key: cord-255103-ckecol7i authors: Lau, Hien; Khosrawipour, Tanja; Kocbach, Piotr; Ichii, Hirohito; Bania, Jacek; Khosrawipour, Veria title: Evaluating the massive underreporting and undertesting of COVID-19 cases in multiple global epicenters date: 2020-06-06 journal: Pulmonology DOI: 10.1016/j.pulmoe.2020.05.015 sha: doc_id: 255103 cord_uid: ckecol7i Abstract Background With continuous global COVID-19 outbreak, differing case numbers and mortality rates are observed. While actual case numbers appear vague, mortality numbers related to COVID-19 seem more precise. In this study, we used the mortality rate as the main indicator to evaluate the extent of underreporting and underdetection of COVID-19 cases. Methods We have analyzed all available data provided by the World Health Organization on the development of international COVID-19 cases and mortality numbers on March 17th, 2020. A crude case-fatality risk (cCFR) and adjusted case-fatality risk (aCFR) was calculated for China, South Korea, Japan, Italy, France, Spain, Germany, Iran and the United States. Additionally, a fold-change (FC) was derived for each country. Results The highest aCFR and FC were detected for Spain. Based on their FC values, an extremely high number of undetected COVID-19 cases was displayed in France, the United States, Italy and Spain. For these countries, our findings indicate a detection rate of only 1-2% of total actual COVID-19 cases. Conclusions Due to limited testing capacities, mortality numbers may serve as a better indicator for COVID-19 case spread in many countries. Our data indicate that countries like France, Italy, the United States, Iran and Spain have extremely high numbers of undetected and underreported cases. Differences in testing availability and capacity, containment as well as overall health care and medical infrastructure result in significantly different mortality rates and COVID-19 case numbers for each respective country. Amid the current COVID-19 pandemic, a continuous rise in mortality rates has been observed. At the same time, concerns have been voiced that COVID-19 testing has been insufficient and that many countries either lack testing kits and infrastructure, fear associated expenses or that cultural factors may impede virus' detection [1 -4] . Once patients are hospitalized, their symptoms are described as flu-like, and their condition often deteriorates and results in death. Prior to a death occurrence, testing is often performed to rule out or confirm a COVID-19 related death. While some countries exhibit such restrictive approaches, others have implemented various measures to contain the virus e.g. social distancing, self-quarantine and lockdown. These measures can also potentially influence the testing procedure. Therefore, to get a better understanding of the spread of the virus in each country, this study compares total reported case numbers for each respective country with total COVID-19 related death numbers. If COVID-19 related mortality remains relatively constant within a certain margin, then this may give a much better estimate of virus spread than the case numbers reported. This study, therefore, aims to assess the extent of COVID-19 undertesting and underreporting based on reported and estimated mortality per case in multiple global epicenters, including China, South Korea, Japan, Italy, France, Spain, Germany, Iran and the United States [5, 6] . The total number of confirmed COVID-19 cases and related deaths for Asia (China, South Korea, Japan), Europe (Italy, France, Spain, Germany), Iran and the United States were sourced from the COVID-19 situation reports made publicly available by the World Health Organization (WHO) on January 20 th , 2020. The present study used data reported by the WHO on March 3 rd , 2020 and March 17 th , 2020 (figure 1) [7] . The crude case-fatality risk (cCFR) of COVID-19 infections on March 17 th , 2020 was calculated by dividing the total number of deaths on March 17 th , 2020 by the total number of confirmed cases on March 17 th , 2020 for each respective country [8 -10] . However, it is important to note that deceased patients were typically infected 14 days prior to death occurrence [11] . Therefore, we must consider the time lag between infection and death when calculating an adjusted CFR. For this purpose, we compared the total reported death numbers with confirmed COVID-19 cases tested 14 days prior. Total COVID-19 cases at date t were calculated using the cCFR for each respective country according to the equations below: On March 17 th , 2020, the cCFR for Germany was the lowest among all investigated countries in the study (0.22% ;95% CI: 0.13% -0.37%). This number was used as a benchmark to calculate total COVID-19 cases in other countries. However, the calculated cCFR was not adjusted to the previously described 14-day shift. Adjusted total COVID-19 cases at date t were also calculated. For this purpose, we used the aCFR value for Germany and South Korea. South Korea had the lowest aCFR with 1.68% (95% confidence interval, (CI): 1.36% -2.09%). aCFR values of both countries were used as a benchmark to calculate adjusted total COVID-19 cases in other countries: The Wilson score interval method was used to calculate cCFR and aCFR at a 95% CI 5,6 . To assess the extent of underreporting and undertesting, we compared adjusted total cases to total reported cases in all countries. This number presents the demonstrated fold change for these countries. Of all countries, the aCFR of South Korea was the lowest on March 17 th , 2020 and thus, it was used to calculate the adjusted total COVID-19 cases for the other investigated countries: Crude case-fatality risks (cCFR) vary between 0.22% and 8.95%. Countries can be grouped into 3 distinct cohorts according to cCFR values of 1% (cohort 1), 1%-3% (cohort 2) and above 3% (cohort 3). South Korea and Germany are in cohort 1, with rates of 0.97% and 0.22%, respectively. The second cohort displays a higher cCFR (1%-3%) and includes countries like France (2.25%) and the 3 and table 2) . When estimating the real total amount of COVID-19 cases using the cCFR value of Germany as the standard, a considerable increase in COVID-19 case numbers compared to total reported cases is observed. Based on these calculations, we estimated the following numbers for investigated countries. All data is presented as reported vs. estimated cases in table 3. Total COVID-19 cases were again estimated based on aCFR for Germany and South Korea. Doing so, we observed that estimated numbers were lower than when cCFR was used. This is true for all investigated countries ( figure 5 ). Germany (4.9). The second cohort includes Japan (6.2). The third cohort includes Iran (33.8), France (46), the United States (53.8), Italy (73) and Spain (161) (see figure 5 ). When analyzing reported death numbers, it becomes apparent that the quality of data on reported case numbers is very heterogenous. Calculated fold change indicates that in some emerging COVID-19 epicenters, (USA CF:54, Italy:57 and Spain CF:161), less than 2 percent of COVID-19 cases were subjected to testing and consequently reported. This data is very concerning and points to extreme undertesting and underreporting. While these numbers may appear extraordinarily high for some epicenters, they may indicate a potentially overwhelmed and exhausted medical system or insufficient medical coverage. This lack of adequate medical services may further increase overall mortality. Impaired medical services can be assumed in countries like Italy, Spain and the United States with constant reports of overwhelmed medical facilities. Quality, quantity and capacity of healthcare systems substantially contribute to the successful management of hospitalized patients and can reduce mortality rates. However, it is very challenging to compare different healthcare systems with respect to COVID-19 mortality rates. While we know that healthcare plays a major role in this pandemic, it is not possible to quantify its effect on current mortality rates. In fact, there is a wide range of factors that may play a significant role in total case numbers like extent, use and safety measures in public transportation, population density, access and quality of health system (quality and quantity/capacity), local temperature and humidity factors, cultural and religious practices, and how media presents the urgency of this immediate health threat. Our findings show that COVID-19 testing has been insufficient, and that many countries either lack testing resources, e.g. test kits and personnel, or fear associated costs. While CFR values for Germany and South Korea are probably close to actual mortality rates, this is not the case for most global epicenters in the third cohort [12] . Containment measures such as isolation, quarantine, lockdown and social distancing are highly effective [13, 14] in provided. For most of these deaths, unspecific pneumonia is listed as the primary cause of death. Our data support concerns about massively insufficient testing in many global COVID-19 epicenters compared to Germany and South Korea. If we assume that mortality rates are roughly stable, COVID-19 related mortality numbers might serve as a better indicator than case numbers to grasp the extent of COVID-19 spread. However, it is important to note that COVID-19 related mortalities typically occur 14 days after infection. The datasets used and/or analyzed during the current study are available from the corresponding author upon reasonable request. Table 3 : Estimating total COVID-19 cases and crude case-fatality risks (cCFR) Internationally lost COVID-19 cases COVID-19 epidemic in Switzerland: on the importance of testing, contact tracing and isolation The global community needs to swiftly ramp up the response to contain COVID-19 Can we contain the COVID-19 outbreak with the same measures as for SARS? Lancet Infect Dis Case-fatality estimates for COVID-19 calculated by using a lag time for fatality Methods for Estimating the Case Fatality Ratio for a Novel, Emerging Infectious Disease World Health Organization. Coronavirus disease 2019 (COVID-19) situation report Estimating risk for death from 2019 novel coronavirus disease, China Case fatality: rate, ratio, or risk? Epidemiology Case fatality ratio of pandemic influenza Updated understanding of the outbreak of 2019 novel coronavirus (2019-nCoV) in Wuhan The reproductive number of COVID-19 is higher compared to SARS coronavirus The positive impact of lockdown in Wuhan on containing the COVID-19 outbreak in China Isolation, quarantine, social distancing and community containment: pivotal role for old-style public health measures in the novel coronavirus (2019-nCoV) outbreak The authors have declared no conflicts of interest. The authors did not receive any funding for this study. No ethical approval was required for this study.