When the World Health Organization (WHO) declared an end to Covid-19 as a global health emergency on May 5, 2023, most governments pointed to a combined reported death toll of seven million lives. 

However, a new report from the WHO has tilted the scale on this number, describing the total death toll as a ‘profound undercount’. The actual death toll, measured through excess mortality data spanning 240 countries across four years, stands at 22.1 million, which is more than three times the official figure, and the single largest documented revision of pandemic mortality in public health history. 

The findings, published in Chapter 2 of the WHO World Health Statistics 2026 and released on 13 May 2026, were produced in collaboration with the University of Washington and represent the most comprehensive, year-by-year accounting of pandemic mortality ever assembled, disaggregated by region, sex, age group, and national income level. 

Official Covid-19 death counts rely on confirmed diagnoses where a person must have been tested, the result must have been recorded, and the death must have been attributed to the virus. However, in much of the world, the report notes, particularly across sub-Saharan Africa, South and South-East Asia, and parts of Latin America, those conditions were rarely met simultaneously. 

The actual death toll is calculated by comparing the total number of deaths recorded in a given period against the number that would have been expected under normal, pre-pandemic conditions. The gap between the two captures not only direct Covid-19 deaths, but also lives lost to disrupted healthcare, economic collapse, and the downstream effects of overwhelmed health systems, as well as deaths directly caused by the virus that were never formally attributed to it. 

This approach, the WHO states, is the most reliable tool available for understanding the pandemic’s true human cost as it does not require a Covid-19 test, but only that deaths are counted. 

Analysing the Covid-19 pandemic year by year in four chapters  

2020: The First Wave – Alpha 

In the pandemic’s first year, an estimated 4.5 million excess deaths occurred globally, against 1.8 million officially reported Covid-19 fatalities, translating to an undercount ratio of 2.5 times. All-cause deaths ran approximately 6.2 per cent above expected levels, signalling serious systemic strain even before health systems had been overwhelmed. 

The hardest-hit regions were the Americas and Europe, where crude excess death rates reached approximately 150 per 100,000 population. In the Americas, this figure rose to 180 per 100,000. The African region recorded comparatively lower rates of 40 per 100,000, a figure that almost certainly reflects data scarcity more than genuine mortality protection. 

Even in this first year, WHO notes that the gap between reported and actual deaths was telling. For every confirmed Covid-19 death, roughly one additional excess death went unrecorded, whether because of limited testing, strained death registration infrastructure, or deaths that occurred from other conditions worsened by the pandemic. 

2021: The Delta Peak 

The year 2021 became the deadliest of the pandemic. Global excess deaths reached 10.4 million, which is nearly three times the officially reported 3.5 million. The P-score, which shows the percentage difference between observed deaths and expected deaths, rose to +17.9 per cent, meaning that almost one in five deaths in 2021 would not have occurred in a non-pandemic year. 

The Delta variant collided with health systems still recovering from 2020, producing cascading indirect mortality: patients who could not access care for heart disease, cancer, or other conditions as hospitals prioritised Covid-19 cases. South-East Asia surged dramatically, matching the Americas and Europe at approximately 200 excess deaths per 100,000. The African region jumped to 110 per 100,000, a tripling from the previous year. 

Age-standardised male mortality hit 143 per 100,000, compared with 119 per 100,000 for women. Among men aged 85 and above, excess mortality reached 3,305 per 100,000 people, reflecting the acute vulnerability of elderly men to both direct infection and healthcare disruption. 

2022: The Omicron variant phase 

The Covid-19 Omicron variant was a highly mutated strain of the coronavirus that was first identified in late 2021. It was designated a “Variant of Concern” by the WHO because it spread much faster than earlier variants such as Alpha and Delta.  

In 2022, excess deaths fell to 3.9 million globally, against 1.2 million reported deaths. The undercount ratio widened to 3.3 times, driven largely by countries abandoning widespread testing as containment measures eased. 

The Western Pacific region became the highest-burden area in 2022, recording 130 excess deaths per 100,000 the clearest illustration of how the Omicron variant, though less severe in clinical outcomes, found new populations without prior immunity or adequate vaccine coverage when restrictions lifted. 

The male-female gap narrowed significantly as the direct lethality of later variants declined: male rates fell to 55 per 100,000, female rates to 46. This convergence reflected a shift away from acute respiratory deaths toward a broader mix of pandemic-related causes. 

2023: The endemic phase  

By 2023, excess deaths had returned to approximately the 2020 level of 3.3 million. Officially reported deaths, however, collapsed to 400,000, producing an undercount ratio of 8.2 times, the highest of any year in the series. 

WHO points out that this is not primarily a sign of genuine mortality improvement but a data-reporting failure because countries that stopped widespread testing also stopped attributing deaths to Covid-19, even as the virus continued to circulate. The European region led in 2023 with 50 excess deaths per 100,000. Africa recorded 25 per 100,000, though that figure remains subject to severe data limitations. 

The WHO notes that 32 per cent of countries had never submitted cause-of-death data to international databases as of 2023. The 2023 excess mortality estimate is therefore the most uncertain of all four years, and the 8.2 undercount ratio reflects a reporting collapse as much as it reflects the pandemic’s trajectory. 

High-income countries recorded the highest crude excess death rate of 103 per 100,000 over the full period, largely because they have older populations and, crucially, because they have the vital registration infrastructure to count their dead.  

Middle-income countries together account for 72 per cent of total excess deaths in absolute numbers, split evenly between lower-middle (37 per cent) and upper-middle (35 per cent) income brackets. Low-income countries account for just 4 per cent, not because they were spared, but because they lack the systems to document what happened. 

Over the course of the full pandemic period, men experienced a much higher crude excess death rate than women 

One of the most consistent findings across all four years is the disproportionate mortality burden borne by men and older adults. Over the full pandemic period, men experienced a crude excess death rate of 75 per 100,000 against 63 per 100,000 for women. This gap was most pronounced in 2021, at the height of the Delta wave. 

Age-stratified data tell an even sharper story. Among those aged 85 and above, male excess deaths over the pandemic period reached approximately 7,370 per 100,000, which is about 12 times the rate seen in the 55-59 age group. The age curve is steep and exponential: below the age of 40, excess mortality rates remain low for both sexes; above 60, the divergence accelerates sharply. 

The African region recorded the lowest crude excess death rates in each year of the WHO’s analysis at 40 per 100,000 in 2020, 110 in 2021, 50 in 2022, and 25 in 2023. Over the full period, Africa accounts for 7 per cent of global excess deaths. These numbers are widely regarded by researchers as significant underestimates. 

A 2023 longitudinal study of Kenya’s Kilifi Health and Demographic Surveillance System, published in Nature Communications, found an overall excess mortality of 4.8 per cent during five Covid-19 waves between April 2020 and May 2022, with a markedly higher excess of 11.6 per cent among adults aged 65 and above.  

The study’s age-standardised excess mortality rate was estimated at 27.4 per 100,000 person-years for 2020–2021, substantially lower than comparable figures for high-income countries but significant nonetheless, particularly for older Kenyans. 

Kenya has limited capacity for counting births, deaths, and causes of death as a result of a weak civil registration system 

A parallel study drawing on population-based infectious disease surveillance data from Asembo in rural western Kenya and Kibera in Nairobi, published in PLOS Global Public Health in 2023, found increased mortality rates from acute respiratory infections during the pandemic period, confirming that Covid-19’s indirect effects reached even populations with limited formal health contact. 

The key constraint, however, is not the pandemic itself but the infrastructure for measuring it. According to a WHO SCORE assessment, Kenya has limited capacity for counting births, deaths, and causes of death as a result of a weak civil registration system; a gap that constrains the government’s ability to plan, allocate resources, and monitor health progress equitably. 

In response, Kenya’s Ministries of Health and Interior, with support from WHO, UNICEF, UNDP, UNFPA, and UN Women, launched an initiative to strengthen civil registration and vital statistics systems. A rapid mortality surveillance mechanism was deployed in six high-burden and border counties in June 2021, improving the timeliness and completeness of death notification. Across sub-Saharan Africa more broadly, only 18 countries on the continent record and report annual death totals at all, and of those, only one in three deaths was officially registered at the time of the pandemic. 

The WHO’s World Health Statistics 2026 report contextualises this as a structural failure in global health data infrastructure, not simply a Covid-19 accounting problem. As of the end of 2025, only 18 per cent of countries were reporting mortality data to WHO within one year of the death occurring. Nearly one-third of countries have never submitted cause-of-death data to international databases at all. 

Of the estimated 61 million deaths that occurred globally in 2023, only 21 million were reported with a cause of death. Just one third of countries meet WHO standards for high-quality mortality data; approximately half have low, very low, or no usable data. 

Dr Alain Labrique, WHO’s Director for Data, Digital Health, Analytics and Artificial Intelligence, was direct at the report’s launch: data gaps severely limit the ability to monitor real-time health trends, compare outcomes across countries, and design effective public health responses.  

Countries without accurate mortality data cannot identify which populations are dying, from what causes, and at what rates. They cannot allocate resources efficiently, detect emerging epidemics in time, or hold health systems accountable.  

Sources: WHO World Health Statistics 2026, Nature Communications 14, 6620, PLOS Global Public Health 3(8): e0002141, WHO (2022), WHO (2026), WHO estimates of excess mortality associated with the COVID-19 pandemic 

Data analytics & visualisation: Stanley Njihia    

Text: Yvonne Kawira