Search

ImportanceReliable estimates of COVID-19 mortality are crucial to aid control strategies and to assess the effectiveness of interventions.ObjectiveProject COVID-19 mortality trends to October 1, 2020, in 12 countries or regions that constitute >90% of the global COVID-19 deaths reported as of April 12, 2020.Design, Setting, and ParticipantsThe Global COVID-19 Assessment of Mortality (GCAM) is an open, transparent, and continuously updated (www.cghr.org/covid) statistical model that combines actual COVID-19 mortality counts with Bayesian inference to forecast COVID-19 deaths, the date of peak deaths, and the duration of excess mortality. The analyses covered a total of 700 million population above age 20 in 12 countries or regions: USA; Italy; Spain; France; UK; Iran; Belgium; a province of China (Hubei, which accounted for 90% of reported Chinese deaths); Germany; the Netherlands; Switzerland; and Canada; and six US states: New York, New Jersey, Michigan, Louisiana, California, and Washington.ResultsForecasted deaths across the 12 current high-burden countries sum 167,000 to 593,000 (median 253,000). The trajectory of US deaths (49,000-249,000 deaths; median 86,000)—over half of which are expected in states beyond the initial six states analysed in this study—will have the greatest impact on the eventual total. Mortality ranges are 25,000-109,000 (median 46,000) in the UK; 23,000-31,000 (median 26,000) in Italy; 21,000-37,000 (median 26,000) in France and 21,000-32,000 (median 25,000) in Spain. Estimates are most precise for Hubei, China—where the epidemic curve is complete—and least precise in California, where it is ongoing. New York has the highest cumulative median mortality rate per million (1135), about 12-fold that of Germany. Mortality trajectories are notably flatter in Germany, California, and Washington State, each of which took physical distancing and testing strategies seriously. Using past country-specific mortality as a guide, GCAM predicts surge capacity needs, reaching more than twice existing capacity in a number of places., In every setting, the results might be sensitive to undercounts of COVID-19 deaths, which are already apparent.Conclusion and RelevanceMortality from COVID-19 will be substantial across many settings, even in the best case scenario. GCAM will provide continually updated and increasingly precise estimates as the pandemic progresses.The coronavirus disease (COVID-19) pandemic has already caused over 115,000 deaths, with global deaths doubling every week.1-3Mortality is less biased than case reporting, which is affected by testing policies. However, the daily reporting of COVID-19 deaths is already known to undercount actual deaths, varying over time and place.4-6Reliable estimates of total COVID-19 mortality, the date of peak deaths, and of the duration of excess mortality are crucial to aid responses to the current and potential future pandemics. We have developed the Global COVID-19 Assessment of Mortality (GCAM), a statistical model to project COVID-19 mortality trends to October 1 2020 in 12 countries or regions that constitute >90% of the global COVID-19 deaths reported as of April 12th. We report also on six US states that account for 70% of the American totals to date (Supplementary Appendix).1We quantify the COVID-19 mortality trajectory ranges in each setting. A semi-automated website (www.cghr.org/covid) provides daily updates. GCAM is open, transparent, and uses a reasonably simple method that employs publicly reported mortality data to make plausible projections. The method is designed to improve as more mortality data become available over longer time periods.

Background: Canada had substantial mortality from COVID in the first SARS-CoV-2 wave, mostly in nursing homes. Efforts to stem the pandemic (including vaccination, now being introduced) can benefit from direct understanding of the relationship of age-specific mortality to SARS-CoV-2 seroprevalence and from information on asymptomatic infection. Methods: The Action to Beat Coronavirus (Ab-C) study surveyed a reasonably representative sample of 19,994 adult Canadians about COVID symptoms and analyzed IgG antibodies against SARS-CoV-2 from self-collected dried blood spots (DBS) in 8,967 adults. A sensitive and specific chemiluminescence ELISA detected IgG to the spike trimer. We compared seroprevalence to deaths to establish infection fatality rates (IFRs) and used mortality data to estimate infection levels in nursing home residents. Findings: The best estimate (high specificity) of adult seroprevalence nationally is 1.7%, but as high as 3.5% (high sensitivity) depending on assay cutoffs. The highest prevalence was in Ontario (2.4-3.9%), the most populous province, and in younger adults aged 18-39 years (2.5-4.4%). Based on mortality, we estimated 13-16% of nursing home residents became infected. The first viral wave infected 0.54-1.08 million adult Canadians, half of whom were 70 times higher than comparably-aged adults living in the community. Seropositivity correlated with COVID symptoms, particularly during March. Asymptomatic adults constituted about a quarter of definite seropositives, with a greater proportion in the elderly. Interpretation: Canada had relatively low infection prevalence and low IFRs in the community, but not in nursing homes, during the first viral wave. The Ab-C study demonstrates the practicability of using self-collected DBS for antibody testing in national surveys to monitor the pandemic and vaccine-induced immunity in the population. Funding Statement: Pfizer Global Medical, Unity Health Foundation, Canadian COVID-19 Immunity Task Force Declaration of Interests: We declare no competing interests. Ethics Approval Statement: Ethics approval was provided by St. Michael's Institutional Review Board. REB 20-107