Will We Ever Know India’s Actual COVID Death Toll?

On 7 May, as focus remained on Operation Sindoor and its aftermath, another major event went largely unnoticed. The Office of the Registrar General of India released the long-overdue 2021 civil registration data on births and deaths, possibly offering critical insight into the actual human cost of the COVID-19 pandemic.According to this data, India recorded 76.41 lakh deaths in 2019, 81.16 lakh in 2020, and 102.25 lakh in 2021. A simple subtraction helps establish that the number of deaths recorded in 2020 was 4.75 lakh more than in 2019, and was over 21 lakh deaths more in 2021 than the year before.Many suggest this data indicates India’s true COVID-19 death toll could be up to seven times higher than the government's official figure of 5.33 lakh deaths to date.On 18 May, Congress General Secretary Jairam Ramesh highlighted this on his X account, stating, "While we always knew that the government had systematically underreported COVID-19 deaths, we now learn that there were an additional 20 lakh deaths reported across India in 2021 alone."But can this increase in all-cause mortality truly reveal India’s COVID-19 death toll?Let's break it down.'Don't Stonewall, Share the Granular Data,' Say Experts to Govt on COVID Deaths The Civil Registration System (CRS) records all registered births and deaths in India. It is complemented by the Sample Registration System (SRS), a survey-based method that estimates unregistered births and deaths to help address gaps caused by underreporting.According to the 2021 CRS report, deaths in India rose by about 6.22 percent from 2019 to 2020 and 25.98 percent from 2020 to 2021. From 2019 to 2021, the jump is staggering at 33.82 percent.Meanwhile, according to the 2021 SRS report, India’s death rate rose from 6.0 (deaths per 1,000 population) in 2019 to 7.5 in 2021. It’s also worth noting that this rate had been steadily declining in the years leading up to 2019. The SRS data also shows that death rate sharply rose in these states from 2019 to 2021:Gujarat: 5.6 (2019) to 8.5 (2021)Chhattisgarh: 7.3 (2019) to 10.1 (2021)Delhi: 3.2 (2019) to 5.4 (2021)Haryana: 5.9 (2019) to 8.2 (2021)Additionally, the 2021 SRS data reveals a sharp urban-rural divide, with rural areas recording more deaths during the pandemic years. The death rate in rural regions was 7.9 (deaths per 1,000 population) compared to 6.6 in urban areas.The Trap of the Data Gap: How Do We Know India’s COVID Vaccines Are Working?In epidemiology, excess deaths refer to the number of deaths above what’s normally expected during a specific time or within a group, based on past trends.By calculating a trend model based on registered deaths from previous years, we can roughly estimate the expected number of deaths for 2020 and 2021. Comparing these projections to actual figures helps us gauge excess deaths. One simple method is to use the average annual increase from 2012 to 2019 to project expected deaths for 2020 and 2021.As per The Quint's calculations, the expected number of deaths in India in 2020 and 2021 should have been around: Approx. 78.97 lakh in 2020Approx. 81.53 lakh in 2021Based on estimates compared with the latest CRS data, India recorded 2.19 lakh excess deaths in 2020, 20.72 lakh in 2021, and a combined total of 22.91 lakh excess deaths over both years — possibly even higher.Ariel Karlinsky, a statistician who co-authored the World Health Organization (WHO)’s 2022 report on excess mortality during the COVID-19 pandemic, tells The Quint, “I think it’s safe to say that excess deaths in India in 2020–2021 are at least 2 million (20 lakh).” Now, this analysis is based solely on registered deaths. For a more accurate picture, one would need to consider what fraction of total deaths were registered before and during the pandemic, and whether that changed over time.“The fraction of deaths registered varies widely by state, and estimates often conflict depending on the data source," says mathematician Murad Banaji, who has been tracking COVID-19 data since the early days and co-authored a 2022 paper on the uncertainty surrounding India’s COVID-19 death estimates due to gaps in mortality data and registration systems.He goes on to explain, "For instance, the SRS tends to overestimate registration levels compared to surveys like the NFHS — while also underestimating overall mortality. So, we treat SRS estimates with caution.” "The main message is that many excess deaths estimates in the media based on the new data will likely be underestimates, because they do not account for incomplete registration," he adds.“So, depending on adjustments for data completeness, population size, and missing deaths in states like Rajasthan, India's excess deaths in 2020-21 could be as high as 4 to 6 million (40 to 60 lakh),” says Karlinsky.This figure is in line with the WHO's estimates, which had projected 4.74 million (47.4 lakh) excess deaths in India during the same period.Now, the government’s official cumulative COVID-19 death toll stands at just 5.33 lakh. In light of the newly released data, this suggests that the true number of COVID-related deaths may have been significantly undercounted.The government argues that a high number of excess deaths doesn’t necessarily indicate a higher COVID-19 death toll. It maintains that these figures could be due to other factors.From Census to NFHS, the Void in India's National Data Is Now Bigger Than EverNot all excess deaths can indeed be directly linked to COVID-19. As the government has noted, factors like expanded population coverage, improved registration systems, and greater awareness may have led to more deaths being recorded. A government official, quoted by The Hindu, pointed to these reasons to explain the spike when questioned about the excess deaths in 2020 and 2021.There’s also no one definition of what constitutes a 'COVID death'. Classification and reporting systems varied across states, with inconsistencies in how COVID cases and related deaths were recorded.Moreover, COVID-19 can lead to fatal complications like heart attacks, respiratory failure, or strokes, which may have been recorded as the primary cause of death rather than COVID itself. Some patients may have also died from post-COVID complications long after the infection cleared, making attribution even more challenging. Not to mention, a significant number of deaths in India occur outside hospitals, making medical certification challenging even beyond the pandemic.Given these factors, it’s hard to say for certain that all excess deaths in 2020 and 2021 were caused by COVID-19. However, experts agree that the sharp rise in deaths during this period indicates an unusually high mortality rate, with a significant portion likely linked — directly or indirectly— to the pandemic.Experts also reject the idea that improved death reporting explains the spike between 2020 and 2021, especially since the increase represents a substantial jump. "Nearly all of the excess deaths in 2020 and 2021 will be due to COVID based on worldwide analyses. It’s hard to think that in the middle of a pandemic, Indian citizens started reporting non-COVID deaths far more than earlier—that just is not sensible," Prabhat Jha, founding director of University of Toronto’s Centre for Global Health Research and former senior scientist at the WHO, tells The Quint.In fact, some causes of deaths, such as traffic accidents and other infectious diseases, likely declined during the pandemic due to the lockdowns and social distancing."CRS coverage of deaths has been expanding. However, only by about 4 percent a year on average. So, nearly all of the excesses of 2020 and particularly of 2021 can be attributed to COVID."Prabhat Jha, former senior scientist at the WHOSome of the increase in reported deaths between 2015 and 2019 may be due to improved reporting rather than an actual rise in deaths. This could also apply to changes between 2019 and 2020. Bhramar Mukherjee, Senior Associate Dean for Public Health Data Science and Data Equity at Yale School of Public Health, argues that this does not explain the 26 percent jump from 2020 to 2021. She notes, "Death reporting in India has been over 90 percent complete for a few years now. So there is no explanation for why there will be more (and not less) reporting during 2021.""When the Government of India released data for 2020 (in 2022), it claimed that death registration had become '100 percent complete' — and that this registration improvement was responsible for the increase in CRS deaths from 2019 to 2020. But if they say registration was already complete in 2020, they can't now claim that the further increase in the death toll in 2021 was also due to better registration."Ariel KarlinskyCOVID-19 Vaccination Before Infection Prevented Deaths By 60%: ICMR StudyIndia’s COVID-19 death toll has been a source of controversy from the outset, marred by denial, inconsistencies, and major data gaps.In the absence of comprehensive and transparent data, independent researchers and international bodies have attempted to estimate the true human cost of the pandemic, but these figures will always remain calculated estimates.A 2022 report by the WHO projected that India saw 47.4 lakh excess deaths during 2020–21. The Indian government strongly pushed back against these findings, calling the WHO’s modelling approach “flawed,” “unscientific,” and “intended to show the country in a poor light.”Speaking to the Quint now, Jha, who's one of the authors of the WHO report, notes, "The Indian government sticking to 0.5 million (5.3 lakh) death totals was a technical and political misstep. Multiple studies, including ours, have estimated that India’s actual COVID-19 death toll may be five to seven times higher than the official count." Experts also point out that this gap isn’t unique to India. For many countries, the WHO’s estimated COVID-19 death toll is significantly higher than their officially reported numbers. While governments around the world reported a combined total of around 6 million (roughly 60 Lakh) COVID-19 deaths by the end of 2021, the WHO estimated the global excess death toll to be approximately 14.9 million (149 lakh) for 2020 and 2021."Attribution and classification of deaths due to COVID is a tricky issue," says Mukherjee, adding, "States and countries that do not accurately report deaths ultimately face setbacks in prevention, decision-making, and public health actions."Experts also emphasise that meaningful analysis requires a stronger system for data collection, recording, and transparent publishing."Not just COVID, getting more reliable data on both, all-cause-mortality and cause-specific mortality is needed to understand what is killing people, this data should be reliably collected at a local level to guide where prevention and treatment resources should be allocated."Bhramar Mukherjee, Senior Associate Dean for Public Health Data Science and Data Equity, Yale School of Public Health"These are not just statistics for debate — millions of Indians have lost a loved one and much of that could have been averted with more science and less hubris," says Prabhat Jha.Jha and Karlinsky are part of a team of researchers currently working on a fresh analysis incorporating the newly released CRS and SRS data to estimate the true human cost of the pandemic in India.5 Years of COVID-19: Those Who Died Aren’t Merely Numbers

BECOME A MEMBER

Will We Ever Know India’s Actual COVID Death Toll?

Latest data suggests India’s actual COVID-19 death toll could be up to seven times higher than official figures.