Relatives carry a coffin for a man who died from COVID-19 before his burial, New Delhi, June 8, 2020. Photo: Reuters/Danish Siddiqui.
‘Almost 50% of COVID-19 deaths in India among those aged below 60 years,’ the newspapers seemed to scream on May 22. Alarmed, I grabbed my glasses to confirm what my Netflix-induced blurry eyes had just read. It was my turn to scream! Something didn’t fit.
Infective organisms have specific, sometimes peculiar, traits and it’s very rare for these traits to change suddenly. HIV for example is transmitted only through body fluids. It can also spread if a needle used by an infected individual is used by another individual without it being sterilised. And yet, a mosquito that bites an infected individual and then bites another person without ‘sterilising’ its needle cannot transmit HIV. Thank god for this universal trait.
The novel coronavirus should be no different. Its traits and behaviour are peculiar to it. It’s precisely because it is behaving the same way everywhere that we’re calling our crisis a pandemic. Few subspecies may behave slightly differently but much of the crucial behaviour is the same.
It is this very fundamental notion that the headline challenges. Reports in late April claimed that over 40% COVID-19 deaths in India were in those younger than 60 years. Comparisons of COVID-19 age-wise deaths are available in the US, the UK and Italy. A cursory look tells us that 10-15% of those who died are younger than 65 years. Why then is this figure closer to 50% in India?
Statements about COVID-19, like ‘new cases are reducing’, ‘recoveries surpass active cases’ and ‘doubling time has increased’, only add to the confusion. And before we know it, people actually believe there are three types of lies, à la Mark Twain: simple lies, damned lies and statistics.
The true interpretation of these numbers lies in the accuracy of two components: numerator and denominator. An accurate figure of the number of people infected or cured depends on the diagnostic tests performed or even the definition of ‘cured’, which in turn would affect all the above statements.
The Government of India changed the discharge protocol such that an infected, asymptomatic person can be declared ‘cured’ and be discharged after 10 days, but without the previously mandated RT-PCR test. I am reminded of how the status of lakhs of Indians changed from ‘below poverty line’ to ‘above poverty line’ by a slight change in the definition of poverty – but their actual plight hadn’t changed.
The ‘doubling time’ can’t be used as an indicator of slowing spread of the virus when we aren’t testing enough. The latest guidelines decree that only symptomatic patients will be tested. We could therefore be missing nearly 80% of asymptomatic people who should actually have been part of the numerator or denominator. As a public health expert recently remarked on NDTV, the simplest way to flatten the curve is to stop testing! He also said that it’s good to look at mortality instead of considering the number infected.
So let’s consider the age-specific mortality among Indians to understand what is really happening in the country. The Telegraph reported that 49.5% of COVID-19 deaths in India are in those younger than 60 years – a ratio of around 50:50 of those younger than 60 versus those over 60. If COVID-19 actually behaves in India as it does in other countries, this ratio would have to be 50:350. Ergo, there ought to be 300 more deaths in the 60+ age group taking the total number of deaths to be 400 – four-times the number being quoted now.
If indeed people have died, they will escape being part of the numerator or the denominator if their official causes of death don’t include a mention of COVID-19. Deaths of older people can be attributed to causes like diabetes, stroke or – quite commonly – cardiac arrest.
Cardiac arrest?! A cardiac arrest simply means the heart has arrested – stopped beating. And when the heart stops beating, a person is declared dead. So technically everyone dies of a cardiac arrest. I haven’t come across a situation where the heart is beating in a dead person. (The sole exception is brain-death: even though the heart is beating, the brain is dead, and this definition of death is used for organ donation.) A cardiac arrest is just the final event in the life of every individual. The cause of death can’t therefore be cardiac arrest. Instead, it has to be the event, situation or condition that led to the arrest.
Examples of such causes include cerebral stroke, septicaemia, multi-organ-failure or a myocardial infarction, a.k.a. a heart attack. A heart attack — not to be confused with cardiac arrest – is due to a sudden stoppage of blood in the coronary arteries supplying the heart’s muscles, and can result in a cardiac arrest. However, cardiac arrest itself can’t be the cause of death; it would be myocardial infarction in this case. We have often heard of individuals having succumbed to cardiac arrest. It should really have been “succumbed with cardiac arrest as the final event” – and in the relevant cases possibly due to COVID-19.
The missing COVID-19 deaths are probably hiding under comorbidities common among Indians aged over 60 years.
So together with the facts that RT-PCR tests typically have a sensitivity of 30-70%, that not all suspected cases are tested, that the definition of ‘cured from COVID-19’ has changed, that a COVID-19 patient can die before being tested, and that deaths data is not very well recorded, it’s possible that the death of many older patients is being attributed to a ‘non-COVID’ causes.
Whatever the cause of death, this theory assumes that more people have died – up to four-times the official number. The Wire reported on May 13, “While the Delhi government has claimed that only 68 people have died so far in the city from COVID-19, 225 cremations (at two designated crematoria), and 89 burials had been carried out in adherence with the protocol.” A similar situation seems to be playing out in Chennai.
When we are dealing with a deadly pandemic, it’s important to know the truth. Correct numbers, reliable data are the only way we can get a firm grip on this crisis. Mortality rate or case fatality rate can’t be used as indicators simply because either the numerator or the denominator isn’t accurate.
So I humbly suggest that we start recording ‘suspected COVID-19’ as cause of death (when the RT-PCR wasn’t performed or was negative but clinical or CT scan features indicate otherwise), and include this data in our statistics. We should also record a death as ‘COVID-19-related’ when COVID-19 has been designated as the antecedent cause of death. Doctors in all sectors can be trusted to give this data – only if they aren’t threatened with dire consequences for reporting it.
No statistic being doled out at the moment is reliable simply because the numbers used for these calculations are not accurate. Mortality might be a better indicator only if we include those suspected to have died of COVID-19 and those who were supposedly ‘cured’ of COVID-19 but died soon after – and all these from a robust database. We need transparent, accessible data or we run the risk of languishing as “rotten apples”.
As a cardiac surgeon and not an epidemiologist or a statistician, even I can make out that there is something amiss with the official numbers. You may say I am wrong, but being wrong is not a trait of cardiac surgeons – universally! 😀
Dr Sanjeeth Peter is the director and chief cardiac surgeon at the DDMM Heart Institute, Nadiad, Gujarat. He has over 25 years’ experience as a cardiac surgeon and is a critical care team member for COVID-19 patient care.