The 1918-19 influenza pandemic in Portugal: a regional analysis of mortality impact

Baltazar Nunes, Susana Silva, Ana Rodrigues, Rita Roquette, Inês Batista, Helena Rebelo-de-Andrade

Research output: Contribution to journalArticlepeer-review

10 Citations (Scopus)
217 Downloads (Pure)

Abstract

Although many archeo-epidemiological studies have assessed the mortality impact of the 1918-19 influenza pandemic, detailed estimates are not available for Portugal. We applied negative binomial models to monthly mortality data from respiratory and all-causes at the national and district-level from Portugal, 1916-1922. Influenza-related excess mortality was computed as the difference between observed and expected deaths. Poisson regression was used to estimate the association between geographic, socio-demographic factors and excess mortality. Two waves of pandemic influenza were identified between July 1918-January 1919 and April- May 1919, representing an excess all-cause death rate of 195.7 per 10,000. All districts of Portugal were affected. The pandemic hit earlier in southeastern districts and the main cities, while excess mortality was highest in the Northeast, in line with the high mortality burden experienced by northern Spanish provinces. During the period of intense excess mortality (fall winter 1918-19), population density was negatively associated with pandemic impact. This pattern changed in the March 1919-June 1920 wave, where excess mortality increased with population density, and north and west directions. Portuguese islands were less and later affected. Given the geographic heterogeneity evidenced in our study, subnational socio-demographic characteristics and connectivity should be integrated in pandemic preparedness plans.

Original languageEnglish
Pages (from-to)2541-2549
JournalAmerican Journal of Epidemiology
Volume187
Issue number12
Early online date7 Aug 2018
DOIs
Publication statusPublished - 2018

Fingerprint

Dive into the research topics of 'The 1918-19 influenza pandemic in Portugal: a regional analysis of mortality impact'. Together they form a unique fingerprint.

Cite this