Thread by @UntergrundmannG

#Covid19India

Worked out the Testing performance for the States of India (based off partially reconstructed series, since ICMR has ceased to provide information). A major finding is that testing rate growth is strongly correlated with how well a state is doing ...

+

States thate have maintained a testing growth well above the growth of daily positives are able to HOLD DOWN the infection count growth. Those that manage to ramp it up, if not keep increasing, are able to fight down spikes and surges well.

Notice Bihar blowing up now? Why?
+

Delhi is an example of 'ramp up on need' - which explains why each spike/surge has been beaten down. Notive Haryana maintain a consistently higher growth level of tests over positives? Ditto for J&K - except for now.
Observe Gujarat here, the worst of ALL states.
#Covid19India
+

Jharkhand ramps up on need - a little delayed. Karnataka has started maintaining that edge and so it is able to check surges now, when they arise. Observe how inconsistent MP is.
Kerala maintains a very small edge, because it can afford to - with the community programs ...
+

Any surprises now - that Maharashtra's growth is unchecked? Observe Odisha for contrast.
Punjab and Rajashthan ramp up on need, but, of late the former appears of registering the importance of maintaining a decisive edge/margin.
#Covid19India
+

Now you can see why TN has achieved negative Actives growth.
And how UP lost the early control it seemed to have.
And why Uttarakhand is doing okay while West Bengal is exploding unnchecked.

#Covid19India
+

Now when you see the state for India Aggregate - you can think of WHY the #Lockdown has not been able to do more than merely damp the daily growth of infection (i.e. slow it down to a middling rate, preventing it from exploding).
Explains why the data shows no containment yet.

CC: @vavasthi

Data Source: api.covid19india.org/csv/

Rules for series reconstruction: Interpolation by compunded growth between gaps in series (e.g. missing data for some dates). Series termination before last date - extended to the end (for maintaining integrity of totals).

Source: api.covid19india.org/csv/

Growth Rate (Y-axis): Total Count on Day N / Total Count on Day N-1
X-Axis: Date

#Covid19India

Since it may still not be clear to some, continuing this thread with a view of the Impact of State Testing on National Case Count.

Plotted the % of Total (National) Tests that States have used with the % of Total (National) Positives they've contributed ...

+

Impact of AP's edge is clearly seen in the decline in its contribution to the National Total of Positives. Delhi's ramp-up when needed also manages to suppress its contribution. Bihar testing growth declined and see the blow-up now. Any doubt on why I called Gujarat the worst?
+

Haryana maintains a constant edge and the effect is clear to see. Karnataka has decided to do the same, with the same effect. J&K is ramp-up on need. Needs to do better now. Kerala has maintained a nice clear (but slight) edge throughout and the effect is loud and clear.
+

MP didn't increase testing and see the effect on rising positives as contribution to the National Count. Ditto for Maharashtra.
Punjab's been ramp-up on need and is now taking an edge. See the effect. Rajashtan is still ramp-up on need, so unable to drive positives growth down.+

Tamil Nadu has taken the edge on testing and driven the positives down. UP has clearly lost its way (by the 10th of April) and now we're seeing how difficult it finds to drive the growth down. West Bengal hasn't done much and now we see positives blowing up ...

+

The Q is : What else do you need - to know what to do? And HOW?

It isn't rocket science at all.
#Lockdown MUST be accompanied by commensurate and effective Testing with #TraceTestIsolate.

We can clearly see which States are failing the National Effort.

Adding a couple more things for clarity and to drive home the sense of impact.

#Covid19India

Impact on daily new case counts between Haryana (maintaining an edge for testing growth) & Gujarat (doing the opposite). Likewise for the two hotspots - Delhi & Maharashtra.

+

#Covid19India

if growth rates are not easy to understand, here's the same picture of the States above - in terms of Doubling Times. States doing well maintain a shorter doubling time for tests against that for positives.

Doubling Time = LogN(2)/LogN(Growth Rate)

+

As can be seen, Bihar and Gujarat are doing the opposite of what needed to be done - hence doubling time of positives has declined (i.e. faster growth of positives) which means increased counts of daily new cases.

#Covid19India
+

All four states here are doing well.
And good for Karnataka that it seems to have figured out that the Haryana approach is the way to go - if it means business.

#Covid19India
+

Not difficult to undertand why MP and Maharashtra are struggling - they've not been able to accelerate the growth of tests past the rise of positives they've encountered.
Notice the contrast with Kerala and Odisha.
#Covid19India
+

You can see how UP lost its early grip on the situation - by allowing the rate of growth of positives to outpace that of tests.
Punjab appears to have chosen to give a boost to testing.
Tamil Nadu has learnt well from the superspreader surge.
#Covid19India
+

Uttarakhand is consistently maintaining its edge on testing. While West Bengal appears to have barely responded to the growth of positives. No wonder the count is exploding.

#Covid19India