geographically separate and a systematic method for non-overlapping area coverage.
4/ In high-rises with more than 5 apartments, N/5 households were selected from separate floors (one household for every 5 apartments in a building). No more than one sample were collected per household/family and as per the age/gender stratification viz.
5/ Females and Males belonging to age brackets (a) 12-24 years (b) 25-39 years (c) 40-60 years and (d) >60 years.
6/ Sample sizes were calculated based on statistical considerations, which took into account
reported cases and population sizes in each of these wards as below (for slums and non-slums
separately)
reported cases and population sizes in each of these wards as below (for slums and non-slums
separately)
7/ 
8/
9/
10/ Key findings: 57% prevalence in slums (impoverished tenements) and 16% prevalence in non-slums. These numbers have not been corrected for 93% sensitivity of the CLIA test used, which puts the numbers at a lower conservative estimate.
11/ Sero-prev in impoverished tenements
12/ Sero-prev in non-slums
13/ tl;dr - many infections are missed, the virus spread very quickly in impoverished tenements (anyone suprised?). Higher prev d/t high population density, shared common facilities (toilets,water points).
14/ interesting questions on IFR (young demographics, less pre-existing illness, underreporting of deaths, better pre-existing cross immunity from common cold CoVs, a combination). Next sampling round should give more info re herd immunity @AdamJKucharski
