3.8 million deaths a year associated with severe fungal disease

The figure comes from David Denning's paper published in 2024 in The Lancet Infectious Diseases. For 2019–2021 he estimated more than 6.55 million invasive infections a year and more than 3.75 million deaths associated with severe fungal disease. About 2.55 million—68% of the unrounded value—were considered directly attributable to the disease, with a 35–90% range for that share. The panel rounds the three values to 6.5, 3.8 and 2.5 million, so dividing the displayed labels produces a slightly different ratio.

The method rests on populations at risk as denominators. For each condition the author established how many exposed patients there are — people with chronic obstructive pulmonary disease, patients in intensive care, people with lung cancer or with a haematological malignancy — and then applied incidence proportions drawn from the literature from 2010 to 2023 and more than 85 country-level and global burden papers. Data from over 120 countries were included.

Mortality was computed in two forms, crude and attributable, using untreated mortality, the proportion of patients actually treated and survival among those treated. Awareness of the problem, the existence of guidelines and access to diagnostics and therapy were used to estimate the ratio of treated to untreated cases. That means the result depends not only on biology but on how well a given country's health system works.

Three caveats belong with the number. The estimates exclude influenza and COVID-19 outbreaks, although both raise the risk of secondary fungal infection. The value is not a measurement but a model built on denominators and transferred proportions. And fungal asthma, which affects roughly 11.5 million people and may contribute to 46,000 deaths a year, is counted separately, outside the invasive-infection total.

Infections, associated deaths and attributable deaths

The burden model estimated 6.5 million invasive fungal infections a year, 3.8 million deaths associated with severe fungal disease and 2.5 million deaths directly attributable to it. These are not values from successive years; they are increasingly narrow levels of the same model. “Associated” includes patients with serious coexisting diseases, while “attributable” attempts to isolate the infection’s contribution.

The author built denominators from populations at risk, data covering more than 120 countries and over 85 burden studies. The result chiefly describes 2019–2021, with evidence reviewed through 2023. It is an estimate, not a global register of death certificates.

Where the 68% comes from

The author calculates the share from unrounded values of more than 2.55 million attributable deaths and more than 3.75 million associated deaths. The panel deliberately shortens these to 2.5 and 3.8 million, so the labels alone do not reproduce exactly 68%. The more important distinction is between association with disease and direct attribution.