Influenza in India 1918: epicenter of an epidemic

Routine demographic data from censuses and civil registration for India for the period 1901 to 1921 are used to assess the size and regional distribution of the impact of the 1918 influenza epidemic. Methods of data evaluation are used to adjust for possible under(or over-) reporting of deaths relative to census counts; given the nature of the age groups used these methods are given some modification. The analysis presents estimates for the country, provinces (now called states) and finally focuses on the districts of Central Provinces and Berar, the area which appears to have been most severely affected of all the areas of India. Analysis of adjusted data suggests that previous estimates of “excess” deaths in the range of 17.5 to 22.5 million were probably too high, having not adequately taken into account the effect of the epidemic on births; our estimates are in the range of 11.0 to 13.5 million. Even so, the “excess” crude death rate between August 1918 and January 1919 is estimated to have averaged over 30 per 1,000, and in Central Provinces to have exceeded 60 per 1,000. For the districts of Central Provinces, associations between excess death rates and other characteristics are explored; severity is found to be positively associated with rainfall, and negatively associated with the sex ratio of the population and emigration rates, but a number of socio-economic indicators including initial mortality level showed no association with excess mortality.