key: cord-0817740-pps56i3b authors: Kenyon, C. title: Intensive COVID-19 testing associated with reduced mortality - an ecological analysis of 108 countries date: 2020-05-30 journal: nan DOI: 10.1101/2020.05.28.20115691 sha: bbeb87f696872655b775e593d5ac594eb6a4777b doc_id: 817740 cord_uid: pps56i3b Background Intensive screening and testing for COVID-19 could facilitate early detection and isolation of infected persons and thereby control the size of the epidemic. It could also facilitate earlier and more targeted therapy. These factors could plausibly reduce attributable mortality which was the hypothesis tested in this study. Methods Linear regression was used to assess the country-level association between COVID-19 attributable mortality per 100 000 inhabitants (mortality/capita) and COVID-19 tests/capita (number of tests/100 000 inhabitants) controlling for the cumulative number of COVID-19 infections/100 000 inhabitants (cases/capita), the age of the epidemic (number of days between first case reported and 8 April), national health expenditure per capita and WHO world region. Results The COVID-19 mortality rate varied between 0.3 and 3110 deaths/100 000 inhabitants (median 30, IQR 8-105). The intensity of testing per 100 000 also varied considerably (median 21,970, IQR 2,735-89,095) as did the number of COVID-19 cases per 100 000 (median 1,600, IQR 340-4,760 cases/100 000). In the multivariate model, the COVID-19 mortality rate was negatively associated with tests/capita (Coef. -0.036, 95% CI -0.047- -0.025) and positively associated with cases/capita (Coef. 0.093, 95% CI 0.819- 1.034). Conclusions The results are compatible with the hypothesis that intensive testing and isolation could play a role in reducing COVID-10 mortality rates. Pearson's correlation revealed deaths/capita to be positively correlated with 139 cases/capita (r=0.80, P<0.001), which was in turn positively correlated with 140 tests/capita (r=0.55, P<0.001), and health care expenditure (r=0.51, P<0.001; Table 141 1). Tests/capita was positively correlated with health care expenditure (r=0.45, 142 P<0.001). 143 In the multivariate model, the COVID-19 mortality rate was negatively associated 145 with tests/capita (Coef. -0.036, 95% CI -0.047--0.025) and positively associated with 146 cases/capita (Coef. 0.093, 95% CI 0.819-1.034; Table 2 ). Sensitivity analyses 147 limited to countries with epidemics older than 15 March 2020 made little difference to 148 the findings (Table 4) . 149 150 Our analysis confirms the logical association between increased testing intensity and 152 increased detection of cases/capita as well the association between cases/capita 153 and deaths/capita. The key finding was that on multivariate analysis, testing intensity 154 was negatively associated with mortality/capita. This is compatible with the theory 155 that intensive testing is associated with reduced mortality via reducing the spread of control for other variables such as national age structure and prevalence of 163 comorbidities which may have influenced mortality [9] . 164 165 These limitations notwithstanding, our analysis provides additional evidence to 166 promote calls to intensify national COVID-19 testing to not only control the spread of 167 this disease but also to reduce associated mortality. (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. (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 30, 2020. . https://doi.org/10.1101/2020.05.28.20115691 doi: medRxiv preprint r a n e a n 0 . 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 30, 2020. . https://doi.org/10.1101/2020.05.28.20115691 doi: medRxiv preprint (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 30, 2020. 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 30, 2020. . https://doi.org/10.1101/2020.05.28.20115691 doi: medRxiv preprint As COVID-19 cases, deaths and 232 fatality rates surge in Italy, underlying causes require investigation. The Journal of Infection 233 in Developing Countries How will country-235 based mitigation measures influence the course of the COVID-19 epidemic? The Lancet Similarity in Case Fatality 238 Rates (CFR) of COVID-19/SARS-COV-2 in Italy and China Estimating 241 clinical severity of COVID-19 from the transmission dynamics in Wuhan, China Defining the Epidemiology of Covid-19 -244 Studies Needed Impact of non-pharmaceutical interventions (NPIs) to reduce COVID19 mortality 248 and healthcare demand Estimating the asymptomatic 250 proportion of coronavirus disease 2019 (COVID-19) cases on board the Diamond Princess 251 cruise ship Wuhan and Hubei COVID-19 mortality 255 analysis reveals the critical role of timely supply of medical resources Estimating the infection and case fatality ratio for coronavirus disease (COVID-19) using 259 age-adjusted data from the outbreak on the Diamond Princess cruise ship