Based on the current trend line of recoveries and active cases, India should hit its peak around the first week of September, and then decline or plateau downward as more people recover than are infected every day.
Just over a week ago, Prime Minister Modi asked his ministers to prepare “emergency plans” to deal with the spike in COVID cases in the five most severely affected states of the country. If one is looking for an admission from the government that its lockdown had failed, then this is it: instead of taking 21 days, Modi’s Mahabharata has lasted over 100 days. And the battle is only growing more intense by the day.
So how long will it last? How long before the case count reaches its peak and starts to decline? After how many deaths? If anyone in the government has an idea, she or he has kept it a deep secret. Mercifully, we have enough data now to make a reasonable estimate by ourselves. The news is not all that good, but it is sufficiently reassuring to make panic unnecessary. Better still, it shows us what needs to be done if the duration of the pandemic is to be shortened.
The following table gives the data from which these conclusions can be drawn. The method is relatively simple – doubtless there are more complex ways to model the future with the same data set – but is sufficiently robust to inspire some confidence in its results. The table covers the period from May 15 till July 4, and presents the three crucial parameters upon which its predictions are based: the total number of cases, the number of patients who have recovered, and the number of patients under active care.
May 15 has been chosen as the starting date because phase 3 of the lockdown had ended and normal life was just being resumed. It therefore gave the data a fairly uniform base, free of policy change-induced shocks. A comparison of the rates of change in these parameters makes it possible to credibly predict the approximate date on which, barring unforeseen shocks, the pandemic will peak. It can also, with less certainty, predict when it will die out.
Of course, a caveat is needed. Two big confounding factors remain testing rates and the government’s denial of community transmission. Recovery rates do increase because that’s the natural course during a pandemic and we won’t discover more new cases unless we test more, which is why leading epidemiologists like Dr Jayaprakash Muliyil think most cases are going under the radar. For this exercise, however, we will take the government’s data at face value.
What the graph tells us
Column 1 shows that although the absolute number of new cases has kept increasing, the rate of growth has slowed from doubling in 14 days till May 29, to 17 days by June 14. It has taken 20 days to double again by July 3-4. This slowdown is welcome, but it is the least reliable (although most quoted) datum for charting the future course of the disease, because it is the one most affected by government interventions.
In India, this is particularly so because a good part of the increase is a product of more intensive – though by no means adequate or even geographically uniform – testing, which has gone up from around 4200 tests per day on March 24, to more than 150,000 today. One consequence of this is the inclusion of a higher proportion of those who are mildly affected by the disease.
Column 2 gives a sense of the number of patients who recover every day. This has been rising much more rapidly than the infection rate throughout the 46 days of the survey, but the rate of recovery too has been slowing down. Recoveries per day doubled in just 9 days by May 24, then doubled again in another 12 days by June 5. This slowed down further after the lockdown was lifted to doubling in 15.5 days ( June20/21). In the last days of June the rate of recovery has slowed down still more, with the doubling rate likely to be 17 days. This is still higher than the rate of growth of new cases, but the difference had narrowed sharply, suggesting that, in India as a whole, the pandemic will subside far more slowly than one had earlier hoped.
Column 3 shows that the higher recovery rate has been mirrored by a slower growth of active cases. Whereas new cases increased by 6.5 times between May 15 and June 30, active cases increased by only four times. Better still, the doubling rate of active cases has also slowed sharply from 18 days between May 15 and June 3, to 25 days from June 2 to June 30.
Columns 4 and 5, plotted in the graph above, give us a sense of the average rate of improvement in these two parameters per day. It is 0.63% per day in recoveries and 0.54% per day in active cases. Moving averages of these rates have been sufficiently stable over the six weeks of the study to be taken as a basis for projecting how long it will take for the pandemic to peak.
They show that, barring presently unforeseen shocks, of which there can be many, recoveries will close in on, and may even equal, new cases – leading, then, to very small or even nil additions to the existing active cases in about 63 to 64 days. The total number of COVID-19 cases may, therefore, peak in the first week of September. The experience of European countries, China, Malaysia, Turkey and several others in Asia suggests that, barring new developments in the spread of the virus – particularly the big ‘if’ of community transmission in the rural areas and smaller cities and towns – the disease count is likely to fall fairly rapidly after that as the recoveries continue to grow. But then there are also countries which have either plateaued after peaking, or had a resurgence of cases (eg. Iran) thereafter. It is too early to tell which way India will go.
Apart from the inadequate amount of testing, particularly in some populous states like Uttar Pradesh and Bihar, the most disruptive of developments that could derail these predictions is the inability of even the augmented treatment facilities we are putting in place to cope with the sheer number of active cases there will be by the beginning of September. At the present 25-day doubling rate active cases will rise to around 1.2 million patients by that date.
Where will so many beds come from? In 2015, Price Waterhouse had estimated that there were 1.3 beds per 1000 population, in India. This amounted to 1.75 million beds, which was far from sufficient to meet the requirements that arose out of accidents, diseases, localised epidemics, pregnancies and premature births. Even if the number of beds has gone up 50 percent since then, and that all of the increase can be sequestered for coping with the pandemic, There will still be a shortage of 600,000 beds. The inevitable decline in the quality of treatment that will result from trying to cope, will slow down recovery and increase the number of deaths further.
As the enormity of the problem dawns upon the state governments, the knee-jerk reaction of many of them could be to reimpose lockdowns, or harden them where they already exist. This must not happen.
Our bitter experience of the past three months has shown that any lockdown that is not back-stopped by financial sustenance of the kind that was given by the government of Malaysia, will again fail to contain the disease. But the harm that it will do to the economy will persist for years to come. This is because once an airline, a hotel chain, a bank, a mobile telephony company, an industrial enterprise, a restaurant or a cinema hall goes bankrupt, debt recovery procedures kick in that make it impossible for it to restart again. Thus the end of the pandemic will not therefore re-start these enterprises automatically.
Delhi example worth emulating
One way to drastically shorten the fight against the pandemic and minimise the damage to the economy at the same time would be to follow the example that Delhi has set, first during the lockdown, and then after it was lifted at the end of May. Kejriwal, Sisodia, and their ministers understood from the very first day that the lockdown would deprive millions of workers of their livelihoods for no fault of theirs or their employers, so they had a moral duty to minimise the hardship the poor, in particular, would have to suffer.
The government, helped by cadres of the Aam Admi Party, did not wholly eliminate the distress caused by the sudden loss of jobs because a large proportion of the wage earners, especially the migrant workers, lived in distant suburbs across the Uttar Pradesh and Haryana borders, and were therefore prevented from entering the city by the stoppage of metro trains and buses. But it did reduce the stress and anxiety in the resident working class of the National Capital Territory and therefore its willingness to abide by the lockdown in the containment areas.
The other key decision of the Delhi government – one that the lieutenant governor tried recently to countermand before taking a U-turn – was to allow mildly affected patients to be treated at home. This reduced the pressure on the city’s medical infrastructure
The second is the introduction of rapid antigen testing on June 18. While this test has been criticised by many, including the WHO, on the grounds that it misses many people in the early stages of the infection, it allows for a much earlier start of treatment of those it has detected, because the results become available in half an hour instead of two days. This has led to a rapid increase in the number of recoveries, and consequently a plateauing of active cases after June 20. It has also raised the recovery rate to 0.71 %, substantially above the national average of 0.54%. As the table below shows, this speeded up the daily recoveries dramatically from an average of 400 a day till June 8 to 3,000 a day from June 17. This stopped the number of active cases from rising any further, and may have begun a marginal decline from June 27.
As a result, the ratio of recovered to total cases has risen from 42.7% on May 31 to 66.5% on June 30, at an average of 0.71% per day. At this rate, the number of recoveries will equal the total number of infected persons (less the 3% who will have died) in about 42 days, i.e by the middle of August.
Since three-quarters of all COVID-19 cases are concentrated in the dozen or so largest cities of India (Delhi and Mumbai alone account for almost a third), if all of them follow Delhi’s example, the number of cases could peak, and the pandemic begin to draw down, three weeks earlier than it is likely to do today.
Courtesy The Wire
