Student or educator? Sign up with your school email and get 6 months of Pro free.
CoChat

The DOI matches its Crossref record; 19 of the 19 references it cites were found in a registry or an index; 9 of the 14 sources its claims cite were checked for support.

The WHO estimates of excess mortality associated with the COVID-19 pandemic

Open access

Published version of a preprint

Abstract

. Reported statistics on COVID-19 mortality are problematic for many countries owing to variations in testing access, differential diagnostic capacity and inconsistent certification of COVID-19 as cause of death. Beyond what is directly attributable to it, the pandemic has caused extensive collateral damage that has led to losses of lives and livelihoods. Here we report a comprehensive and consistent measurement of the impact of the COVID-19 pandemic by estimating excess deaths, by month, for 2020 and 2021. We predict the pandemic period all-cause deaths in locations lacking complete reported data using an overdispersed Poisson count framework that applies Bayesian inference techniques to quantify uncertainty. We estimate 14.83 million excess deaths globally, 2.74 times more deaths than the 5.42 million reported as due to COVID-19 for the period. There are wide variations in the excess death estimates across the six World Health Organization regions. We describe the data and methods used to generate these estimates and highlight the need for better reporting where gaps persist. We discuss various summary measures, and the hazards of ranking countries' epidemic responses.

Overview

CoChat Summary

TL;DR

WHO estimated global excess mortality from COVID-19 for January 2020–December 2021 at 14.83 million deaths, using a Bayesian Poisson framework to fill gaps where all-cause mortality data were incomplete. The estimate is 2.74 times the 5.42 million COVID-19 deaths reported to WHO for the same period.

Key findings

  • Globally, the paper estimates 14.83 million excess deaths over 2020–2021 (95% UI 13.23–16.58 million), compared with 5.42 million reported COVID-19 deaths; the excess-to-reported ratio is 2.74 (UI 2.44–3.06).
  • Excess deaths were estimated at 4.47 million in 2020 and 10.36 million in 2021 globally; the global P-score increased from 7.97% in 2020 to 18.30% in 2021.
  • The paper reports large country-level totals, including India: 4.74 million excess deaths (UI 3.31–6.45 million), Russian Federation: 1.07 million (UI 1.05–1.10 million), Indonesia: 1.03 million (UI 0.75–1.29 million), and the United States: 932,000 (UI 887,000–978,000).
  • The 20 countries with the highest excess estimates account for 48.9% of the global population and over 80% of estimated global excess deaths for January 2020–December 2021.

Methods

  • The study used a WHO-led, consultative modelling process: a Technical Advisory Group developed methods, WHO member states reviewed preliminary estimates, and feedback was used to update the final estimates.
  • The analysis used all-cause mortality (ACM) data from multiple sources; monthly national data were available for 100 countries (52%), while other locations had annual, subnational, or no data.
  • Excess deaths were estimated with a Bayesian Poisson / negative binomial framework: historical ACM (often 2015–2019) was used to model expected deaths with spline-based seasonal and annual trends; where pandemic data were incomplete, a covariate-based log-linear model and, for some countries, subnational-to-national multinomial/proportionality models were used.

Limitations

  • The paper states that the biggest limitation is data paucity: for almost half of WHO member states, ACM data during the pandemic were unavailable, so estimates relied on statistical models; the text says these countries should be interpreted with caution.
  • The paper states that the counterfactual trend is sensitive to modelling assumptions, including the choice of spline models and how recent historical years are weighted; it notes splines can produce inappropriate extrapolations.
  • The text makes evident that comparisons across countries/regions are difficult because countries differ in age structure, data completeness, and reporting quality; this is stated in the paper and means rankings can be misleading.
  • The text makes evident that the paper cannot quantify how much excess mortality is directly COVID-19 versus indirect or non-COVID causes; the discussion explicitly says excess mortality combines these and cannot separate them.

Model-written summary. Check it against the paper before relying on it.

Integrity signals

Registry

Crossref · 10.1038/s41586-022-05522-2

Cites

19 of 19 references verified0 citation errors · 0 not checkable

Statistics

Nothing to recompute.

Figures

Fig. 1 Global excess and reported COVID-19 deaths and death rates per 100,000 population. a, Cumulative global excess death estimates and the cumulative reported COVID-19 deaths by month from January 2020 to December 2021. b, Global excess death rates per 100,000 population and the reported COVID-19 death rates per 100,000 population, also by month, from January 2020 to December 2021. On both plots, the central lines of the excess mortality series show the mean estimates and the shaded regions indicate the 95% uncertainty intervals. Source data
Fig. 1 Global excess and reported COVID-19 deaths and death rates per 100,000 population. a, Cumulative global excess death estimates and the cumulative reported COVID-19 deaths by month from January 2020 to December 2021. b, Global excess death rates per 100,000 population and the reported COVID-19 death rates per 100,000 population, also by month, from January 2020 to December 2021. On both plots, the central lines of the excess mortality series show the mean estimates and the shaded regions indicate the 95% uncertainty intervals. Source dataCC BY · PubMed Central
Fig. 2 Global and WHO region P-scores (excess deaths relative to expected deaths). Monthly estimates of P-scores, expressed as a percentage, aggregated globally and for the six WHO regions for the period January 2020 to December 2021. All plots show the mean estimates and the 95% uncertainty intervals. Source data
Fig. 2 Global and WHO region P-scores (excess deaths relative to expected deaths). Monthly estimates of P-scores, expressed as a percentage, aggregated globally and for the six WHO regions for the period January 2020 to December 2021. All plots show the mean estimates and the 95% uncertainty intervals. Source dataCC BY · PubMed Central
Fig. 3 The 25 countries with the highest total estimated excess deaths between January 2020 and December 2021. The red dots show the total reported COVID-19 death numbers. The purple dots show the mean total estimated excess death numbers with the width of the bars showing the 95% uncertainty intervals. Source data
Fig. 3 The 25 countries with the highest total estimated excess deaths between January 2020 and December 2021. The red dots show the total reported COVID-19 death numbers. The purple dots show the mean total estimated excess death numbers with the width of the bars showing the 95% uncertainty intervals. Source dataCC BY · PubMed Central
Fig. 4 Mapping estimated P-scores (excess deaths relative to expected deaths). The map shows the geographic distribution of the mean P-scores for years 2020 and 2021 across all 194 WHO member states. The darker the colour the higher the estimated mean P-score. The patterns indicate the quality of the all-cause mortality data that were available for each respective country with the solid pattern showing full or partial data, dots for mixed data and diagonal lines for no data. Source data
Fig. 4 Mapping estimated P-scores (excess deaths relative to expected deaths). The map shows the geographic distribution of the mean P-scores for years 2020 and 2021 across all 194 WHO member states. The darker the colour the higher the estimated mean P-score. The patterns indicate the quality of the all-cause mortality data that were available for each respective country with the solid pattern showing full or partial data, dots for mixed data and diagonal lines for no data. Source dataCC BY · PubMed Central
Fig. 5 The 25 countries with the highest mean P-scores (excess deaths relative to expected deaths). The plot shows the 25 countries with the highest mean P-scores for years 2020 and 2021 after ranking all WHO member states with populations greater than 200,000 by mean P-score from highest to lowest. The mean for each country is shown using the central grey dots and the widths of the bars show the 95% uncertainty intervals. Source data
Fig. 5 The 25 countries with the highest mean P-scores (excess deaths relative to expected deaths). The plot shows the 25 countries with the highest mean P-scores for years 2020 and 2021 after ranking all WHO member states with populations greater than 200,000 by mean P-score from highest to lowest. The mean for each country is shown using the central grey dots and the widths of the bars show the 95% uncertainty intervals. Source dataCC BY · PubMed Central
Fig. 6 Mapping the ratio of total excess deaths to total reported COVID-19 deaths. The map shows the geographic distribution of the mean ratio of the total excess deaths to total reported COVID-19 deaths for years 2020 and 2021 across all 194 WHO member states. The darker the colour the higher the estimated mean ratio. The patterns indicate the quality of the all-cause mortality data that were available for each respective country with the solid pattern showing full or partial data, dots for mixed data and diagonal lines for no data. Source data
Fig. 6 Mapping the ratio of total excess deaths to total reported COVID-19 deaths. The map shows the geographic distribution of the mean ratio of the total excess deaths to total reported COVID-19 deaths for years 2020 and 2021 across all 194 WHO member states. The darker the colour the higher the estimated mean ratio. The patterns indicate the quality of the all-cause mortality data that were available for each respective country with the solid pattern showing full or partial data, dots for mixed data and diagonal lines for no data. Source dataCC BY · PubMed Central
Extended Data Fig. 1 Mapping the availability of all-cause mortality data. The countries in dark blue have all 24 months of data available for January 2020 to December 2021 whereas those in purple have monthly data available but for less than 24 months. For the countries in green we only have either subnational or annual data for some or all of the period and for those in yellow do not have any representative all-cause mortality data available for the pandemic period.
Extended Data Fig. 1 Mapping the availability of all-cause mortality data. The countries in dark blue have all 24 months of data available for January 2020 to December 2021 whereas those in purple have monthly data available but for less than 24 months. For the countries in green we only have either subnational or annual data for some or all of the period and for those in yellow do not have any representative all-cause mortality data available for the pandemic period.CC BY · PubMed Central
Extended Data Fig. 2 Overview of modelling strategy to produce excess mortality estimates for all countries. The flowchart outlines how the excess mortality associated with the COVID-19 pandemic is estimated. By definition, this is the difference between the all-cause mortality (ACM) during the pandemic period and that which was expected had the pandemic not occurred (light-green). Starting with the historic monthly ACM data (light-blue) we apply the Negative Binomial Spline model to generate the expected deaths for each country. The data that are available inform the modelling strategy employed to calculate the pandemic period ACM (light-pink). For the places with full data, the reported ACM are taken as is. For those countries with only subnational ACM data, a Multinomial Subnational model is employed to derive national level ACM based on the historic fractions of deaths observed in subnational regions. The overdispersed Poisson Covariate model is fitted to countries with monthly pandemic data, and this model is used to estimate pandemic period ACM for countries without any reported ACM. The covariate model is also used to infer the within-year monthly pandemic ACM in countries with only annual data, via the Multinomial Covariate model.
Extended Data Fig. 2 Overview of modelling strategy to produce excess mortality estimates for all countries. The flowchart outlines how the excess mortality associated with the COVID-19 pandemic is estimated. By definition, this is the difference between the all-cause mortality (ACM) during the pandemic period and that which was expected had the pandemic not occurred (light-green). Starting with the historic monthly ACM data (light-blue) we apply the Negative Binomial Spline model to generate the expected deaths for each country. The data that are available inform the modelling strategy employed to calculate the pandemic period ACM (light-pink). For the places with full data, the reported ACM are taken as is. For those countries with only subnational ACM data, a Multinomial Subnational model is employed to derive national level ACM based on the historic fractions of deaths observed in subnational regions. The overdispersed Poisson Covariate model is fitted to countries with monthly pandemic data, and this model is used to estimate pandemic period ACM for countries without any reported ACM. The covariate model is also used to infer the within-year monthly pandemic ACM in countries with only annual data, via the Multinomial Covariate model.CC BY · PubMed Central
Extended Data Fig. 3 Excess mortality densities and ranking probabilities for select countries. Left: Ridgeplots representing the uncertainty in the cumulative excess monthly mortality rate over January 2020–December 2021 for six European countries. Right: Bivariate plots of pairs of excess rates (lower triangular), 1-dimensional summaries for individual countries (diagonal), and probabilities that the excess for the country labelled on the left exceeds the rate for the country labelled at the top (upper triangular). These probabilities are the fraction of points that lie below the red line in the corresponding bivariate plot.
Extended Data Fig. 3 Excess mortality densities and ranking probabilities for select countries. Left: Ridgeplots representing the uncertainty in the cumulative excess monthly mortality rate over January 2020–December 2021 for six European countries. Right: Bivariate plots of pairs of excess rates (lower triangular), 1-dimensional summaries for individual countries (diagonal), and probabilities that the excess for the country labelled on the left exceeds the rate for the country labelled at the top (upper triangular). These probabilities are the fraction of points that lie below the red line in the corresponding bivariate plot.CC BY · PubMed Central
Extended Data Fig. 4 Excess rates over time and ranking probabilities for select countries. : Left: excess monthly mortality rate (per 100,000 of population), with 95% uncertainty intervals, by month, for six European countries. The panels to the right display the ranking probabilities for each of the individual countries. The probabilities sum to one for each country within each month, i.e., a country has to be in one of the six ranking positions. Also for a fixed month and ranking position across all countries the probabilities sum to one because one of the countries has to be in rank 1, 2, etc, in each month.
Extended Data Fig. 4 Excess rates over time and ranking probabilities for select countries. : Left: excess monthly mortality rate (per 100,000 of population), with 95% uncertainty intervals, by month, for six European countries. The panels to the right display the ranking probabilities for each of the individual countries. The probabilities sum to one for each country within each month, i.e., a country has to be in one of the six ranking positions. Also for a fixed month and ranking position across all countries the probabilities sum to one because one of the countries has to be in rank 1, 2, etc, in each month.CC BY · PubMed Central

Verification

Claims the paper makes

8 Valid · 2 Not valid · 6 Not checked · 81 Own results and interpretations

49 more claims

Create a free account to see all 97 claims and every source verdict.

See all claims — free account
  • Probably not a substantial claim. 19
  • It may state two separate claims as one. 30
  • The quote only partly states this claim. 2

Create a free account to see these claims.

Claims other papers make about this one

Each is checked against this paper’s own text: does it say what the citing paper says it does?

Citations

Cites

Show all 17 references

Research works printed but not linked: 2

  1. National Bureau of Statistics of China. China Statistical Yearbook 2021 (China Statistics Press, 2021).
  2. Hale, T. et al. Variation in Government Responses to COVID-19 Technical Report (Univ. Oxford, 2020).

Cited by

See all 1,067

Metadata

Authors
Msemburi, W., Karlinsky, A., Knutson, V., Aleshin-Guendel, S., Chatterji, S., & Wakefield, J.
Affiliations
  • World Health Organization (CH)
  • Hebrew University of Jerusalem (IL)
  • University of Washington (US)
Corresponding author
Msemburi, William
Venue
Nature
Published
2022-12-14
Volume
613
Issue
7942
Pages
130-137
ISSN
  • 0028-0836 (ISSN-L)
  • 1476-4687
OpenAlex
W4311472757
Cited by
1,064
References
17
Citations by year
2026: 176 · 2025: 283 · 2024: 346 · 2023: 252 · 2022: 6
FWCI
131.36
Type
journal article
Language
en
Open access
Licence
Locations
  • Nature (publishedVersion, cc-by)
  • Research Square (acceptedVersion, cc-by)
  • PubMed Central (publishedVersion, cc-by)
  • PubMed (publishedVersion)
  • RePEc: Research Papers in Economics (submittedVersion)
Funders
World Health Organization
Topics
  • Physical Sciences › Mathematics › Modeling and Simulation › COVID-19 epidemiological studies
  • Health Sciences › Medicine › Epidemiology › Data-Driven Disease Surveillance
  • Social Sciences › Demography › Insurance, Mortality, Demography, Risk Management
Keywords
COVID-19 pandemic, excess mortality, all-cause mortality, underreporting of deaths, cause of death certification, Bayesian uncertainty quantification, WHO regions
MeSH terms
COVID-19, Bayes Theorem, Humans, World Health Organization, Poisson Distribution, Uncertainty, Pandemics

About these checks

Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.

The claim verdicts this report counts are experimental: their accuracy is not yet measured.

A language model drew the claims from the paper on 2026-10-10, under version 3 of the claim definition (model deepseek-v4.1-flash, prompt extract-v3).

A second model labelled each claim on 2026-10-10 (model jev-1.13-20260917, definition revision gold-guideline-v0.2).

Cited sources are checked with a language model’s help (gpt-6-sol and qwen3.8-omni-flash).

Statistics and arithmetic are recomputed, not judged.

Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.

What counts as a claim ›How we measure confidence ›How we check this ›

Are you an author, or do you see a problem with this page? Request a review

How to cite

Cite CoChat for the checks on this page, and the paper for its own findings.

This page
CoChat. (2026). Citation check: The WHO estimates of excess mortality associated with the COVID-19 pandemic. https://app.cochat.ai/papers/10.1038/s41586-022-05522-2
The paper
Msemburi, W., Karlinsky, A., Knutson, V., Aleshin-Guendel, S., Chatterji, S., & Wakefield, J. (2023). The WHO estimates of excess mortality associated with the COVID-19 pandemic. Nature. https://doi.org/10.1038/s41586-022-05522-2

Data sources

Bibliographic data: Semantic Scholar, OpenAlex, Crossref.

Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).

Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.