Since everyone has jumped on the leaked government report from Punjab and making their own conclusions without actually looking at the study and its details, here's a thread on why it was better that this document was not released and why it is an erroneous document.
1. Firstly, the document is incomplete without the complete study. Nobody knows the methodology of sample selection and determination of conclusions based on sample results. There is no way to know if there was sample bias and the straight extrapolation of positivity in sample to
the whole population of Lahore is just amateur statistics. 2. Now, assuming that the results of the study are accurate and it was a perfectly designed study, the conclusions and recommendations make no sense. The first group (RTS) finds prevalence in essential services that
operated even during the lockdown. The second 'smart sample' finds prevalence in the randomly selected sample of population as the name and description of strategy suggests (can't be sure without full study). If the prevalence is higher in random population compared to
essential services which operated even during the lockdown, the recommendation for further lockdown makes no sense at all. Your results negate your recommendation because those not observing the lockdown are found to have lower prevalence than those who might have observed some
lockdown. The expert working group would need to explain how it reached that conclusion because the results don't support that. They say it should be done because the disease has spread everywhere, but will the lockdown stop further spread? The results don't suggest.
3. Now onto the numbers since everyone is fixated on them and throwing around the 670,000 figure. As mentioned earlier, the document is incomplete at best and completely erroneous at worst due to the simplistic statistical modelling. Assuming that the figure they have quoted is
completely correct and there were actually 670,000 infections 15 days ago. This would mean that actual infections today would be around 1.4 million because our doubling time is 15 days roughly. Till now Lahore has reported approximately 200 deaths. Assuming Punjab has grossly
under-reported the deaths and they have missed 4 deaths for every 1, they'd have at maximum 1,000 deaths which is still a liberal figure. This would bring the IFR to 0.07%. Assume the worst case scenario that it infects 60% of Pakistan till the end of year to reach herd
immunity, it would cause 92,000 fatalities till December before reaching herd immunity. A big number as it seems, but go talk to any epidemiologist or public health expert (not the random expert) and they wouldn't really be worried about it. With such numbers they
would advise for minimum lockdown to flatten the curve, extra precautions for at risk group and enhancing HC capacity while doing better clinical management. Since, epidemiologists and public health experts are not even invited to our mainstream media, we never hear from them and
assume that they all advocate for draconian lockdowns which they absolutely don't. Even with higher numbers and IFR they don't suggest for strict lockdowns.
4. Now before you all jump on me for advocating herd immunity or something, I am not advocating for it. I think
this study or at least the document leaked is erroneous and not helpful at all. Before jumping on any study or document, we need to look at the numbers and make sense of them. The only thing Punjab govt should be criticized for is of conducting this poorly designed study and
coming up with faulty conclusions. And to the person who leaked it, this was not helpful and was panic inducing. Now that the document has leaked, Punjab govt should at least release the whole study so that a better understanding can be developed with better conclusions and
recommendations. Till then you all need to stop jumping on the study and going hysterical over it. At best it's a study that proves the disease to be much milder and at worst it is a completely rubbish study.
You can follow @smzrz.
Tip: mention @twtextapp on a Twitter thread with the keyword “unroll” to get a link to it.

Latest Threads Unrolled: