Officially, governments around the world have reported over seven million COVID-19 deaths since the pandemic began in early 2020. But that number captures only a fraction of the real toll. The most rigorous estimates, including one published by the World Health Organization, put global excess deaths associated with the pandemic at roughly 14.8 million through the end of 2021 alone, nearly three times the official figure for that period. Understanding why such a large gap exists, where the deaths were concentrated, and who bore the greatest burden reveals a pandemic far deadlier than the headline numbers suggest.
Why Official Counts Fall Short
Every country tracks COVID-19 deaths differently, and those differences create enormous blind spots. Some nations count only deaths in people who tested positive; others include clinically diagnosed cases without a lab confirmation. The timing of the test matters too: a person who dies days after their last positive result may or may not appear in the statistics, depending on local rules. And in many low-income countries, large portions of the population die without ever seeing a doctor or receiving a test, meaning their deaths never enter the COVID-19 column at all.
Researchers have catalogued the specific factors that widen or narrow this measurement gap. The ratio between officially attributed COVID-19 deaths and “true” COVID-19 deaths depends on how testing is conducted near the time of death, how broadly the diagnostic criteria are applied, the overall infection rate in the population, and whether the epidemic wave coincides with a period of already-elevated mortality from other causes.
1PubMed Central. Over- and under-estimation of COVID-19 deathsVaried definitions of what counts as a COVID-19 death, uncertainty about how many people were actually infected, and biased testing (which tends to oversample the sickest patients) all compound the problem.
2PubMed Central. Using excess deaths and testing statistics to determine COVID-19 mortalitiesExcess Mortality and the True Scale
Because official COVID-19 death counts are unreliable in many settings, epidemiologists turn to a broader measure: excess mortality. The concept is straightforward. You look at how many people died in a given period, compare that to how many would normally have died based on recent years, and the difference is the “excess.” Not all excess deaths are necessarily from COVID-19 itself; some reflect healthcare systems buckling under pandemic pressure. But excess mortality is widely considered the most honest single number for capturing a pandemic’s full impact.
The WHO commissioned a detailed statistical analysis covering January 2020 through December 2021. That study estimated 14.83 million excess deaths globally, at a time when only 5.42 million deaths had been officially attributed to COVID-19. The true toll, in other words, was about 2.74 times higher than the reported figure.
3PubMed Central. The WHO estimates of excess mortality associated with the COVID-19 pandemicA separate analysis published in The Lancet used a different modeling approach, building predictions for places where all-cause mortality data were unavailable and selecting covariates through a statistical technique that identifies the most informative predictors.
4The Lancet. Estimating excess mortality due to the COVID-19 pandemic: a systematic analysis of COVID-19-related mortality, 2020–21 That study found the highest excess mortality rate globally was in Bolivia, at roughly 735 deaths per 100,000 people, compared to a global average of about 120 per 100,000.5The Lancet. Estimating excess mortality from the COVID-19 pandemic in 191 countries and territories and 252 subnational units for selected countries from Jan 1, 2020, to Dec 31, 2021
These models have limitations. For countries with poor civil registration systems, the estimates rely on statistical proxies rather than direct death counts. Researchers have used creative data sources in these settings, including funeral counts, cemetery records, burial-site satellite imaging, and demographic surveillance systems.
6PubMed Central. Excess mortality during the COVID-19 pandemic in low-and lower-middle-income countries: a systematic review and meta-analysis Still, the direction of the finding is not in doubt: official counts dramatically understate the pandemic’s death toll, and the undercount is worst in the countries with the weakest health data infrastructure.
Where the Reporting Gap Was Widest
India stands out as perhaps the most striking example of the gap between official and estimated deaths. A nationally representative survey published in Science estimated that COVID-19 accounted for about 29% of all deaths in India between June 2020 and July 2021, translating to roughly 3.2 million COVID deaths in that period. India’s cumulative toll by September 2021 was estimated at six to seven times higher than the government’s official figure.
7PubMed Central. COVID mortality in India: National survey data and health facility deaths A separate review of multiple Indian studies confirmed that excess mortality ran at roughly three times the official count across the board, with model-based estimates generally running higher than those drawn from civil registration data.
8PubMed Central. Systematic Review of Excess Mortality in India during the Covid-19 Pandemic with Differentiation between Model-Based and Data-Based Mortality EstimatesThe Delta wave in mid-2021 was especially devastating in rural India. An analysis of health facility data in eight Indian states found a 270% increase in deaths among people aged ten and older during April through June 2021 compared to the pre-pandemic baseline. Extrapolating those findings nationally yielded an estimate of about 2.6 million excess rural deaths in that three-month window alone.
9BMJ Open. Cause-specific excess mortality in rural India during the COVID-19 pandemic 2020–2023: longitudinal analyses of deaths in 0.2 million rural health facilitiesLatin America was hit hard as well. Peru reported the highest per-capita mortality in the region, exceeding 600 deaths per 100,000 inhabitants, with Brazil second at over 290 per 100,000.
10PubMed Central. COVID‐19 in Latin America and the Caribbean: Two years of the pandemic Across Mexico, Peru, Brazil, Chile, and Ecuador, researchers found that as the pandemic progressed, excess mortality became more visible in areas with poorer living conditions, and that life expectancy in some of these countries dropped by two to ten years.
11PubMed Central. Investigating regional excess mortality during 2020 COVID-19 pandemic in selected Latin American countriesAt the regional level, the highest excess mortality rates during 2020-2021 clustered in Andean Latin America, eastern Europe, central Europe, and southern sub-Saharan Africa. But elevated rates also appeared in scattered locations outside those regions, including Lebanon, Armenia, Tunisia, parts of Italy, and several southern U.S. states.
5The Lancet. Estimating excess mortality from the COVID-19 pandemic in 191 countries and territories and 252 subnational units for selected countries from Jan 1, 2020, to Dec 31, 2021Who Was Most Vulnerable
Age was by far the strongest predictor of dying from COVID-19. The relationship between age and the infection fatality rate is exponential: a systematic analysis found that the lowest fatality risk occurred around age seven, at roughly two thousandths of a percent, then climbed steeply through middle age to about 1% at age 60 and over 20% at age 90.
12THE LANCET. Variation in the COVID-19 infection-fatality ratio by age, time, and geography during the pre-vaccine era: a systematic analysis Another meta-analysis confirmed this pattern, estimating the fatality rate at about 0.4% for 55-year-olds, 1.4% at 65, 4.6% at 75, and 15% at 85.
13PubMed Central. Assessing the age specificity of infection fatality rates for COVID-19: systematic review, meta-analysis, and public policy implications For children and young adults, the risk of death from infection was vanishingly small, with a median fatality rate of 0.0003% for those under 20.
14PubMed Central. Age-stratified infection fatality rate of COVID-19 in the non-elderly populationPre-existing health conditions amplified the risk considerably. A meta-analysis found that people with cardiovascular disease had the highest odds of a critical or fatal outcome, at roughly 3.4 times the risk compared to those without it. Chronic lung disease, high blood pressure, and diabetes each roughly doubled or tripled the risk as well.
15PubMed Central. Impact of metabolic and cardiovascular disease on COVID-19 mortality: A systematic review and meta-analysis A separate systematic review reinforced these findings, additionally identifying chronic kidney disease as carrying a particularly high mortality risk at about 3.25 times the baseline.
16PLOS ONE. Association of cardiovascular disease and 10 other pre-existing comorbidities with COVID-19 mortality: A systematic review and meta-analysisHow Variants Shifted the Death Toll Over Time
The virus did not stay the same throughout the pandemic, and the shift from Delta to Omicron in late 2021 marked a turning point for severity. In England, Omicron infections carried about a 66% lower risk of death compared to Delta infections after adjusting for age, sex, vaccination status, prior infection, and a long list of other factors.
17BMJ. Risk of covid-19 related deaths for SARS-CoV-2 omicron (B.1.1.529) compared with delta (B.1.617.2): retrospective cohort study Hospital admission risk was similarly lower: the adjusted hazard ratio for admission with Omicron versus Delta was about 0.41, meaning Omicron patients had less than half the hospitalization risk.
18The Lancet. Comparative hospitalisation and mortality risk for the omicron and delta variant in EnglandSome of this reduced severity was clearly due to rising population immunity from prior infections and vaccination. But evidence from laboratory studies suggests that Omicron was also intrinsically less dangerous, in part because it preferentially infected the upper airways rather than penetrating deep into the lungs the way Delta did.
19PubMed Central. Severity and outcomes of Omicron variant of SARS-CoV-2 compared to Delta variant and severity of Omicron sublineages: a systematic review and metanalysis The combination of a milder variant and wider immunity meant that while Omicron infected enormous numbers of people, it killed a much smaller fraction of them than earlier waves had.
Lives Saved by Vaccination
How many deaths vaccines prevented depends heavily on the modeling assumptions, and two major estimates illustrate the range. A study in The Lancet Infectious Diseases estimated that in just the first year of vaccine rollout (December 2020 through December 2021), vaccinations prevented about 14.4 million deaths based on official COVID-19 counts, and roughly 19.8 million when using excess mortality as the baseline, representing a 63% reduction in the total deaths that would otherwise have occurred.
20The Lancet Infectious Diseases. Global impact of the first year of COVID-19 vaccination: a mathematical modelling study A more recent analysis in JAMA Health Forum, using a different methodology covering the full 2020-2024 period, arrived at a lower estimate of about 2.5 million deaths averted globally, with a sensitivity range of 1.4 to 4.0 million.
21JAMA Health Forum. Global Estimates of Lives and Life-Years Saved by COVID-19 Vaccination During 2020-2024These two figures are not as contradictory as they look. The Lancet estimate asks a counterfactual question: how many people would have died if vaccines had never been developed? The JAMA estimate uses a comparative effectiveness approach that accounts for real-world rollout timing and the interplay between vaccination and evolving natural immunity. Both agree that vaccines substantially reduced deaths; they disagree on how to measure the baseline scenario.
Non-pharmaceutical interventions mattered too, especially early on. A study of 130 countries found that earlier and stricter school and workplace closures during the first wave were associated with lower COVID-19 death rates, though other interventions did not reach statistical significance across countries.
22PubMed Central. The relative effects of non-pharmaceutical interventions on wave one Covid-19 mortality: natural experiment in 130 countries However, prolonged strict lockdowns carried their own costs. A study of 22 European countries found that the most intensive and prolonged restrictions were associated with a 44% increase in excess mortality, an effect that vaccination partially offset.
23PubMed Central. The association of intensity and duration of non-pharmacological interventions and implementation of vaccination with COVID-19 infection, death, and excess mortality: Natural experiment in 22 European countries That finding likely reflects the indirect toll of healthcare disruption, economic hardship, and delayed medical care that accompanied extended shutdowns.
Deaths Beyond the Virus Itself
Not every pandemic-era death was caused by the virus directly. When hospitals filled with COVID-19 patients, care for everything else suffered. An analysis of England’s National Health Service found that for roughly every 42 COVID-related deaths in the population, there was at least one additional excess death among patients who came to the hospital for completely unrelated reasons. The mechanism was straightforward: pandemic pressures reduced the system’s ability to deliver timely, high-quality care to non-COVID patients.
24PubMed. The hidden toll of the pandemic: Excess mortality in non-COVID-19 hospital patientsCancer screening delays offer a concrete example of how indirect effects ripple outward. A study of U.S. adults found that about a quarter delayed mammograms and Pap tests during the pandemic, more than a quarter delayed HPV testing, and over a third delayed colonoscopies.
25PubMed Central. Impact of the COVID-19 Pandemic on Cancer Screening Delays Delayed diagnoses mean cancers caught at later stages, which translates into worse survival rates. The full mortality impact of these delays will take years to fully quantify, but they are part of the pandemic’s broader death toll that official COVID-19 counts never capture.
Income Inequality and the Distribution of Death
COVID-19 did not kill equally across the income spectrum. A review of 95 studies found that 91% of them reported significantly higher COVID-19 death rates in socially disadvantaged areas compared to affluent ones.
26PubMed Central. COVID-19 mortality and deprivation: pandemic, syndemic, and endemic health inequalities This pattern held both within and between countries. At the national level, each percentage-point increase in a country’s income inequality index was associated with a roughly 14% increase in the hazard of higher COVID-19 mortality in a survival analysis across 74 countries.
27Preventive Medicine Reports. Income inequality and the disease burden of COVID-19: Survival analysis of data from 74 countriesThe reasons are layered. People in lower-income settings are more likely to work in jobs that cannot be done remotely, to live in crowded housing, to rely on public transportation, and to have less access to healthcare. They also carry a higher burden of the chronic diseases that amplify COVID-19 severity. The pandemic did not create these inequalities, but it exposed and deepened them, concentrating death among those with the fewest resources to protect themselves.
The Long Shadow on Heart Health
The pandemic’s health toll extends beyond the acute infection. Research using the UK Biobank found that COVID-19 patients faced increased risks of cardiovascular disease and death not just during the infection but for up to a year afterward, with the elevated risk linked to long COVID.
28Cardiovascular Research. Association of COVID-19 with short- and long-term risk of cardiovascular disease and mortality: a prospective cohort in UK Biobank A large U.S. Veterans Affairs study confirmed and expanded this picture, showing increased risk across a range of cardiovascular problems including stroke, heart failure, irregular heartbeat, and blood clots in the year following infection, even among people who were never hospitalized.
29Nature Medicine. Long-term cardiovascular outcomes of COVID-19For people with diabetes, the long-term risks were especially pronounced. Hospitalized diabetic patients who had COVID-19 faced roughly double the risk of developing coronary heart disease and stroke compared to similar patients without COVID-19, and a 37% higher risk of death from any cause among middle-aged adults.
30PubMed Central. Association between COVID-19 and incidence of cardiovascular disease and all-cause mortality among patients with diabetes These downstream cardiovascular deaths do not appear in COVID-19 mortality statistics, but they represent a real and ongoing cost of the pandemic that will unfold over years.
Millions of Children Lost a Parent or Caregiver
Behind every death statistic is a family, and one of the pandemic’s most devastating legacies is the number of children left without a parent. By the end of October 2021, an estimated 5.2 million children worldwide had experienced the death of a parent or caregiver due to COVID-19, a 90% increase from six months earlier.
31PubMed Central. Global, regional, and national minimum estimates of children affected by COVID-19-associated orphanhood and caregiver death, by age and family circumstance up to Oct 31, 2021: an updated modelling study More than three-quarters of these children lost their father, reflecting the higher fatality rate among men throughout the pandemic.
By May 2022, when researchers updated their estimates using WHO excess mortality data, the figure had climbed to about 7.5 million children orphaned and 10.5 million who lost either a parent or another caregiver. The burden fell disproportionately on Southeast Asia and Africa, which together accounted for roughly 65% of orphaned children. India alone accounted for an estimated 3.5 million.
32JAMA Pediatrics. Orphanhood and Caregiver Loss Among Children Based on New Global Excess COVID-19 Death Estimates Peru had the highest rate relative to its child population, at about 10 per 1,000 children, followed by South Africa at about 5 per 1,000.
33PubMed Central. Global minimum estimates of children affected by COVID-19-associated orphanhood and deaths of caregivers: a modelling studyHow COVID-19 Compares to the 1918 Flu
The comparison people reach for most often is the 1918 influenza pandemic, and the answer depends on how you measure. In absolute numbers, COVID-19 killed more people in many countries simply because populations are so much larger now. In the Netherlands, for instance, about 50,000 people died of COVID-19 compared to roughly 32,000 from the 1918 flu. But after adjusting for population size and age structure, the 1918 flu was substantially deadlier: age-standardized mortality rates showed the Spanish flu would have caused about seven times as many deaths per capita as COVID-19 if the two pandemics had struck the same population.
34PubMed. Mortality Rates of the Spanish Flu and Coronavirus Disease 2019 in the Netherlands: A Historical ComparisonAn Italian comparison reached the same conclusion. The 1918 pandemic was “certainly more deadly” on a rate basis, though COVID-19’s pre-vaccine death rate in Italy was higher than some previously published estimates for other European countries had suggested.
35The Journal of Infectious Diseases. Historical Comparison Between the Death Rate for Spanish Flu and Coronavirus Disease 2019 in Italy The 1918 flu also killed a very different demographic profile: young adults in their twenties and thirties were disproportionately affected, while COVID-19’s toll was overwhelmingly concentrated among the elderly. That age difference means the 1918 pandemic likely destroyed more total years of life per death than COVID-19 did, even where COVID-19’s absolute death counts are higher.
Context matters here. Medicine in 1918 had no ventilators, no antivirals, no antibiotics for secondary bacterial infections, and no vaccines. COVID-19 arrived in an era of rapid vaccine development, intensive care technology, and global surveillance. The fact that the virus still killed millions despite these advantages is a reminder of how dangerous a novel respiratory pathogen remains, even with modern tools.