Read on Twitter
My #Covid_19 projections for #Pakistan : We have passed the peak, Total Deaths by the end of this wave ~7000, Total actual infections ~35 Million, but total recorded infections may not exceed 50K.
Comparing with absolute worst case scenario (entire population is 60+ and/or has diabetes or BP) Still not in the Millions as some were projecting
Daily recorded vs Projected infections in Log scale to see how they line up
My projections are based on the following assumptions;
1. Results of seroprevelance study in Punjab in May which held infection rate to be ~6%
2. I extrapolated this to be the actual infection rate in all of Pakistan (12.7 Mil) Which incidentally was 431 times the recorded/reported infections.
Fitted this to an SIR table while considering T0 to be in early Dec
3. Its emerging that in order to have herd immunity we may actually need 45% to be infected. Evidence is mounting however that 20% may be enough to achieve what some epidemiologists are calling 'burnout'
https://www.medrxiv.org/content/10.1101/2020.04.27.20081893v3 https://twitter.com/gummibear737/status/1275118332887392256?s=20
https://www.medrxiv.org/content/10.1101/2020.04.27.20081893v3 https://twitter.com/gummibear737/status/1275118332887392256?s=20
This may be in part because we can already have some degree of immunity because of previous infections from relates coronaviruses/flu https://www.cell.com/cell/fulltext/S0092-8674(20)30610-3
4. Infection Fatality Rate IFR is still unclear but we get a range looking at meta studies https://www.medrxiv.org/content/10.1101/2020.05.13.20101253v2
I basically fit the recorded death figures to the projected infections based on the TWG sero study and got an estimated figure for IFR=0.01. This may be wrong and many deaths may go unreported or undiagnosed. My projection considers only what may be finally reported.
Lastly: I hesitate to call this a "model" as its just rough arithmetic based on assumptions and scant available data but i thought i would put my cards on the table.Critique welcome. Special thanks to @UsmaanK84
for his inputs. The projections are mine alone.
for his inputs. The projections are mine alone.
P.S. I think there will be a second peak around November but it will much less than what we had this time.
For some additional information to support my claim of peak having passed https://twitter.com/irtikazr/status/1276630906271420416
About which data sis relevant to make predictions; https://twitter.com/IrtiKazr/status/1270462717443411970?s=20
