COVID-19: India's new daily case counts decline attributed to insufficient testing (Al Jazeera)
“Please note that a huge fall in daily cases … is largely due to a heavy fall in testing,” Rijo M John, a professor and health economist at the Indian Institute of Management in the southern state of Kerala, said on Twitter.
“This should not be taken as an indication of falling cases, rather a matter of missing out on too many positive cases!”
(27 Apr 2021)
https://www.aljazeera.com/news/2021/4/27/india-nears-200000-covid-deaths-army-steps-in
I'd suggested a possible inflection point / actual decline a couple of days ago, though cautioned other explanations might also be responsible. My first caution then seems to be accurate:
India’s case rise has been so large and fast that it may simply be that detection capabilities are limiting new counts, whether by overwhelmed hospitals, labs, transport, or other means.
New daily case counts remain below trend, though at this point, faith in the numbers is low. Until we see deaths data, both attributed to Covid (lagging by 2-3 weeks) and total excess mortality (lagging by 3--6 months), what's happening right now will remain unclear at best.
Again, the take-away is that interpreting reported statistics is an art, and apparent trends are often attributes of the data collection and reporting process rather than the underlying phenomena.
Looking through other country's data earlier today, I noted that Egypt's new daily cases plot shows an unusually low amount of variability and noise. Given that the cases being reported are relatively few (presently just hitting 1,000/day), this leads me to suspect data adjustments. The best case would be that reported data are smoothed. Worst would be that they're simply being fabricated. Either way, this is another case of numbers which don't inspire much faith.
