Background An unbiased estimate of the COVID-19 deaths is essential to compare and contrast the public health actions adopted by governments from time to time. It also helps them in their policy-making in the future. But unfortunately, there is gross underreporting of COVID-19 deaths across the world, and also from India, as reported by many researchers, which can be estimated by modelling exercises only. Methods In this article, we estimated the number of COVID-19 deaths among the Indian states using a time series analysis approach with the mortality data published by the civil registration system since 2011. We excluded the trend in non-COVID mortality due to the prevailing causes to prevent it from being counted erroneously as due to COVID-19. Results Accordingly, the number of COVID-19 deaths in India is estimated as 25.8 lakh (95% confidence interval [CI]: 11.1L, 35.4L), with Bihar being the largest under-reporter of COVID-19 mortality (98.4% [95% CI: 98.0, 98.7] deaths not reported). On the contrary, there are states with zero underreporting (Kerala, Goa, Mizoram, Sikkim, and all the union territories). We also found out that the measures of social and economic development, like the multidimensional poverty index, infant mortality, literacy rate, per capita state domestic product, and the NITI Aayog Health Index, are significantly correlated to the proportion of underreporting. Discussion and Conclusion The underreporting of COVID-19 deaths in India is very much linked to its social and economic development. There is intentional nonreporting of the deaths of children, women, and the poor for a much longer period of time. Poor quality of data-grabbing mechanisms, the social stigma associated with the disease, inadequate testing of the population, the inefficiency of the administration, and omission of the socially and economically vulnerable groups had intensified the scenario.
Read full abstract