• Takayuki Kawashima, Shuhei Nomura, Yuta Tanoue, Daisuke Yoneoka, Akifumi Eguchi, Chris Fook Sheng Ng, Kentaro Matsuura, Shoi Shi, Koji Makiyama, Shinya Uryu, Yumi Kawamura, Shinichi Takayanagi, Stuart Gilmour, Hiroaki Miyata, Tomimasa Sunagawa, Takuri Takahashi, Yuuki Tsuchihashi, Yusuke Kobayashi, Yuzo Arima, Kazuhiko Kanou, Motoi Suzuki, and Masahiro Hashizume.
• Emerging Infect. Dis. 2021 Mar 1; 27 (3): 789-795.

AbstractTo provide insight into the mortality burden of coronavirus disease (COVID-19) in Japan, we estimated the excess all-cause deaths for each week during the pandemic, January-May 2020, by prefecture and age group. We applied quasi-Poisson regression models to vital statistics data. Excess deaths were expressed as the range of differences between the observed and expected number of all-cause deaths and the 95% upper bound of the 1-sided prediction interval. A total of 208-4,322 all-cause excess deaths at the national level indicated a 0.03%-0.72% excess in the observed number of deaths. Prefecture and age structure consistency between the reported COVID-19 deaths and our estimates was weak, suggesting the need to use cause-specific analyses to distinguish between direct and indirect consequences of COVID-19.

18 keywords...

hide…