History of the 1918 Influenza Pandemic

In the United States, the first documented cases appeared in spring 1918 in military training camps in Kansas, including Camp Funston (now Fort Riley). By July and August 1918, excess mortality had identified pandemic emergence in China, India, and parts of northern Europe, and by autumn the pandemic was global. The earliest virus-confirmed fatal case dates to May 1918, which suggests the virus was already circulating before it became visible.

No origin has been established. Étaples in France, China, and the Kansas camps have all been proposed, and the pandemic's “everywhere at once” pattern suggests the virus had already spread worldwide below the threshold of detection before fatal cases clustered in large populations.

The pandemic moved in waves: a milder first wave in spring 1918, a far deadlier second wave from autumn 1918, and further recurrences in 1919 and 1920 with lower mortality. By 1922, the excess deaths of young adults had virtually disappeared.

Most infections were typical, self-limiting influenza. What made the pandemic deadly was a disproportionate share of cases progressing to lower respiratory disease, and an unprecedented age pattern: mortality was “W-shaped,” with the usual peaks in infants and the elderly plus a third peak in adults aged 20 to 40, centred around age 27. The lower-than-expected mortality among the elderly may reflect earlier exposure to related H1 and N1 viruses in the nineteenth century.

Most deaths resulted from secondary bacterial pneumonia, most often involving Streptococcus pneumoniae, Streptococcus pyogenes, or Staphylococcus aureus. No antibiotics existed to treat these infections. About 0.5 to 0.7 percent of the entire United States population died, while mortality was much higher in many lower-income countries and in some crowded or isolated communities.

About 500 million people were infected. Global death estimates range from 50 million to 100 million, with about 675,000 deaths in the United States. The W-shaped curve is documented in city records from 1918, including Breslau (now Wrocław, Poland), where the pandemic struck in October 1918.

With no influenza vaccine and no antibiotics, control depended on non-pharmaceutical measures: isolation, quarantine, hygiene advice, disinfectants, mask rules, school closures, and limits on public gatherings. These were applied unevenly, and their effect depended on local authority and public cooperation.

A comparative study of US cities found that places which imposed restrictions earlier and kept them in place longer recorded lower excess mortality. Philadelphia, which held a large Liberty Loan parade in late September 1918 before acting, suffered one of the worst urban outbreaks; St. Louis, which closed schools and suspended gatherings at the first sign of the second wave, fared better.

Mask mandates, fines, and bans on gatherings provoked disputes, resistance, and fatigue. The pandemic therefore belongs beside quarantine and isolation and Epidemics and Public Health, where older epidemic tools met a disease that moved faster than any they had faced.

Mortality varied with housing, nutrition, occupational exposure, access to nursing, local timing, and the capacity to suspend crowded work or travel. Case-fatality was much higher in many developing countries, where records are also sparser, so a global total hides uneven evidence as well as uneven suffering.

Colonial rule and racial inequality shaped both exposure and survival. Indigenous communities, colonised populations, migrants, and poor urban districts could face crowded accommodation, weak services, food insecurity, and authorities that recorded deaths incompletely. In Alaska, the pandemic reached remote Indigenous communities, and a 2025 review reconstructs the non-pharmaceutical measures used there between 1918 and 1920.

Nurses, relatives, neighbours, voluntary organisations, and local officials performed much of the response. They improvised beds, meals, transport, childcare, burial, and home visits when hospitals could not absorb demand. Their labour is less visible than official orders but was essential to care, and connects this topic to the History of Nursing.

The label “Spanish flu” did not identify a Spanish origin. Neutral Spain's press reported the epidemic while wartime censorship limited news elsewhere, helping the name travel. Naming and memory were therefore products of media and war as well as virology.

The 1918 virus was not identified at the time. In 1933, Smith, Andrewes, and Laidlaw at the National Institute for Medical Research in London isolated an influenza A virus from a human patient and propagated it in ferrets; serological work in the 1930s suggested the “classical” swine virus and the 1918 pandemic virus were closely related.

Modern reconstruction began in 1995, when researchers sequenced degraded viral RNA from formalin-fixed autopsy tissue in the collection of the Armed Forces Institute of Pathology. Partial sequences were published in 1997, the complete genome in 2005, and reverse genetics then allowed infectious 1918 virus to be produced and studied under containment.

The 1918 virus was a “founder virus”: every subsequent influenza A pandemic, in 1957 (H2N2), 1968 (H3N2), and 2009 (H1N1), contained gene segments descended from it. It also remains the reference model in pandemic planning, and the 2018 centenary renewed scholarly attention to its unanswered questions.