Weekly COVID-19 fatality predictions from leading epidemiologists

By Dr. Liji Thomas, MDApr 9 2020

While many developed and developing countries are reeling under the impact of the SARS-CoV-2 virus, which causes COVID-19 disease, preventive and therapeutic measures alike need more accurate forecasts of fatal and non-fatal events to achieve preparedness and ensure an optimal outcome. Now scientists at Imperial College, London, have come up with a new weekly forecast of the number of deaths due to coronavirus and the likely number of cases reported in countries where the virus is actively spreading.

NIAIDFollow Novel Coronavirus SARS-CoV-2 Colorized scanning electron micrograph of an apoptotic cell (red) infected with SARS-COV-2 virus particles (yellow), isolated from a patient sample. Image captured at the NIAID Integrated Research Facility (IRF) in Fort Detrick, Maryland. Credit: NIAID

Importance of forecasts of death

Various health data organizations have analyzed the current data to bring out forecasts of the impact of the disease on the healthcare system, in terms of the number of days times hospital beds occupied by COVID-19 patients; the number of days in intensive care; and the number of ventilator days.

The number of deaths is also being predicted. These measures are intended to help the healthcare systems to prepare for the maximum impact of the pandemic, for instance, by reducing the number of non-urgent non-COVID-19 demands for healthcare and increasing the number of temporary hospital beds and ventilator units in active service. The predicted time to peak patient volume is crucial in planning such interventions.

Public health authorities can also gauge the need for lockdown and similar measures and take extra care to enact, enforce, and maintain these throughout expected peak impact to prevent unnecessary overload on the hospital system.

The current prediction

The current projection was produced by researchers from the WHO Collaborating Centre for Infectious Disease Modelling, MRC Centre for Global Infectious Disease Analysis, J-IDEA, Imperial College London, and the University of Sussex.

The researchers rely on the number of deaths due to COVID-19 rather than the reported number of cases, because the former is more reliable and varies less over time than the latter measure. The quality of surveillance and of reporting varies from country to country, and this impacts the accuracy of the predictions.

If the reporting of deaths is unreliable, the forecast for the coming week will be too low. On the other hand, if the country involved is one with very few deaths, the prediction is likely to be unreliable.

The first report was produced to cover April 1-7.

How many deaths are predicted?

The model has produced an estimate of over 5,000 deaths due to the virus in the coming week. The chief contributions will be from France, Spain, Italy, the UK, and the USA. The following are the death estimates for the countries listed below, from highest to least:

• Over 5,000 deaths: France, Italy, Spain, the UK, and the US
• 1,000 to 5,000 deaths: Algeria, Belgium, Brazil, Germany, Iran, Netherlands, Switzerland, and Turkey
• 100 to 1,000 deaths: Austria, Canada, Denmark, Ecuador, India, Indonesia, Ireland, Israel, Mexico, Peru, Philippines, Poland, Portugal, Romania, Sweden
• Less than 100 deaths: Argentina, China, Czech Republic, Dominican Republic, Egypt, Greece, Iraq, Japan, Luxembourg, Malaysia, Morocco, Norway, Panama, South Korea

Transmissibility from April 1-7

Death from the disease has an inevitable time lag from the time of transmission. Thus, the scientists say, “Our estimates of transmissibility are a reflection of the transmissibility at the time at which the reported deaths were infected.” Changes in transmissibility estimates will reflect the efficacy of control measures, but with this delay.

There are 42 countries in which the virus is spreading rapidly. The researchers, therefore, examined the rates of transmissibility. This refers to the number of cases that each individual with the infection, on average, is likely to infect.

The researchers estimate rapid growth in the number of cases in 26 of these countries, including the US, the UK, France, Germany, China, and India. Rapid growth means the average number of people infected by one case of COVID-19 is more than 2. Other countries in this category include African countries like Egypt and Algeria; European countries like Turkey, Belgium, Czech Republic, Denmark, Ireland, Poland, Romania, Switzerland and Sweden; South American countries like Brazil, Ecuador, and Peru; Israel; Mexico; and Asian countries like Japan and the Philippines.

Slow growth, with transmissibility between 1 and 2, is estimated in the Netherlands and Spain. Stabilizing transmissibility, with less than one person being infected by one case, is estimated in 14 countries, such as Italy, Norway, and South Korea. Other countries in this category include Argentina, Austria, Greece, Luxembourg, Dominican Republic, Indonesia, Iran, Iraq, Malaysia,   Morocco.

The aim of generating these forecasts is to arm the governments and public health organizations in these countries with the necessary foreknowledge to take preventive and proactive measures to reduce the transmission rates and increase the level of preparedness in these emergency scenarios.