Is India heading for a worst case scenario with the #WuhanVirus?
Growth trajectory on the log scale is linear for India as for most countries. Keeping with the known exponential growth in the early phase of an epidemic before it flattens out.
Even in an absolute sense, it can be seen from the graph that the growth trajectory is very similar to Australia, Thailand and Philippines. Growth trajectory seems conformant with global trends and standard models.
#WuhanVirus #Takeaway1 - The exponential increase in numbers seen over the last few days is very much on expected lines
#WuhanVirus #Takeaway2 - We typically tend to think of risk factors on a linear scale, but an epidemic works on an exponential scale
Is #WuhanVirus spreading faster in India?
Looking at the time it takes to double in the log(growthRates) to time of onset is a way to understand the “stun” effect at work. Italy and S Korea-similar time periods to double. SKorea strictly enforced counter-measures, Italy failed
Super-spreader events will lead to sudden explosion in cases, as in the case of SKorea, Denmark or Pak.
US-only country with a low K score but still hard-hit. US graph b/w 50 & 150 is non-linear even on log-scale. In fact it is already supralinear-staring at a super spreader?
For larger countries, performance of individual clusters must be monitored separately. Early trends in USA dominated by Washington state, but later steep rise is driven by California, New York and Massachusetts.
#WuhanVirus #Takeaway3-Spread of the virus seems in consonance with several other countries in the initial phase
#WuhanVirus #Takeaway4-It is imperative we clamp down and enforce strict protocols, else we may be looking at a catastrophe #JantaCurfew
State-wise analysis of #WuhanVirus cases
While there is insufficient nos for a scatter plot of different states in India, qualitatively, one can notice the 2 distinct trends.
1st set of which starts early, show constant or progressively decreasing slope (MH, KA, KL, DL, UP)
2nd set - maintain low slopes but suddenly show steep increase at late dates. Comparable to Pakistan, Denmark any one of the fat tails may grow explosively and wag the dog!
Impact on healthcare system #WuhanVirus
Early phase - states like MH, KA, KL, DL and UP account for a large proportion of numbers. Their log growth rate has largely been in the range of 0.1-0.15 log units/day. This would mean doubling in about a week’s time.
Maharashtra can expect to see about 500-700 cases in 2 weeks time, 1500+ in 3 weeks and so on. At 3-5 ICU beds per lakh, Maharashtra with a population of 11.2 crores can very optimistically expect to support about 4K ICU beds and probably about 2K ventilators
Assuming even 30% of patients need ventilation, Maharashtra can survive for about 5 weeks of 0.1-0.15 rate growth in cases.
What next? #WuhanVirus
Indian growth rate on log scale has hovered between 0.14 and 0.22 log units / day. That means roughly doubling every 5-7 days.
At this rate we can expect to hit a count of 400 by 22nd March, 1000+ by 30th March and 2.5K+ by April 5th.
Considering that 50% cases are currently imported and the ban on incoming flights comes into effect on 22nd, we can expect the log rate of increase to dip by the first week of April. But by then effects of community spreading may also come into play. Assuming latency of 7 days before onset of symptoms, the community spreading effects during the week of Mar 18-25 will play a significant role in our future course.
Options for India #WuhanVirus
(1) ‘The dance’ as Tomas Pueyo calls it: Active tracking and quarantining of suspects; large random testing as SKorea did.
Large random testing-Can’t afford to spray bullets around and risk being caught with no ammo when the enemy actually appears.
Instead we went after early tracking of incoming infectors and quarantined them. It delayed the onset of a full blown epidemic, buying us crucial time. Bound to break down sooner or later, given the propensity of our people to break rules & behave recklessly & it has
(2) Suppression-Strict lockdowns & isolation can reduce community spread, but imposes economic costs on lower strata. This siege can last for quite long times
(3) You try your best to slow down the infection rate and help the healthcare system keep pace. Largely “Ram Bharose” !
Which do we choose? None of the above !
Way ahead for India #WuhanVirus
Hyperlocal management - Divide & conquer
Treat problems at population clusters of about a few lakhs to tens of lakhs - typically a district or a couple of districts.
Cut off each of these clusters from the rest of the country
Impose lockdowns on severely affected districts - run internally with no external contact for the general populace
Life goes on normal within unaffected district boundaries
Has several advantages
(1) Administrative machinery at the level of districts is well established.
(2) Targeting high testing in affected districts becomes feasible.
(3) Even if new clusters emerge, the overall infection rates and growths in the country remain stable
(4) Economic activity within less affected parts can remain normal or recover faster to support the affected ones.
Lower income groups can survive longer lockdowns when in their native villages due to community support, small land holdings
In the long run this exercise will boost and strengthen local supply chains.
If managed properly, this crisis can be the beginning of a new economic order.
There's a risk of homebound workers taking the infections deep into the hinterland. This is where hyperlocalism comes into play…. Districts MUST work like independent units and quarantine the new returnees…
covid19org.in for India data
https://coronavirus.jhu.edu/ for Global data
This was published here first -