Measuring excess deaths is the best possible way to estimate the count of COVID-19 deaths
The real time mortality impact of the COVID-19 pandemic is an important statistical measure to guide policy responses. But measuring the actual count is not an easy task. WHO, in January 2021, had estimated, based on excess deaths data in Europe and the American continents, that actual deaths were at least 1.6 times over the official count. The problem of under-counting, even in mature public health systems across the developed world, is largely because patients who die due to cardiovascular issues among others even after apparent recovery from COVID-19 are sometimes not tracked and registered as COVID-19-related deaths. This is why even in Kerala — with 100% registration of deaths and a relatively low case fatality rate — following criticisms about the methodology to evaluate whether a death was related to COVID-19, the health administration in the districts, rather than a State-level audit committee, will now audit deaths. But there is another class of under-counting across States, where health bulletins mislead by reporting a lower number of cases and deaths. This is the case with Bihar where the reported toll was suddenly increased by 72% following a Health Department review after the Patna High Court found discrepancies in figures cited by different agencies in Buxar district. Bihar is among the States in India with the lowest civil registration of deaths, with barely 34.1% of the dead being registered, according to the Civil Registration System (CRS) report of 2018. Estimations of the actual count of the dead are difficult to obtain in other States such as Uttar Pradesh as well, where public health systems are poor and neither the infections nor deaths have been effectively tracked, especially in rural areas, where many have died outside of hospitals.
One method to assess the actual number of deaths due to COVID-19 is by calculating the excess deaths during the given period when the pandemic has raged, compared to the baseline mortality occurring in similar time frames before the pandemic. This exercise also works best if death registrations are relatively high, which should be possible in most districts as registration of deaths has improved to 76% according to CRS 2018. Excess deaths analyses in Gujarat, Chennai and Kolkata based on collations of preliminary registration data by news organisations suggest that they were nearly 10, five and seven times higher, respectively, than reported fatalities during the second wave. If the CRS datasets, maintained by the Office of the Registrar General and Census Commissioner of India besides State registrars and municipal officials with a good quality of registration, are made available, it would enable better estimation of the actual mortality figures. In the meantime, only honest reporting of the deaths will help provide better mitigation strategies.