Mortality: A comprehensive look at the COVID-19 pandemic death toll
More than 18 months into the pandemic, the exact death toll of COVID-19 remains elusive. There are several ways to assess how many people have died in a pandemic, each with their advantages and disadvantages. Official national reports of COVID-19 deaths are useful, but their accuracy depends on the level of testing in that country and may underestimate the true death toll. Combining mortality rates with estimates of the fraction of people in a country who have been infected provides a better estimate, but requires serological studies of antibody prevalence that are not widely available. A tried-and-trusted approach to calculate the death toll is to estimate 'excess mortality' by comparing the total number of deaths during the pandemic period with a baseline level of deaths before the pandemic. This indirect statistical approach does not depend on testing strategy.
A recent study, based on data from 29 high-income countries during 2020, reported substantial excess mortality in some Eastern European countries and no excess mortality in New Zealand, Norway or Denmark (Islam et al., 2021). Now, in eLife, Ariel Karlinsky (Hebrew University) and Dmitry Kobak (University of Tübingen) report the results of an excess mortality study that extends to the summer of 2021 and more than 100 countries, and provides a first look at the substantial pandemic death toll in several middle-income countries (Karlinsky and Kobak, 2021). Excess mortality depends on infection rates, population demographics, COVID-19 interventions and vaccine coverage, and reflects the unique pandemic experience of different countries.
Karlinsky and Kobak compiled a unique database of weekly, monthly or quarterly deaths in 103 countries, including five years of pre-pandemic baselines for most of these. In some of the worst-affected countries, mortality in 2020 and the first half of 2021 exceeded the baseline level by between 50% and 150% (e.g., Peru and Mexico), and in absolute terms, more than 0.4% of the population died of COVID-19 in some countries (e.g., Peru and Bulgaria). For Peru, the most severely affected country, the effects of a poor healthcare system may have been exacerbated by strict lockdowns that fostered severe economic restraints and mass migration (Taylor, 2021). Meanwhile, countries like Japan, Finland, the Philippines, and South Korea had negative excess mortality, reflecting excellent pandemic control, which resulted in a modest COVID-19 death toll and a near absence of influenza deaths during the pandemic. For such countries, the official COVID-19 death count is more accurate than excess mortality estimates. The impact of COVID-19 was intermediate in South Africa and Russia, with mortality about 30% higher than the baseline and a death rate of about 0.3%.
Excess mortality reflects the sum of positive and negative changes from baseline years, meaning that some of the changes observed may not be directly related to the COVID-19 pandemic itself, but due instead to interventions. For example, social distancing measures decrease circulation and mortality due to influenza and other non-SARS-CoV-2 pathogens, but other factors – such as overwhelmed healthcare systems, violence and drug overdoses – increase mortality. However, Karlinsky and Kobak convincingly argue that most excess deaths reflect the direct consequences of COVID-19 (see also the study on excess death in Russia: Kobak, 2021).
Karlinsky and Kobak refrain from calculating the global mortality burden of COVID-19. Such an estimate would be heavily biased due to the lack of data for populous countries like China and India, and because many low-income countries in Asia and Africa cannot participate due to a lack of timely national vital statistics. Such gaps in the data can impact a global mortality estimate greatly: for example, one recent study computed a likely toll of about 4 million COVID-19 deaths in India, which is about 10 times higher than official death counts (Anand et al., 2021). Karlinsky and Kobak estimate that for the 103 countries they collected data for, the true number of COVID-19 deaths is on average 1.4 fold greater (range between 1 to 100-fold) than reported.
Global mortality estimates for past influenza pandemics range from 0.4 million deaths for the 2009 pandemic to 75 million deaths for the 1918 pandemic (Murray et al., 2006; Simonsen et al., 2013; Viboud et al., 2016; all adjusted to 2020 population, see Table 1). Globally, 4.3 million deaths due to SARS-CoV-2 have been reported as of August 11, 2021: this corresponds to about 6 million deaths when applying the mean underreporting factor of 1.4 reported by Karlinsky and Kobak. This is clearly a low global estimate due to missing data from key countries and the continued circulation of new SARS-CoV-2 variants in 2021 and in the coming years. Even so, the COVID-19 pandemic is already deadlier than the 1957 pandemic, but has nowhere near the death toll of the pandemic of 1918.
Estimates of global excess mortality for five pandemics.
Estimates of the global per capita excess mortality rate (row 2), the number of global excess deaths adjusted to 2020 population (row 3), and the mean age at death (row 4) for the ongoing COVID-19 pandemic (column 2) and the influenza pandemics of 2009, 1968, 1957 and 1918 (columns 3–6). Each study used different statistical models, assumptions and country data. The levels of mortality in non-participating countries were estimated using various extrapolation/imputation strategies.
COVID-191 | 20092 | 19683 | 19574 | 19185 | |
---|---|---|---|---|---|
Per capita excess mortality rate | >0.08%6 | 0.005% | 0.03% | 0.04% | 1.0% |
Global excess deaths adjusted to 2020 population | >6 million6 | 0.4 million | 2.2 million | 3.1 million | 75 million |
Mean age at death (years) | 707 | 37 | 62 | 65 | 27 |
-
1Karlinsky and Kobak, 2021. 103 wealthier countries; an under-reporting factor of 1.4 was applied.
2Simonsen et al., 2013. Based on 2009 data from 20 countries covering approximately 35% of the world population and using an allcause imputation method that uses 10 factors. Estimates based on 300,000–400,000 pandemic excess deaths from all causes.
-
3CDC, 2019. Based on 1 million excess deaths in the US, UK, Canada, Australia and France.
4Viboud et al., 2016. For the entire pandemic period (1957–1959), using data from 39 countries; extrapolated globally based on GDP, latitude and baseline death rate.
-
5Murray et al., 2006. 13 countries or regions for the entire pandemic period between 1918–1920; allcause mortality; extrapolated by GDP, latitude; (62 million deaths in 2004 population).
6 Based on the official COVID-19 global death toll as of 11/8/2021 multiplied by 1.4 to allow account for underreporting (Karlinsky and Kobak, 2021). This is an underestimate as the 103 participating countries in this study are wealthier, but harder-hit populous countries like India (which may account for approximately 4 million excess deaths) are not included. Also, the burden is incomplete because the COVID-19 pandemic is not yet over.
-
7The COVID-19 pandemic mainly kills the elderly, but the exact mean age of deaths is not currently known. Mean age at death is likely lower in middle-income countries: for example, it is reported to be 60 years in South Africa (Guimarães et al., 2021; Statistics South Africa, 2020).
Importantly, these historical comparisons do not consider long-term decreases in baseline mortality due to better healthcare, longer life expectancy and other factors, which make the COVID-19 pandemic stand out sharply against low background mortality levels. Another key consideration is age: the mean age of people who die of COVID-19 is around 70 years, similar to the 1957 pandemic, but dramatically higher than the 1918 and 2009 pandemics (Table 1). The mean age at death is likely lower for less wealthy countries with younger populations (e.g., 60 years in South Africa; Table 1). Mortality age patterns are critically needed for estimating years of life lost, which is an alternative metric used to understand and compare pandemic death tolls (Viboud et al., 2010; Pifarré I Arolas et al., 2021).
We applaud Karlinsky and Kobak’s efforts to compile, release and update timely mortality data in over 100 countries – a major achievement that would have been impossible even 10 years ago. This is a data revolution that parallels that seen in vaccine development and pathogen sequencing. Future work should focus on including incomplete or subnational mortality data from low- and middle-income countries in Asia, the Middle East and Africa (United Nations) to begin filling the data gap. The database should also be expanded to include age breakdowns whenever available, as in other international mortality databases (such as the Human Mortality Database, COVerAGE-DB, and EuroMOMO). Going forward, we call for resources to maintain these valuable databases in the post-SARS-CoV-2 era, as these can uniquely monitor the complete impact of the COVID-19 pandemic. These databases can also be used to track the death toll of increasingly frequent heat waves and other effects of climate change, and help us be ready for future pandemics.
Note
Disclaimer: This article does not necessarily represent the views of the NIH or the US government.
References
-
WebsiteThree new estimates of India’s all-cause excess mortality during the COVID-19 pandemicCenter for Global Development. Accessed August 4, 2021.
-
Younger Brazilians hit by COVID-19 – What are the implications?The Lancet Regional Health - Americas 373:100014.https://doi.org/10.1016/j.lana.2021.100014
-
Years of life lost to COVID-19 in 81 countriesScientific Reports 11:3504.https://doi.org/10.1038/s41598-021-83040-3
-
Global mortality impact of the 1957–1959 influenza pandemicJournal of Infectious Diseases 213:738–745.https://doi.org/10.1093/infdis/jiv534
Article and author information
Author details
Publication history
- Version of Record published: August 12, 2021 (version 1)
Copyright
This is an open-access article, free of all copyright, and may be freely reproduced, distributed, transmitted, modified, built upon, or otherwise used by anyone for any lawful purpose. The work is made available under the Creative Commons CC0 public domain dedication.
Metrics
-
- 8,368
- Page views
-
- 374
- Downloads
-
- 23
- Citations
Article citation count generated by polling the highest count across the following sources: Crossref, PubMed Central, Scopus.
Download links
Downloads (link to download the article as PDF)
Open citations (links to open the citations from this article in various online reference manager services)
Cite this article (links to download the citations from this article in formats compatible with various reference manager tools)
Further reading
-
- Epidemiology and Global Health
Background:
Short-term forecasts of infectious disease burden can contribute to situational awareness and aid capacity planning. Based on best practice in other fields and recent insights in infectious disease epidemiology, one can maximise the predictive performance of such forecasts if multiple models are combined into an ensemble. Here, we report on the performance of ensembles in predicting COVID-19 cases and deaths across Europe between 08 March 2021 and 07 March 2022.
Methods:
We used open-source tools to develop a public European COVID-19 Forecast Hub. We invited groups globally to contribute weekly forecasts for COVID-19 cases and deaths reported by a standardised source for 32 countries over the next 1–4 weeks. Teams submitted forecasts from March 2021 using standardised quantiles of the predictive distribution. Each week we created an ensemble forecast, where each predictive quantile was calculated as the equally-weighted average (initially the mean and then from 26th July the median) of all individual models’ predictive quantiles. We measured the performance of each model using the relative Weighted Interval Score (WIS), comparing models’ forecast accuracy relative to all other models. We retrospectively explored alternative methods for ensemble forecasts, including weighted averages based on models’ past predictive performance.
Results:
Over 52 weeks, we collected forecasts from 48 unique models. We evaluated 29 models’ forecast scores in comparison to the ensemble model. We found a weekly ensemble had a consistently strong performance across countries over time. Across all horizons and locations, the ensemble performed better on relative WIS than 83% of participating models’ forecasts of incident cases (with a total N=886 predictions from 23 unique models), and 91% of participating models’ forecasts of deaths (N=763 predictions from 20 models). Across a 1–4 week time horizon, ensemble performance declined with longer forecast periods when forecasting cases, but remained stable over 4 weeks for incident death forecasts. In every forecast across 32 countries, the ensemble outperformed most contributing models when forecasting either cases or deaths, frequently outperforming all of its individual component models. Among several choices of ensemble methods we found that the most influential and best choice was to use a median average of models instead of using the mean, regardless of methods of weighting component forecast models.
Conclusions:
Our results support the use of combining forecasts from individual models into an ensemble in order to improve predictive performance across epidemiological targets and populations during infectious disease epidemics. Our findings further suggest that median ensemble methods yield better predictive performance more than ones based on means. Our findings also highlight that forecast consumers should place more weight on incident death forecasts than incident case forecasts at forecast horizons greater than 2 weeks.
Funding:
AA, BH, BL, LWa, MMa, PP, SV funded by National Institutes of Health (NIH) Grant 1R01GM109718, NSF BIG DATA Grant IIS-1633028, NSF Grant No.: OAC-1916805, NSF Expeditions in Computing Grant CCF-1918656, CCF-1917819, NSF RAPID CNS-2028004, NSF RAPID OAC-2027541, US Centers for Disease Control and Prevention 75D30119C05935, a grant from Google, University of Virginia Strategic Investment Fund award number SIF160, Defense Threat Reduction Agency (DTRA) under Contract No. HDTRA1-19-D-0007, and respectively Virginia Dept of Health Grant VDH-21-501-0141, VDH-21-501-0143, VDH-21-501-0147, VDH-21-501-0145, VDH-21-501-0146, VDH-21-501-0142, VDH-21-501-0148. AF, AMa, GL funded by SMIGE - Modelli statistici inferenziali per governare l'epidemia, FISR 2020-Covid-19 I Fase, FISR2020IP-00156, Codice Progetto: PRJ-0695. AM, BK, FD, FR, JK, JN, JZ, KN, MG, MR, MS, RB funded by Ministry of Science and Higher Education of Poland with grant 28/WFSN/2021 to the University of Warsaw. BRe, CPe, JLAz funded by Ministerio de Sanidad/ISCIII. BT, PG funded by PERISCOPE European H2020 project, contract number 101016233. CP, DL, EA, MC, SA funded by European Commission - Directorate-General for Communications Networks, Content and Technology through the contract LC-01485746, and Ministerio de Ciencia, Innovacion y Universidades and FEDER, with the project PGC2018-095456-B-I00. DE., MGu funded by Spanish Ministry of Health / REACT-UE (FEDER). DO, GF, IMi, LC funded by Laboratory Directed Research and Development program of Los Alamos National Laboratory (LANL) under project number 20200700ER. DS, ELR, GG, NGR, NW, YW funded by National Institutes of General Medical Sciences (R35GM119582; the content is solely the responsibility of the authors and does not necessarily represent the official views of NIGMS or the National Institutes of Health). FB, FP funded by InPresa, Lombardy Region, Italy. HG, KS funded by European Centre for Disease Prevention and Control. IV funded by Agencia de Qualitat i Avaluacio Sanitaries de Catalunya (AQuAS) through contract 2021-021OE. JDe, SMo, VP funded by Netzwerk Universitatsmedizin (NUM) project egePan (01KX2021). JPB, SH, TH funded by Federal Ministry of Education and Research (BMBF; grant 05M18SIA). KH, MSc, YKh funded by Project SaxoCOV, funded by the German Free State of Saxony. Presentation of data, model results and simulations also funded by the NFDI4Health Task Force COVID-19 (https://www.nfdi4health.de/task-force-covid-19-2) within the framework of a DFG-project (LO-342/17-1). LP, VE funded by Mathematical and Statistical modelling project (MUNI/A/1615/2020), Online platform for real-time monitoring, analysis and management of epidemic situations (MUNI/11/02202001/2020); VE also supported by RECETOX research infrastructure (Ministry of Education, Youth and Sports of the Czech Republic: LM2018121), the CETOCOEN EXCELLENCE (CZ.02.1.01/0.0/0.0/17-043/0009632), RECETOX RI project (CZ.02.1.01/0.0/0.0/16-013/0001761). NIB funded by Health Protection Research Unit (grant code NIHR200908). SAb, SF funded by Wellcome Trust (210758/Z/18/Z).
-
- Epidemiology and Global Health
Background:
Affectionate touch, which is vital for mental and physical health, was restricted during the Covid-19 pandemic. This study investigated the association between momentary affectionate touch and subjective well-being, as well as salivary oxytocin and cortisol in everyday life during the pandemic.
Methods:
In the first step, we measured anxiety and depression symptoms, loneliness and attitudes toward social touch in a large cross-sectional online survey (N = 1050). From this sample, N = 247 participants completed ecological momentary assessments over 2 days with six daily assessments by answering smartphone-based questions on affectionate touch and momentary mental state, and providing concomitant saliva samples for cortisol and oxytocin assessment.
Results:
Multilevel models showed that on a within-person level, affectionate touch was associated with decreased self-reported anxiety, general burden, stress, and increased oxytocin levels. On a between-person level, affectionate touch was associated with decreased cortisol levels and higher happiness. Moreover, individuals with a positive attitude toward social touch experiencing loneliness reported more mental health problems.
Conclusions:
Our results suggest that affectionate touch is linked to higher endogenous oxytocin in times of pandemic and lockdown and might buffer stress on a subjective and hormonal level. These findings might have implications for preventing mental burden during social contact restrictions.
Funding:
The study was funded by the German Research Foundation, the German Psychological Society, and German Academic Exchange Service.