India appears to be headed for gradual improvement on the pandemic front through the next few weeks, with greater relief expected by early July, statisticians said. The first signs of a peak, coming in mid-May, are in line with predictions made by a number of scientists last month.
India reported 281,386 new covid-19 cases on Sunday, the fewest since 20 April. The number last crossed the 400,000 mark on 8 May. Despite a modest increase in tests, the positivity rate—or the proportion of a covid-19 test turning positive—dropped to 18% from a peak of 23% last week.
The slowdown follows weeks of lockdowns in most parts of India.
The daily caseload could further decline to below 100,000 by 2 June and below 10,000 by 7 July, shows a model developed by faculty at the Indian Institutes of Technology in Kanpur and Hyderabad. A four-digit figure was last seen in February.
Another mathematical model, built by professors at the Indian Institute of Science, Bengaluru, shows the daily count could may slip below 100,000 by 26 May. This model assumed a 30-day lockdown from 27 April. The model projects covid cases inching up again once the curbs are lifted but show deaths below the 1,000 mark by early June.
However, any statistical prediction model is only as good as the data that goes into it, warned Gautam Menon, a professor at Ashoka University. Lower levels of testing than needed, as well as evidence of serious undercounting of both cases and deaths, may make such models detached from reality, he said.
Another study by the Institute of Health Metrics and Evaluation (IHME) at Washington University takes this factor into account. It estimates actual infections, including unreported cases. India is adding nearly 5 million cases a day, around 16 times the reported count, according to this model. The silver lining is that this, too, is on a sharp downward path, set to halve in a month.
Most statistical models assume various scenarios to make their predictions, ranging from lockdown strictness to mask usage.
Without any such rules, daily cases could have been half a million by now and a million by mid-June, shows a model built by the COV-IND-19 Study Group, a collective led by Bhramar Mukherjee of the University of Michigan. Mukherjee was one of the first to predict the imminent second wave, back in February.
Predicting a peak around mid-May, Christopher Murray, the IHME director, said on 23 April that the high volume of infections in India showed the virus “may run out of people to infect pretty soon”.
But another problem with the predictions is the spread of infections to rural areas. Murad Banaji, who teaches mathematics at Middlesex University, said: “If disease winds down in urban areas, but continues to spread in rural areas, it may look as though the worst is over in terms of cases and deaths simply because testing and death recording is generally higher in urban areas,” he said.
The current slowdown in infections is far from pan-India. Four major hotspots—Maharashtra, Uttar Pradesh, Delhi, and Chhattisgarh—have recorded a sharp dip in daily cases, but most others remain in the danger zone. The positivity rate remains high in Karnataka and Kerala.
Statistical models don’t clearly predict a third wave yet, which will depend on public behaviour, vaccinations, and government steps after the second wave subsides. Live Mint