Hyderabad/Bhopal: India implemented the largest and most stringent lockdown worldwide on 25 March for 21 days to control the Covid-19 spread. This was further extended until 3 May, with some sectors and districts slated for relaxations from 20 April onwards.
The Ministry of Health & Family Welfare has claimed it is working, by showing that without the lockdown and containment measures, projected cases would have been 8.2 lakh by 11 April, against the actual 11,933 cases recorded on that date.
It has also claimed that the time needed for cases to double has increased from three days before the lockdown to 6.2 days after it.
Scientists at the Institute of Mathematical Sciences in Chennai have also estimated that the Rt, the real-time effective reproductive number, which shows how many people a Covid-19 patient can infect on average at a given time, had declined to 1.55 for the week of 6-13 April, indicating a slight flattening of the curve. This data indicates that the national lockdown is indisputably slowing down the transmission, but is it the right metric to measure the success of our Covid-19 strategy?
The cricketing equivalent
In a cricket match, say if the current run rate is 4 runs per over and the required run rate is 10, would it be a cause for relaxation if a pinch-hitter comes in and raises the current run rate to 6 (and required run rate comes down to 9)? Even more so, if you know that your star batter will probably get out in the next over (3 May)?
As we know, in the larger picture, it does not matter if pinch-hitters hit exquisite fours and sixes in the five overs they play to bump up the current run rate, if it does not help cross the required run rate, and more importantly, win the match.
In the same way, it is essential to have appropriate benchmarks to measure the success of the Covid response. Therefore, what is the ‘required run rate’ in the context of the Covid-19 pandemic?
We need to calculate backwards from the target, which is “not to let Covid-19 overwhelm the health system capacity”. Let’s use the bed capacity to illustrate this calculation. The government announced on 12 April that there were 1,05,980 dedicated hospital beds kept on standby for Covid-19 across the country.
This translates the ‘required run rate’ to a tangible question — how long will the 1,05,980 beds last for a given doubling rate or Rt number?
India’s ‘required run rate’
Assuming that the available beds remain constant, a simple model which assumes 20 per cent of all infections require hospitalisation (based on data from China and the US and reiterated by the Union health ministry) and a weighted average length of stay of 12 days per Covid patient (based on data from Kerala), will reveal that at a doubling rate of 6.2 days, the bed capacity will be exhausted by 22 May. Does this indicate that we have everything in control?
With a doubling rate of 10 days, the beds will be exhausted by 17 June, and at a rate of 30 days, by 25 November.
Using reproduction number data, instead, at an Rt of 1.55, the number of beds will be exhausted by 16 July. At 1.35, the date will be 27 August, with an Rt of 1.15, it would be 29 January 2021.
All this assumes that the doubling rate or Rt achieved during the lockdown is maintained beyond 3 May, which may not necessarily be the case. The lockdown is the highest level of forced social distancing that can be achieved — the star batter in our cricketing analogy. But when sectors and districts are opened (other batters in the middle to lower order), the doubling rate may decrease or Rt increase since people interact more, offering an opportunity for the infection to spread — maybe even rapidly.
The cruel maths of Covid
The question, perhaps, not to ask is if the model is accurate for our diverse country. It is imperfect, as all current models are, since they tend not to adequately factor in urban-rural disparity, the personnel required to run the infrastructure and beds, PPE kits for the personnel, other medical devices like ICU equipment and ventilators, and other pertinent factors.
Instead, the relevant question to ask is if the model is useful and can insights gained from this be used to guide public health action and policy.
What the model reveals is, just how cruel the maths of Covid-19 could be. It is a highly contagious disease, no doubt, but what makes it so dangerous is it gives a very narrow margin for error. Not surprisingly, high income countries like Italy, France, US have struggled to contain it, and those who initially succeeded in containing it, like Singapore or South Korea, are having a hard time maintaining that.
German Chancellor Angela Merkel, a trained scientist with a doctorate in quantum chemistry, while addressing lockdown relaxation said: “If we get to the point where everybody infects (Rt) 1.1 people, then by October, we will reach the capacity level of our health system, with the assumed level of intensive care beds. If it edges up further still, to 1.2…Germany’s health care system will reach its limit in July. At an Rt of 1.3, the healthcare system maxes out in June. So you see what little leeway we have.”
Germany has the highest number of ICU beds per 1,000 people in Europe, and has so far managed Covid-19 exceptionally well. But it is still grappling with concerns around “fragile intermediate success”. The situation will not be different for India, and may in fact be relatively worse beyond 3 May.
Communities need to participate actively
The benefits of the lockdown, which were gained at huge economic and social cost, could easily be lost if communities do not participate actively in helping to maintain them. The government needs to enhance its efforts on community engagement.
One way to increase community participation is for the government to be transparent, honest and upfront. It needs to convey the truth that this is a Test match, not a T20 that will end on 3 May. The Covid response will likely last for more than a year at a minimum, and everyone must participate for this duration, without complacency setting in once lockdown is lifted.
Prime Minister Narendra Modi, in his four national addresses to date, has been wary about emphasising this point, perhaps this is to avoid creating anxiety, but the prevailing uncertainty about a further extension in the lockdown and relaxations (if any) is already rising.
By helping people understand the protracted nature of the response, it’s more likely we will be able to reinforce that people need to increase their civic responsibility substantially and sustainably post lockdown. Reciprocally, the government needs to ensure it supports the citizenry by protecting the vulnerable, initiating and maintaining additional protection measures, and supporting a well-thought-through economic revival.
Target doubling rate and Rt
The second insight is gained by asking what the target for doubling rate of cases and Rt should be. If we assume that a vaccine will be developed in 12 months, ambitious as that is, we will need a doubling time of 45 days at least, or an Rt of less than 1.12. If this is achieved, then the bed capacity will last until April 2021.
This leads to another, deeper understanding that the ideal reproductive number to target is less than 1 (since there is a small difference between 1.12 and 1), a rate at which transmission stops. This can be achieved by aggressively focusing on continued contact-tracing to break chains of transmission, strengthen Integrated Diseases Surveillance Programme (IDSP) to detect outbreaks faster, and contain transmissions in hotspots to localise them.
Govt’s current strategy is short-term
But when one looks at the ads for recruiting epidemiologists across states, all the positions are for a short-term period of three months on contract basis, with a low salary between Rs 25,000 to Rs 60,500 per month offered. This is for trained professionals with usual qualifications of a medical undergraduate with a postgraduate public health degree.
There are no figures provided on the number of contact-tracers hired, something we advocated for in a prior article in ThePrint, to augment the response; the current contact-tracers seem to be mostly drawn from the existing health workforce (ASHAs, ANMs, etc.), which means other routine health services are on the back foot.
This is indicative of a short-sighted response mindset. Sadly, the government has adopted a stop-gap strategy in place of a sustainable robust public health strategy, and this could cost us adversely in the long run.
American statistician Nate Silver said, “Before we demand more of our data, we need to demand more of ourselves”. This alludes to the third insight — that instead of focusing on ephemeral gains of lockdown success in terms of doubling rate, or that only 1,671 beds of the 1,05,980 allotted for Covid-19 were occupied as on 12 April, we need to focus on critical indicators that show if we are ahead of the curve or resilient enough to face Covid.
For instance, the government updates the daily number of confirmed cases, which is then diced and sliced multiple ways without throwing light on other important matters. If we concur that contact-tracing is important, then the critical question we need to ask the government is that of the daily new cases, how many came from known contacts or people under observation/ quarantine?
The percent of new cases coming from known contacts will be a key performance indicator for the contact-tracing process.
In addition to giving the number of samples tested every day, the average turnaround time from sample collection to getting results would show the efficiency of the system. The average duration from symptom onset to admission or diagnosis will not only reveal the health system strength or weakness, but also indirectly assess community awareness, faith in health services, and participation in the response.
A dashboard of such critical indicators will increase data transparency and accountability and build trust between the public and government. Moreover, additional data, say, on the percentage requiring ICU care or ventilators, or average length of stay will help to build better nuanced national models, rather than models using foreign data.
We need to think out of the box. Since it is going to be a tightrope walk, once the lockdown is lifted, the only way out is to either increase bed capacity to act as a buffer (without affecting non-Covid services) or come up with creative solution like reverse isolation, as in Kerala.
Kerala’s modelling revealed that bed capacity will soon be overrun once the lockdown is lifted, but reverse isolation of those considered most vulnerable, such as the elderly or those with co-morbidities, who account for a major portion of hospitalisations, can buy some time.
Modelling is not necessarily meant to answer a question, but is rather a way of thinking, an approach to finding solutions. The Covid-19 virus is relentless and gives very little room for error. We all need to understand that as a nation. India, a vast, diverse country with limited resources, is in a precarious position.
We need to look beyond the lockdown, beyond sector-wise opening, beyond hotspots, to evolve a mid-to-long term sustainable strategy for Covid response, which balances the economic, social and public health benefits and costs equitably.
The past two months were an opening gambit; now India needs to handle the middle- and end-game well. Perhaps then, and only then, we might be able to win the Covid-19 Test match.
Dr Manjunath Shankar is a public health specialist. He participated in the US CDC Emergency Response (Modelling Task Force) to the West African Ebola outbreak in 2014-15. He tweets at @MonJunNot.
Dr Anant Bhan is a researcher in global health, bioethics and health policy. He tweets at @AnantBhan.