key: cord-0882696-yjny2xcf authors: Sonoo, M.; Kanbayashi, T.; Shimohata, T.; Kobayashi, M.; Hayashi, H. title: Estimation of the true infection rate and infection fatality rate of COVID-19 in the whole population of each country date: 2020-05-16 journal: nan DOI: 10.1101/2020.05.13.20101071 sha: 1824c22fbd8041b407a3e38c8a3ef1279c458cd3 doc_id: 882696 cord_uid: yjny2xcf The true infection rate of COVID-19 and the infection fatality rate in the whole population have been estimated for each country. The estimate well coincided with local surveys. The fact that several tens of times, or more, of identified cases are already infected might require reconsideration of the strategy. Coronavirus disease 2019 (COVID-19) pandemic is now a worldwide peril. Recently, a few local surveys revealed surprisingly high infection rates in general population (1) (2) (3) (4) . However, the True Infection Rate (TIR) in the whole population of each country is unknown. We devised a simple method to infer TIR, as well as the Infection Fatality Rate (IFR) based on the open data. We analyzed the data at a website on April 12th and 26th, 2020 (5) . Countries or regions having more than 1000 cases on April 12th were included. Positive rate of PCR tests (Infection Rate; IR) and the rate of PCR tests among the population (Examination Rate; ER) were calculated. The cardinal assumption is that IR and ER are negatively correlated because PCR examinations will be restricted to cases with strong suspicion while ER is low, and vice versa. Then TIR of a specific region can be estimated using the linear regression. The slope value calculated from the data on April 12th was applied to the data on April 26th, because of the reasons explained later. The ratio of the infected persons in the population to the already identified cases was calculated and named as True/Identified Case Ratio (TICR). IFR was estimated using the number of total deaths at the same website (5) . For each parameter, 95% confidence interval (CI) was calculated. All statistical calculation was done using Microsoft Excel for Mac. Included were 67 countries or regions. IR was negatively correlated with ER on logarithmic scales on April 12th (r = -0.394, p = 0.00097; Figure 1 ) and on 26th (r = -0.362, p = 0.0026). The slope value was -0.234 on 12th and -0.234 on 26th. The estimated TIR, TICR and IFR on April 26th of representative countries are presented ( Table 1 ). The data of Tokyo and Hyogo prefectures in Japan are also added. The data for the whole countries or regions investigated are presented in the Appendix Table. TIR took on values between 1% to 10% for All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint this version posted May 16, 2020. . https://doi.org/10.1101/2020.05.13.20101071 doi: medRxiv preprint most countries. TICR was around 20 to 40 in major western countries, and over 100 in some countries, including Japan. IFR also took on wide range of values, generally higher in western countries than in eastern countries. A number of factors obscure the correlation between IR and ER. These include the variation of actual situations of conducting PCR-tests among different countries. However, these varieties will be averaged over many countries included in this study. The observed correlation was highly significant, indicating that the inverse relation of IR and ER is sufficiently robust. The variation of TIR among countries will become higher as the time lapses, which will result in decreasing correlation (actually, -0.394 to -0.362 during 2 weeks). As another disturbing factor, if countries with high TIR actively expand the PCR tests due to a sense of crisis, the negative correlation will be reduced, which will also become evident as the time lapses. These are the reasons why we used the slope value calculated from the earlier data. As a matter of fact, however, the slope value remained almost constant during 2 weeks (-0.234 to -0.234) despite the increase of the y-intercept from 0.810 to 0.833 (corresponding to 5% increase of IR), which indicates the reliability of our analysis. Comparison of the present results with local surveys supports the validity of our estimate. Keio University in Tokyo found 6% IR using PCR tests for in-patients unrelated to COVID-19 (1). Our estimate of TIR in Tokyo is 6.8%. Others are antibody surveys. Kobe City Medical Center General Hospital in Hyogo prefecture, Japan found 3.3% IR for unselective out-patients (2) . Our estimate of TIR in Hyogo is 1.8% (95%CI: 0.8 -4.5). Reported IRs were 1.5% in Santa Clara (3) and 13.9% in New York state (4), and estimated TIR in the whole USA is 6.9%. The pandemic is still expanding and uncontrolled in most countries as evident in the trajectory analysis (6, 7) . Only limited countries, such as China or Iceland, succeeded in All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint this version posted May 16, 2020. . https://doi.org/10.1101/2020.05.13.20101071 doi: medRxiv preprint containing the pandemic (present estimation is not applicable to these countries). The fact that several tens of times of identified cases are already infected, often asymptomatic, may urge reconsideration of the strategy in the management of COVID-19. Finding out all infected persons and completely containing the infection is unrealistic since ER is less than 3% in most major countries. Other than complete containment, herd immunity, and effective treatment/vaccines, attenuation of the viral virulence is another way to make the pandemic less damaging (8) . This is achieved in animals by natural selection (9) but in humans the prevention of in-hospital infection, especially from severe cases, would be the key. This might be a choice, especially in countries with already low estimated IFR, mainly Asian countries such as Singapore, India, Indonesia or Japan (Table 1) . We would like to thank Dr. Masashi Idogawa, Sapporo Medical University School of Medicine, for providing information on the trajectory analysis. All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint this version posted May 16, 2020. . https://doi.org/10.1101/2020.05.13.20101071 doi: medRxiv preprint Appendix Table. Parameters of the COVID-19 Pandemic in 68 countries and regions. (Appendix Table is submitted as a separate word file) All rights reserved. No reuse allowed without permission. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. The copyright holder for this preprint this version posted May 16, 2020. . https://doi.org/10.1101/2020.05.13.20101071 doi: medRxiv preprint Update on 2019 Novel Coronavirus Situation at Keio University Hospital Three percent of out-patients have antibodies New York coronavirus antibody study shows just how far the virus has spread Reported Cases and Deaths by Country, Territory, or Conveyance Trajectory analysis of new coronavirus COVID-19 cases and deaths by country Interactive web-based graphs of novel coronavirus COVID-19 cases and deaths per population by country. Clinical Infectious Diseases in press The spread of the COVID-19 coronavirus: Health agencies worldwide prepare for the seemingly inevitability of the COVID-19 coronavirus becoming endemic All rights reserved. No reuse allowed without permission.(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.The copyright holder for this preprint this version posted May 16, 2020. . https://doi.org/10.1101/2020.05.13.20101071 doi: medRxiv preprint