Counting the uncounted – India’s Covid-19 death toll: The Indian Express obtained latest registered death data from 8 states that point to a spike but limited to April & May and varying sharply across states.
Written by Amitabh Sinha
Pune | Updated: August 4, 2021, 9:49:15 am
How many people died of Covid during the brutal second wave in April and May?
That question is key to not only understanding the scale of the unprecedented health tragedy but also to framing a policy prescription for any future re-surge.
There’s a practical imperative too — by August 14, the Centre needs to get back to the Supreme Court on how it plans to pay compensation to the next of kin of those who lost their lives to Covid-19.
One ready answer is 1.69 lakh — the official Covid-19 death count for the two months reported by state governments to the Centre.
An investigation by The Indian Express team of reporters suggests that there is an undercount. But exactly how much may never be known — it can only be estimated and, that too, in at least a year given the way India counts its dead.
As part of the investigation, the newspaper approached several state governments of whom eight — these accounts for almost a third of all Covid deaths as of last week — provided records of deaths they have registered in April and May this year as per the Civil Registration System (CRS).
This data, so far unpublished, shows that the total number of “all-cause deaths” (number of deaths whatever be the cause, but excluding official Covid deaths) of these eight states is 2.04 times that of all-cause deaths for April-May 2019 (2019 has been chosen for comparison because it is a non-pandemic year).
This surge ranges from the lowest 1.23x (x being times) in Kerala to 2.92x in Madhya Pradesh. If the official Covid deaths are subtracted from these all-cause deaths, the surge multiplier comes down for all states: from 1.12x in Kerala to 2.86x in Madhya Pradesh. Cumulatively, for all states, it drops to 1.87x.
The investigation identified three patterns:
Pattern 1
The surge in deaths during the peak of the second wave is not uniform across the country. So, death numbers cannot be assumed to increase at the same rate across states — big and small.
The increase in the number of deaths varies significantly amongst states (see chart). Minus their official Covid death figures, the surge works out to these multipliers: 2.86x for Madhya Pradesh; 2.03x for Bihar; 1.21x for Jharkhand; 1.73x for Punjab; 2.44x for Haryana; 1.4x for Delhi; 1.37x for Karnataka and 1.12x for Kerala.
Pattern 2
Deaths spiked during the peak but did not rise as dramatically in the non-peak months (January-March 2021). So the high multiples during peak Covid months do not hold well for other months and, hence, cannot be extrapolated for the full year.
In most states, the surge in 2021 is restricted to just April and May. The death numbers in April-May 2021 show a surge ranging 1.23x to 3.12x in the eight states. In Punjab and Haryana though, the number of deaths during January-March this year is less than the number of deaths in January-March 2019.
Jharkhand is an exception: the death numbers are high across all five months.
Pattern 3
In states with relatively better reporting standards, the multiple is lower. Take for example, Kerala. Adjusting for official Covid-19 deaths, the total deaths in April-May 2021 is just 1.12x more than in April-May 2019. To a lesser extent, Jharkhand (1.21x), Karnataka (1.37x) and Delhi (1.4x) too.
This is corroborated by the 100% Level of Registration of deaths in Delhi, Karnataka and Kerala, as per the report on the Vital Statistics of India Based on the Civil Registration System, 2019. A 100% level means the state has registered all the estimated deaths; lower levels suggest inefficiencies in the system of registration. Jharkhand, with 84% is again an exception. For the other states, this level is much lower: Madhya Pradesh 78%, Punjab 88%, Bihar 89%, Haryana 90%.
For states that said they could not provide preliminary CRS data for deaths in April-May this year, The Indian Express looked at another data set for all-cause deaths: the Health Management Information System maintained by the National Health Mission under the Union Ministry of Health and Family Welfare.
In HMIS, set up in 2008, service delivery information from about 2 lakh health facilities across India (with dominant rural, public institution bias) is uploaded every month.
This shows that at the all-India level, the number of deaths in April-May 2021 at 8.31 lakh was 2.11 times that in April-May 2019.
In the eight states that account for 60% of all official Covid-19 deaths — West Bengal, Maharashtra, Tamil Nadu, Uttar Pradesh, Chhattisgarh, Andhra Pradesh, Gujarat, Rajasthan — the multiple is highest at 3.62x for Gujarat and lowest at 1.46x for West Bengal. Adjusting for official Covid-19 deaths, the multiple drops to 3.29x for Gujarat and 1.33x for West Bengal.
But HMIS data, too, needs to be read with caveats.
For instance, Bihar’s death count during April-May 2021 in HMIS is only 2,034, which is lower than even the official Covid-19 death count for the state at 3,587.
HMIS death count for April-May 2021 as a percentage of CRS death count for April-May swings between 2% for Bihar to 72% for Karnataka. This could be because its coverage is comprehensive for only public institutions and rural areas.