Excess and Exiguous Deaths Dashboard in Japan
NEWS
About the dashboard
This dashboard displays the calculated number of excess and exiguous deaths for all-cause deaths and by major causes of deaths for the period during the novel coronavirus disease (COVID-19) pandemic (i.e., from January 2020 onwards) in Japan and the period before that (i.e., from 2017 to 2019). By selecting a prefecture and a time period, the number of excess and exiguous deaths in that prefecture and period will be displayed.The number of excess and exiguous deaths displayed on the dashboard is calculated by subtracting the actually observed number of deaths from the number of deaths predicted in previous data using a statistical model. An explanation about the statistical model used and the data analysis can be found in the research article or from the National Institute of Infectious Diseases’'website [in Japanese].
All-cause death: When all causes of death are included in the analysis, the observed excess number of deaths is not only the sum of deaths directly caused by COVID-19, but also includes deaths due to indirect effects, such as deaths due to aggravation of pre-existing conditions that may have resulted from curtailment of going out or other lifestyle changes to mitigate the risks of infection. The exiguous deaths are also likely to include both a decrease in deaths related to the increased implementation of infectious disease control measures and the improved personal health management in response to COVID-19 and a decrease in deaths not attributable to the pandemic. For details about the interpretation of the excess and exiguous all-cause deaths, please refer to “Interpretation of the number of excess and exiguous all-cause deaths” below.
By major causes of death: To assess the indirect mortality impact of COVID-19, this dashboard also reports the number of excess and exiguous deaths by major causes of death. By calculating the number of exiguous deaths, it can be used to assess the positive health impact of increased infectious disease control measures and personal health management than in previous years. The following six disease categories, which includes the major causes of death in Japan, are covered our analysis (The classification numbers in parentheses are based on the International Statistical Classification (ICD-10)).
- (1) Deaths from all causes, excluding deaths due to COVID-19 (U07.1)
- (2) Deaths due to diseases of the respiratory system (J00-J99, R09.2, U04)
- (3) Deaths due to diseases of the circulatory system (I00-I99)
- (4) Deaths due to malignant neoplasm (cancer) (C00-C97)
- (5) Deaths due to senility (R54)
- (6) Death due to suicide (X60-X84)
By excluding deaths from COVID-19 as the primary cause of death, we can yield excess and exiguous deaths due to other causes in (1). This gives us an insight into the overall indirect mortality impact of COVID-19. (2) to (5) are the top five disease categories [in Japanese] in accordance with the ranking of causes of death in Japan in 2019, using the categories of the International Statistical Classification (ICD-10) [in Japanese]. (6) has been included in the analysis as an important cause of death as evident from a previous study suggesting excess deaths. Analyses using (2) to (6) are intended to identify the indirect mortality effects of COVID-19 separately.
The aim of this dashboard is to contribute to open, evidence-based debate about the impact of COVID-19 in Japan. In order to ensure an open debate, all data on this dashboard and all program codes used in the analysis are made public (see Supplementary documents). In addition, to facilitate appropriate interpretation of the number of excess and exiguous deaths displayed on this dashboard, appropriate understanding of the concept of excess and exiguous deaths and the estimation methods used in the statistical model is important. The homepage of the National Institute of Infectious Diseases provides a detailed explanation and a Q&A page [in Japanese], but please feel free to submit your comments and inquiries if you have any further questions.
Definition of the number of excess and exiguous deaths on the dashboard
On this dashboard, the numbers of excess and exiguous deaths are displayed as ranges. In these ranges, the upper limit is the difference between the actually observed number of deaths and the predicted number of deaths as a point estimate (threshold 1), based on the number of deaths occurring in previous years. The lower limit is the difference between the actually observed number and the 95% one-sided prediction interval（upper and lower bounds） (threshold 2).
For example, based on the number of deaths occurring in previous years, if the point estimate is 100 deaths with 125 deaths for the 95% one-sided prediction interval (upper bound) and the actual number of deaths at 130, then the displayed range of number of excess deaths would be “5-30” (if the actual number of deaths is less than the predicted number of deaths, then the number of excess deaths is considered 0). If the point estimate is 155 deaths with 140 deaths of 95% one-sided prediction interval (lower bound) and the actual number of deaths at 130, then the exiguous deaths range would be "10-25" (if the actual number of deaths exceeds the point estimate, then the number of exiguous deaths is set to zero).
Method of calculating numbers of excess and exiguous deaths
The Farrington algorithm, known internationally as a standard method and used by the US Centers for Disease Control and Prevention,was used to calculate the numbers of excess and exiguous deaths . For details about the methods, please refer to the National Institute of Infectious Diseases homepage [in Japanese].
Interpretation of the number of excess and exiguous all-cause deaths
The number of excess deaths for all-cause deaths is likely to include deaths both directly and indirectly caused by COVID-19. To say that, “there were excess deaths” does not necessarily mean the same as, “there were deaths due to COVID-19,” as it is possible that some of the excess deaths were not directly caused by COVID-19. Similarly, to say that “there were no excess deaths” does not necessarily mean the same as, “there were no deaths due to COVID-19,” as deaths due to COVID-19 did occur, but the excess deaths were possibly cancelled out by a decrease in the number of deaths due to causes other than COVID-19. Attention must be paid in these respects when interpreting the data.
In general, excess and exiguous deaths may occur due to infectious diseases other than COVID-19, changes in atmospheric temperature, or other incidental factors. In fact, incidental increases in the numbers of excess and exiguous deaths were seen from 2017 to 2019, before the COVID-19 pandemic. The increase in the numbers of excess and exiguous deaths in the pre-COVID-19 era can be referred to as the range of the number of excess deaths that can occur without COVID-19.
During the COVID-19 pandemic (i.e., during and after January 2020), in addition to the deaths directly due to COVID-19, it is highly probable that changes in the number of deaths would be complicated by indirect effects of the spread of COVID-19. This can include causes of death that are more common than those seen in previous years, such as deaths due to a failure to be seek medical care during lockdowns and exacerbation of chronic diseases by lifestyle changes due to COVID-19, as well as causes of death that are less common than those seen in previous years, such as deaths due to infectious diseases other than COVID-19. A detailed breakdown of the causes of deaths for the calculated number of excess deaths is presented below.
- COVID-19 has been diagnosed as the direct cause of death; this diagnosis is correct and the cause of death is indeed COVID-19.
- COVID-19 has been diagnosed as the direct cause of death, but this diagnosis is incorrect, and the actual cause of death is something other than COVID-19. For example, the cause of death is influenza, but COVID-19 has been diagnosed as the cause of death. However, in the current situation where COVID-19 is diagnosed using polymerase chain reaction tests, almost no cases may belong to this category.
- COVID-19 has not been diagnosed as the direct cause of death. Instead, a different disease has been diagnosed as the direct cause of death, but the diagnosis is incorrect, and COVID-19 is in fact the cause of death.
- COVID-19 is not the direct cause of death, and the death is due to a different disease, but the COVID-19 pandemic has had an indirect effect on that disease. For example, the person who dies may not have sought medical care or his/her chronic disease may have been exacerbated by lifestyle changes during the pandemic.
- COVID-19 is not the direct cause of death, and the COVID-19 pandemic has not had even an indirect effect on the cause of death. The number of excess deaths from 2017 to 2019 reflects such deaths.
The calculated number of exiguous deaths can be interpreted as the sum of the reduction in the number of deaths, mainly due to the following factors:
- A decrease in the number of deaths due to the increased infectious disease mitigation measures and improved personal health management in response to COVID-19.
- A decrease in the number of deaths not attributable to the increased infectious disease mitigation measures and improved health management in response to COVID-19. (e.g., a decrease in the number of deaths due to the influence of seasonal influenza epidemics, climatic factors, and other coincidental factors).
Supplementary materials
- R_code.R
- All-cause deaths
- By causes of deaths
Recommended Broswers
To make sure you have best experience possible, we recommend using the most up-to-date version of one of the following browsers:
- Microsoft Edge 89.0.x ～
- Google Chrome 89.0.x ～
- Firefox 86.0.x ～
- Safari 14.0.x ～
Other browsers may have difficulty displaying the full functionality of this site.
Member
This work was supported by the Ministry of Health, Labour and Welfare of Japan (23HA2005).
*Research leader
Dr. Masahiro Hashizume*
Department of Global Health Policy, Graduate School of Medicine, The University of TokyoDr. Shuhei Nomura*
Keio University Global Research Institute (KGRI)Dr. Akifumi Eguchi
Center for Preventive Medical Sciences, Chiba UniversityDr. Daisuke Yoneoka
Infectious Disease Surveillance Center at the National Institute of Infectious DiseasesDr. Takayuki Kawashima
Department of Mathematical and Computing Science, Tokyo Institute of TechnologyDr. Yuta Tanoue
Faculty of Marine Technology, Tokyo University of Marine Science and TechnologyDr. Yumi Kawamura
Department of Health Policy and Management, School of Medicine, Keio UniversityDr. Takehiko I. Hayashi
Center for Health and Environmental Risk Research, National Institute for Environmental StudiesDr. Yoonhee Kim
Department of Global Environmental Health, Graduate School of Medicine, The University of TokyoDr. Chris Fook Sheng Ng
Department of Global Health Policy, Graduate School of Medicine, The University of TokyoDr. Paul LC Chua
Department of Global Health Policy, Graduate School of Medicine, The University of TokyoMr. Ganan Devanathan
Department of Global Health Policy, Graduate School of Medicine, The University of TokyoDr. Motoi Suzuki
Infectious Disease Surveillance Center at the National Institute of Infectious Diseases
