Estimates of the excess mortality associated with the COVID-19 pandemic from the World Health Organization (preprint)

As a part of its mandate to compile and disseminate statistics on mortality, the World Health Organization (WHO) has been tracking the progression of the COVID-19 pandemic since the beginning of 2020. However, reported statistics on COVID-19 are problematic for a number of countries due to variations in testing access, differential diagnostic capacity and inconsistencies in the applications of standards to correctly certify COVID-19 as cause-of-death. In addition, the pandemic has caused extensive collateral damage beyond what is directly attributable to it. Consequently, the WHO has estimated excess deaths for each country for the years 2020 and 2021 to quantify the pandemic’s impact more comprehensively and consistently. Defined as the number of deaths in a particular period, relative to that expected during “normal times”, excess deaths capture both the direct and indirect impacts of a crisis. The data required to estimate the excess mortality associated with the COVID-19 pandemic i.e., time-series of known deaths during the pandemic period and historical time-series of the same to forecast into the pandemic period as “expected”, are only available for a subset of countries. This paper describes the methods used to estimate the global, regional, and country specific estimates of excess mortality for the years 2020 and 2021 and provides an overview of the resultant estimates. The full details of the method development, validation and performance are provided in a separate report. In summary, excess deaths have been derived by the WHO using an over-dispersed Poisson count framework that applies Bayesian inference techniques to quantify uncertainty. The framework utilizes data from locations that have recorded national monthly data to build a loglinear regression model with both time-varying and time-invariant coefficients. The model is used to predict excess deaths in locations without any all-cause mortality data reported during the pandemic period. Furthermore, certain countries have only subnational data for the period. For these, the framework is used to build country-specific multinomial models that use pre-pandemic subnational data and subnational data reported during the pandemic to predict national level monthly mortality for years 2020 and 2021. Globally, 14.91 million excess deaths are estimated with a 95% Uncertainty Interval (UI) from 13.32 million to 16.64 million which is 2.75 (UI 2.45 to 3.07) times higher than the 5.42 million COVID-19 deaths reported for the period. There is wide variation in the excess estimates across the six WHO regions. African Region accounts for 1.25 million excess deaths (UI 0.91 million to 1.58 million), Region of the Americas for 3.23 million excess deaths (UI 3.16 million to 3.30 million), Eastern Mediterranean Region for 1.08 million excess deaths (UI 0.87 million to 1.30 million), European Region for 3.25 million excess deaths (UI 3.18 million to 3.32 million), South-East Asia Region for 5.99 million excess deaths (4.50 million to 7.72 million) and Western Pacific Region accounting for 120 thousand excess deaths (UI –65 thousand to 351 thousand). Across the World Bank Income groups, High-income economies account for 2.16 million excess deaths (UI 2.09 million to 2.24 million), Upper-middle-income economies account for 4.24 million excess deaths (UI 4.18 million to 4.31 million). Lower-middle-income economies account for 7.87 million excess deaths (UI 6.30 million to 9.60 million) and Low-income economies account for 638 thousand excess deaths (UI 434 thousand to 846 thousand).

Estimating Global and Country-Specific Excess Mortality During the COVID-19 Pandemic (preprint)

Estimating the true mortality burden of COVID-19 for every country in the world is a difficult, but crucial, public health endeavor. Attributing deaths, direct or indirect, to COVID-19 is problematic. A more attainable target is the "excess deaths", the number of deaths in a particular period, relative to that expected during "normal times", and we estimate this for all countries on a monthly time scale for 2020 and 2021. The excess mortality requires two numbers, the total deaths and the expected deaths, but the former is unavailable for many countries, and so modeling is required for these countries. The expected deaths are based on historic data and we develop a model for producing expected estimates for all countries and we allow for uncertainty in the modeled expected numbers when calculating the excess. We describe the methods that were developed to produce the World Health Organization (WHO) excess death estimates. To achieve both interpretability and transparency we developed a relatively simple overdispersed Poisson count framework, within which the various data types can be modeled. We use data from countries with national monthly data to build a predictive log-linear regression model with time-varying coefficients for countries without data. For a number of countries, subnational data only are available, and we construct a multinomial model for such data, based on the assumption that the fractions of deaths in sub-regions remain approximately constant over time. Based on our modeling, the point estimate for global excess mortality, over 2020-2021, is 14.9 million, with a 95% credible interval of (13.3, 16.6) million. This leads to a point estimate of the ratio of excess deaths to reported COVID-19 deaths of 2.75, which is a huge discrepancy.

A framework for reconstructing SARS-CoV-2 transmission dynamics using excess mortality data (preprint)

Detailed reconstruction of the SARS-CoV-2 transmission dynamics and assessment of its burden in several parts of the world has still remained largely unknown due to the scarcity of epidemiological analyses and limited testing capacities of different countries to identify cases and deaths attributable to COVID-19 [1-4]. Understanding the true burden of the Iranian COVID-19 epidemic is subject to similar challenges with limited clinical and epidemiological studies at the subnational level [5-9]. To address this, we develop a new quantitative framework that enables us to fully reconstruct the transmission dynamics across the country and assess the level of under-reporting in infections and deaths using province-level, age-stratified all-cause mortality data. We show that excess mortality aligns with seroprevalence estimates in each province and subsequently estimate that as of 2021-10-22, only 48% (95% confidence interval: 43-55%) of COVID-19 deaths in Iran have been reported. We find that in the most affected provinces such as East Azerbaijan, Qazvin, and Qom approximately 0.4% of the population have died of COVID-19 so far. We also find significant heterogeneity in the estimated attack rates across the country with 11 provinces reaching close to or higher than 100% attack rates. Despite a relatively young age structure in Iran, our analysis reveals that the infection fatality rate in most provinces is comparable to high-income countries with a larger percentage of older adults, suggesting that limited access to medical services, coupled with undercounting of COVID-19-related deaths, can have a significant impact on accurate estimation of COVID-19 fatalities. Our estimation of high attack rates in provinces with largely unmitigated epidemics whereby, on average, between 10% to 25% individuals have been infected with COVID-19 at least twice over the course of 20 months also suggests that, despite several waves of infection, herd immunity through natural infection has not been achieved in the population.

National Excess Mortality from Sub-National data: Method and Application for Argentina (preprint)

Currently, many countries have not yet reported 2020 or 2021 mortality data to allow an estimate of excess mortality during the COVID-19 pandemic. However, some countries have sub-national mortality data, at the state, province or city level. I present a simple method to allow estimation of national level mortality and excess mortality from sub-national data, using the case of Argentina and projecting excess mortality up to August 2021.

The World Mortality Dataset: Tracking excess mortality across countries during the COVID-19 pandemic (preprint)

Comparing the impact of the COVID-19 pandemic between countries or across time is difficult because the reported numbers of cases and deaths can be strongly affected by testing capacity and reporting policy. Excess mortality, defined as the increase in all-cause mortality relative to the recent average, is widely considered as a more objective indicator of the COVID-19 death toll. However, there has been no central, frequently-updated repository of the all-cause mortality data across countries. To fill this gap, we have collected weekly, monthly, or quarterly all-cause mortality data from 77 countries, openly available as the regularly-updated World Mortality Dataset. We used this dataset to compute the excess mortality in each country during the COVID-19 pandemic. We found that in the worst-affected countries the annual mortality increased by over 50%, while in several other countries it decreased by over 5%, presumably due to lockdown measures decreasing the non-COVID mortality. Moreover, we found that while some countries have been reporting the COVID-19 deaths very accurately, many countries have been underreporting their COVID-19 deaths by an order of magnitude or more. Averaging across the entire dataset suggests that the world's COVID-19 death toll may be at least 1.6 times higher than the reported number of confirmed deaths.