The dashed diagonal line represents the equality between the number of excess deaths and of COVID-19 reported deaths. The 0 marker on the y-axis indicates no excess mortality. Countries are colored according to their decile of the test-to-case ratio. Countries appearing in gray had unavailable or incomplete data on testing.
eAppendix. Supplemental Methods
Customize your JAMA Network experience by selecting one or more topics from the list below.
Identify all potential conflicts of interest that might be relevant to your comment.
Conflicts of interest comprise financial interests, activities, and relationships within the past 3 years including but not limited to employment, affiliation, grants or funding, consultancies, honoraria or payment, speaker's bureaus, stock ownership or options, expert testimony, royalties, donation of medical equipment, or patents planned, pending, or issued.
Err on the side of full disclosure.
If you have no conflicts of interest, check "No potential conflicts of interest" in the box below. The information will be posted with your response.
Not all submitted comments are published. Please see our commenting policy for details.
Sanmarchi F, Golinelli D, Lenzi J, et al. Exploring the Gap Between Excess Mortality and COVID-19 Deaths in 67 Countries. JAMA Netw Open. 2021;4(7):e2117359. doi:10.1001/jamanetworkopen.2021.17359
During the SARS-CoV-2 pandemic, a surge in overall deaths has been recorded in many countries, most of them likely attributable to COVID-19. However, COVID-19 confirmed mortality (CCM) is considered an unreliable indicator of COVID-19 deaths because of national health care systems’ different capacities to correctly identify people who actually died of the disease.1,2 Excess mortality (EM) is a more comprehensive and robust indicator because it relies on all-cause mortality instead of specific causes of death.3 We analyzed the gap between the EM and CCM in 67 countries to determine the extent to which official data on COVID-19 deaths might be considered reliable.
In this cross-sectional study, we retrieved aggregated country-level data on population and COVID-19 overall confirmed cases, deaths, and testing as of December 31, 2020, from Our World in Data. Data on countries’ overall deaths from 2015 to 2020 were obtained from the World Mortality Data set (eAppendix in the Supplement). This research was based on public use datasets that do not include identifiable personal information and, per the Common Rule, was exempt from Institutional Review Board review and approval. For the same reason, no informed consent was required. This study follows the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) reporting guideline.
Negative binomial regression models were used to estimate projected deaths in 2020 using mortality data from 2015 to 2019. Two-sided 95% CIs for country-specific projected deaths were calculated applying the normal approximation to the Poisson distribution. EM in the pandemic period (ie, February 26 to December 31, 2020) was estimated as the difference between cumulative observed deaths and projected deaths. Countries’ testing capacity was assessed with their cumulative test-to-case ratio (eAppendix in the Supplement). The association between country-specific cumulative CCM and EM per 100 000 population of 2020 was displayed using a scatterplot, in which the identity line discriminates countries with EM exceeding CCM from those with EM lower than CCM. A color was assigned to countries based on their decile of testing capacity. All analyses were performed using R version 4.0.4 (R Project for Statistical Computing). Details on the analytic approach are available in the eAppendix in the Supplement.
Most of the 67 countries experienced an increase in mortality during 2020 (Table). Among countries with increased mortality (ie, those located above 0 on the y-axis in the Figure), a small number appeared under the identity line, showing lower-than-expected mortality after subtracting COVID-19 deaths. Countries located above the identity line can be visually classified into 2 groups: 1 with several Latin American and East European countries, which exhibit a large gap between EM and CCM (eg, Mexico, 212 excess deaths vs 96 COVID-19 deaths per 100 000 population); the other, more heterogeneous group showed a moderate EM beyond CCM (eg, Greece, 57 excess deaths vs 45 COVID-19 deaths per 100 000 population). Countries with negative EM also had very low CCM and were mainly located in East Asia. The lowest figures of EM and CCM generally belonged to countries with higher testing capacity (in green) and the largest differences between EM and CCM to countries with poorer testing capacity (in red).
This comparison of CCM and EM revealed the different national health systems’ capacity to test and diagnose COVID-19 and their responsiveness to the health crisis. Underreporting of COVID-19 deaths because of strained health care systems’ capacity might explain our findings for countries where EM exceeded CCM.2,4 In contrast, the effects of nonpharmaceutical interventions on populations’ main causes of deaths, such as the decrease in work and road accidents, could be responsible for the reduction in overall mortality in countries where CCM exceeded EM.5 Notably, most of the countries that presented reduced overall mortality during 2020 had extremely high testing capacity and were praised for their effective response measures against the pandemic.6
Limitations of our analysis include the lack of stratification by age and sex, the underrepresentation of some areas of the world, and not considering nonpharmaceutical interventions. Despite these drawbacks, our findings corroborate the evidence that in many countries the accuracy in quantifying the death toll of COVID-19 is still a missed target. The global action against the pandemic is being conditioned by diverse responses to the crisis, but reliable evidence should be the pillar on which effective prevention measures are built.
Accepted for Publication: May 13, 2021.
Published: July 16, 2021. doi:10.1001/jamanetworkopen.2021.17359
Open Access: This is an open access article distributed under the terms of the CC-BY License. © 2021 Sanmarchi F et al. JAMA Network Open.
Corresponding Author: Davide Golinelli, MD, Department of Biomedical and Neuromotor Sciences, Alma Mater Studiorum–University of Bologna, Via San Giacomo 12, 40126 Bologna, Italy (firstname.lastname@example.org).
Author Contributions: Dr Sanmarchi had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.
Concept and design: Sanmarchi, Golinelli, Capodici, Gibertoni.
Acquisition, analysis, or interpretation of data: All authors.
Drafting of the manuscript: Sanmarchi, Golinelli, Capodici.
Critical revision of the manuscript for important intellectual content: Sanmarchi, Golinelli, Lenzi, Esposito, Reno, Gibertoni.
Statistical analysis: Sanmarchi, Lenzi, Capodici, Gibertoni.
Supervision: Golinelli, Gibertoni.
Conflict of Interest Disclosures: None reported.
Create a personal account or sign in to: