Verification of the Overestimation of the “Deaths Associated with Influenza Pandemic of 1918–1919, Japan” Claimed in a Demographic Study

抄録

Official records show that the Japanese influenza death toll in 1918–1920 was 385 thousand. However, a recently published study claims that the records are skeptical, claiming the figure to be “an anomaly by Asian standard,” and re-estimated the number to be about 2 million by its unique demographical calculation. However, it is not sound from the following socio-historical and statistical perspectives: i) Japan had developed accurate registration and surveillance systems which might not have existed in other developing countries; ii) there were unique socio-economic situations that claim that the relatively low mortality rate in Japan was not a “myth”; iii) the proposed re-estimation was an overestimation, because if it was a fact, about 1.6 million influenza deaths should have not been detected nationwide, i.e., about 3% of the population. Also, the influenza death toll was unrealistically large compared to the all-causes mortality of that period; iv) Japan started census in 1920 and it should have significantly affected the demographic data analysis. However, the effect by this artifact was not taken into account. Consequently, it caused the severe overestimation of the death; v) we recalculated the mortality using the same method and dataset but we could not reach an estimated figure similar to that claimed in the paper.