How reliable is India's data on COVID-19 deaths?

The COVID-19 deaths are based on those reported or occurring in hospitals. More than half the deaths in the country, especially those in the large states in the north, occur at home without the cause of death known.

There should be a sizeable number of deaths and associated morbidity due to COVID-19 in these domiciliary deaths that are not accounted in the official figures.

It is also possible that many deaths occurring in hospitals are wrongly attributed to COVID-19 for various reasons.

That is why it is necessary to have a dramatic increase in the number of rapid tests for COVID-19 conducted on a large number of people and that should become the basis for estimating the morbidity and mortality due to COVID-19, not relying only on hospital reports, says Prof K Srinivasan, former director, International Institute of Population Sciences, Mumbai.

• THE WAR AGAINST COVID-19

The COVID-19 pandemic that has ravaged most of the first world countries, causing large scale virus-induced morbidity and mortality, is also taking a heavy toll on India, though on a much lower scale as compared to developed countries like the United States, Russia, Spain, Italy, the United Kingdom and Germany, in terms of affected population per million.

Since the beginning of the epidemic in early February till May 30, India has recorded 176,792 COVID-19 infections and 5,100 deaths.

India's population in mid-2019 has been estimated at 1,364,418 million and that of the US is 329,065 million (as per the United Nations World Population Prospects, 2019).

Thus, India's population is slightly more than four times that of the US and when computed per million population, the COVID-19 infections rate in India is just 60 as compared to 5,256 in the US (88 times more) and the death rate is 1.94 as compared to 258.26 in the US (133 times more) (as on May 19).

India's achievements in controlling the infection and mortality due to COVID-19 are phenomenal compared to what has happened in the US.

Of course, India launched its nationwide preventive and control measures, such as stopping all international and domestic flights and imposing a nationwide lockdown from March 25, much earlier and more effectively across the length and breadth of the country than what the US could do.

But the differences between the data of the two countries are staggering, with the US recording 88 times more infections and 133 times more deaths as compared to India.

The nagging question that crops up is, how reliable is Indian data on infections and deaths due to COVID-19?

The rest of the world has doubts on the quality of India's COVID-19 data, and there are valid reasons for this.

Let us take the case of deaths due to COVID-19. In the US, all the deaths occur in hospitals or certified as such in hospitals with the cause of death ascertained, even if the patient is brought dead.

It is recorded according to the International Classification of Diseases (ICD-10), developed by the World Health Organisation and updated at periodic intervals based on latest medical research on diagnostic techniques for the existing and newer diseases.

The immediate or antecedent causes of death have to be recorded as per the ICD code and that can be done only by a doctor in a hospital setting.

All the deaths are not only recorded almost immediately, but also the cause of death is accurately known in most of the developed countries including the US.

But this is not the case in India.

The registration of deaths in India is woefully deficient in terms of the percentage of deaths registered, though there have been improvements in recent years, and the actual cause of death is known only for a very small percentage of deaths that take place in hospitals and nursing homes, and that too among those that follow the ICD-10 coding system.

First, the Completeness of Registration of Deaths (CoRD) in terms of percentage of deaths registered to estimated deaths in 2017 was 79.6 per cent for India as a whole, varying widely across the states; 18 states had CoRD over 80 per cent, 10 of them southern and western states having 100 per cent CoRD; while five Hindi-speaking states in the north having less than 50 per cent, with Bihar just at 42.7 per cent and Uttar Pradesh at 38.3 per cent.

The CoRD deaths highly correlated with the percentage of deaths that occurred in institutions since these are more likely to be registered.

The percentage of deaths that occurred in institutions in India as a whole in 2017 was 47 per cent. In Kerala it was over 78 per cent while those in Bihar (31.3%), Assam (34.7%), West Bengal (44.9%), Jharkhand (33.8%) and Chhattisgarh (39.5%) were far less than the national average.

The COVID-19 deaths are based on those reported or occurring in hospitals. More than half the deaths in the country, especially those in the large states in the north, occur at home without the cause of death known.

There should be a sizeable number of deaths and associated morbidity due to COVID-19 in these domiciliary deaths that are not accounted in the official figures.

It is also possible that many deaths occurring in hospitals are wrongly attributed to COVID-19 for various reasons.

That is why it is necessary to have a dramatic increase in the number of rapid tests for COVID-19 conducted on a large number of people and that should become the basis for estimating the morbidity and mortality due to COVID-19, not relying only on hospital reports.

'Tests, tests and tests' is the WHO slogan for ascertaining the magnitude of the spread of COVID-19 and planning the appropriate action.

The number of persons tested by the rapid testing kits for COVID-19 per million population, as on May 15, is 52,784 in Spain, 37,585 in Germany, 33,175 in the US -- and only 1,480 in India.

India has to conduct almost 50 million tests, compared to the two million tests carried out so far, in order to ascertain the real magnitude of the prevalence of the disease mostly in the rural areas.

Thus, it is possible that we are grossly underestimating the prevalence of COVID-19 in the country.

Carrying out such a large number of tests and establishing treatment facilities across the country needs massive expansion in the health infrastructure, density of health personnel and medical equipment facilities across the country. We must undertake this now.

India is currently spending just 1.28 per cent of its gross domestic product on public health (about Rs 62,400 crores), lower than 1.57 per cent spent in much poorer countries and around 4 per cent in many developed countries.

India is spending less than Rs 500 per capita on health. Its health infrastructure at all levels needs expansion both in number of sub-centres, community health centres, secondary health centres and district level hospitals and also qualitatively in the number of medical and paramedical personnel employed in each of them and the medical facilities available there.

The ministry of health and family welfare is committed to spend 2.5 per cent of India's GDP on public health by 2025.

It is encouraging to see that the Government of India recently announced a massive economic and health revival package with the catchy title of 'Atmanirbhar Bharat' (or self-reliant India).

About 10 per cent of the Rs 20 lakh crore package -- or 1 per cent of India's GDP -- that amounts to Rs 2 lakh crore, should be spent on public health, including quantitative and qualitative expansion at all levels.

This will come to approximately 40,000 crores per year to be spent or committed to be spent on public health.

This is an opportunity that we can ill-afford to lose now to revamp and modernise our public health system.

The need is there and the money is available; so we have to plan and go ahead with the schemes under a suitable commission of public health experts with different specialities.

K Srinivasan is a senior professor and former director of International Institute of Population Sciences, Mumbai.