Background: As an island and a former British colony, Sri Lanka is a case of special interest for the study of 1918-1919 influenza pandemic because of its potential for isolation from as well as integration into the world epidemiologic system.
Objectives: To estimate population loss attributable to the influenza pandemic and weekly district-level excess mortality from the pandemic to analyze its spread across the island.
Methods: To measure population loss, we estimated a population growth model using a panel of 100 district-level observations on population for five consecutive censuses from 1891 to 1931, allowing for a one-time drop in population in 1918-1919. To estimate weekly excess mortality from the pandemic, we estimated a seasonally adjusted weekly time series of district-specific mortality estimates from vital registration records, ranked them, and plotted the ranks on weekly maps to create a picture of the geographic pattern of propagation across Sri Lanka.
Results: Total loss of population from the influenza pandemic was 307 000 or approximately 6·7% of the population. The pandemic peaked in two discrete (northern and southern) regions in early October of 1918 and in a third (central) region in early March 1919.
Conclusions: The population loss estimate is significantly higher than earlier estimates of mortality from the pandemic in Sri Lanka, suggesting underreporting of influenza-attributable deaths and a role for influenza-related fertility declines. The spatial pattern of peak mortality indicates the presence of two distinct entry points and three distinct epidemiologic regions, defined by population density and ethnicity, in colonial Sri Lanka.
Keywords: 1918-1919 influenza pandemic; Ceylon; Sri Lanka; epidemic; influenza; island; mortality; spread.
© 2014 The Authors. Influenza and Other Respiratory Viruses Published by John Wiley & Sons Ltd.