Now Reading
It’s Both Easier and Harder to Work Out How Many Indians Have Died Due to COVID-19

It’s Both Easier and Harder to Work Out How Many Indians Have Died Due to COVID-19

A student wearing a protective face mask and face shield gets her temperature measured as she arrives at an examination centre for the Joint Entrance Examination in Kolkata, September 1, 2020. Photo: Reuters/Rupak De Chowdhuri.

In 17th century Europe, and following the plague, many cities started counting the total number of deaths every week and analysed them by cause to find out how many people had died and of what.

Perhaps London had the best system in the 1660s. Officials would compile all deaths in all parishes of London in the previous week by cause, age and sex, analyse them, print them and distribute them among the people as the so-called ‘bills of mortality’ every Wednesday, at a small price. Officials also put together an annual report of deaths the Wednesday before Christmas every year.

Some interested people – especially John Graunt, regarded as the ‘founder of demography’ – worked with these reports to produce actuarial tables and estimate the prospective mortality rate. The bills and these analysts both attempted to track the plague as well as other important diseases, motivate public health action and observe its impact on disease mortality.

A bill of mortality published in 1665. Photo: Wellcome Collection gallery (March 31, 2018),
CC BY 4.0

In modern India, no city or even district undertakes such an exercise. In fact, to the best of my knowledge, nowhere in India does such a system exist despite the fact that we are in the middle of a pandemic.

Case numbers are unreliable

Experts and commentators of different persuasions and hues have advanced a variety of statistical methods and mathematical models to track COVID-19 cases and predict the pandemic’s course. Some predictions constructed using these tools provided the impetus for governments to impose lockdowns and other steps to arrest the virus’s spread.

However, the real case numbers are very difficult to get as they depend on testing frequency, testing strategy and the type of tests used. The more we test, the more cases we will diagnose. There are also 10-20-times as many people showing no symptoms of a COVID-19 infection as there patients with visible symptoms. (Many close contacts of a confirmed patient are not tested if they don’t have any outward symptoms.) Many people also don’t report to clinics or hospitals, especially when they are mildly symptomatic, and prefer to recover at home.

Indeed, many doctors and other health staff who have participated in community surveillance and testing programmes have said there are many people with cough and fever in areas badly affected by epidemic but who haven’t come forward to be tested. They were concerned that, since their symptoms were fairly mild, a ‘positive’ test would mean being isolated from their families — an undesirable state of affairs.

In addition, the ‘gold standard’ test for COVID-19, the RT-PCR test, typically has a sensitivity of 70% – that is, it will miss three out of every 10 cases. The rapid antigen test, now in wide use in many states, has an even lower sensitivity of around 50% – it will miss five out of every 10 cases.

Many eminent scientists, doctors and some media professionals have said India should test more to control its COVID-19 epidemic. The test positivity rate is currently about 10% – i.e. 10 tests need to be conducted to find the next case. An RT-PCR test costs about Rs 2,000 to Rs 4,000, so the direct cost of finding one case is Rs 20,000 to Rs 40,000.

In addition, fewer than 5% of those who get infected worldwide, although this figure is lower in India, go on to have a severe form of the disease, so the cost of finding a patient who needs hospitalisation due to COVID-19 Rs 2-4 lakh each. Testing and identifying cases is an expensive business. This is on top of the fact that many districts of India simply lack testing centres, so diagnoses of samples in labs are either delayed or don’t happen altogether.

In the final analysis, the case number is at best a rough estimate of the actual total number of cases. Hence even when tests are available free of cost, they don’t pick up all cases. Thus, the reported number of cases can only indicate the level of infection in a population, with a large margin of underestimation.

Also read: India Is Undercounting Its COVID-19 Deaths. This Is How.

Counting deaths

An alternative way to keep track of COVID-19 progression through the population is by the number of deaths. Specifically, we can compare the total number of deaths every week in any city or town with, say, the same week averaged over the previous five years – akin to what Europe did in the 17th century.

For example, based on historical data, we should be able to calculate the number of deaths expected for the city (with a 95% confidence interval), and then compare this figure with the number of deaths recorded in the city in a given week during the epidemic. The extent of this excess mortality will give an indication of the epidemic’s impact.

This method doesn’t depend on knowing the causes of death; instead, we compare the all-cause mortality with past trends. Such a system is already available for the European Union, called EuroMOMO. According to its data, many European countries had high levels of excess mortality during the pandemic.

Level of excess mortality in Europe. Source: EuroMOMO bulletin, week 34, 2020

The process of arriving at the cause of death is well established in public health and epidemiology. There are simple rules to determine the cause of death. However, it does need a strong public healthcare and health statistics system – of the sort that doesn’t yet exist in India. Some cities with good municipal health departments do maintain the number of daily and weekly death numbers, but officials don’t compile and/or publicise this data.

As it happens, there have been news reports of some cities in India where more people have been buried or cremated in accordance with protocols for COVID-19 victims than the number of people hospitals have reported as having died of COVID-19. However, no public health department or entity – including the Indian Council of Medical Research, the National Centre for Disease Control, and registrars of death who have a statutory responsibility to register deaths and analyse them – has produced a public report explaining excess mortality during the pandemic.

The government’s Aarogya Setu app also hasn’t paid any attention to capture the total mortality. But it’s curious that instead of insisting on ground-level available mortality data, which is as simple as it gets, many of us are more interested in mathematical models.

In India, we don’t allow the dead to speak so that the living may be saved. In Europe, the dead spoke, through the bills of mortality, and saved many lives. Even now, compiling death data doesn’t need much money – one clerk can call up the crematoria and burial grounds in a city and collect the data in minutes. But it needs political will and a mature understanding of public health.

Dr Dileep Mavalankar is the director of the Indian Institute of Public Health, Gandhinagar.

Scroll To Top