Uncounted deaths from Covid-19

A new analytical model reveals that the actual number of deaths from Covid-19 globally may be many times higher than published data.
Official figures show there have been 55,000 Covid-19 deaths in South Africa since March 27 last year. That puts the country’s death rate at 92.7 per 100,000 people, the highest in sub-Saharan Africa. However, the reported figures are much lower than the reality.

An estimate of the number of daily excess deaths from Covid-19 around the world. Photo: The Economist In the year to 8 May, South Africa recorded 158,499 excess deaths (deaths that were higher than estimated under previous trends due to demographic changes). Health officials are confident that 85-95% of these deaths are caused by the new strain of corona virus, which is almost three times the official statistics. The disparity stems from the fact that, for a death to be recognized as caused by SARS-CoV-2, the deceased needs to undergo testing and be confirmed. Although South Africa does more testing than its neighbors, rates are still low. And the cause of death is not fully documented for those who die at home. Common phenomenon Mortality rates that exceed official statistics, at least at some point during the pandemic, occur in most if not all of the world. According to the most recent data, the number of excess deaths in the US was 7.1% higher than official statistics between early March 2020 and mid-April 2021. Studies have discovered such a mismatch in several countries. For example, the UK saw an excess death rate that was higher than published in the first wave of infections, but lower in the second, reflecting how measures to contain the spread of the virus helped. save many lives. A similar situation occurs in France. The inadequacy is simply because most countries, especially poor ones, do not provide such data in a timely manner. The best official data only reflect more than half of the real numbers, while the worst data show only about a quarter of reality. Photo: The Economist The Economist has attempted to model the extent of excess deaths during the pandemic in countries that do not report. With 95% probability, the model estimates the number of deaths globally to date between 7.1 and 12.7 million, with an average of 10.2 million. Not surprisingly, most of the missed deaths are in low- and middle-income countries. The actual death toll in most rich OECD countries is 1.17 times the official data, but 14 times higher in sub-Saharan Africa. Notable points The structure of the first and second waves in Europe and the US is less visible in the data of the statistical model for the whole world. Overall, the pandemic is increasingly concentrated in developing economies and continues to increase. In some places, seroprevalence surveys have shown that many people have Covid-19 antibodies, a sign of previous illness. Other factors that may also play a role are the measures governments have taken to limit the spread of the virus. Demographics also have an effect, with more young people often synonymous with lower mortality rates. By collecting 121 indicators for more than 200 countries and regions and machine learning models, The Economist experts estimate that, by May 10, there is a 95% probability that the pandemic has led to 2.4 – 7.1 million excess deaths in Asia (official death toll at 0.6 million), 1.5 – 1.8 million in Latin America and the Caribbean (officially 0.6 million), approx. 0 -2.1 million in Africa (officially 0.1 million), 1.5 – 1.6 million in Europe (officially 1 million) and 0.6 – 0.7 million in the US and Canada ( official 0.6 million). In Oceania, where only 1,218 deaths are officially reported, the model predicts that the actual number of deaths is between 12,000 and 13,000. The estimated range for Africa and Asia is very broad due to lack of data. If the probability is 50%, the estimated range narrows significantly, to only 3.3 – 5.2 million excess deaths in Asia, 0.8 -1.6 million in Africa and 8 ,2 – 10.5 million cases worldwide. In 2020, the number of daily deaths increased during 33/52 weeks. After a brief lull in early 2021, they bounced to new highs, largely due to the tragedy in India. According to estimates, the South Asian country is seeing between 6,000 and 31,000 deaths a day, far exceeding the official figure of about 4,000, in line with independent epidemiological estimates. Remarkably, the death rate from the epidemic is actually worse in richer countries. For Asia and Africa, the estimated death toll per million population is about half that of Europe. India is just the equivalent of Britain, at least for the time being. If two communities have the same level of health care, the community with more elderly people will have a higher number of deaths. If demographics are the only differentiating factor, estimates suggest that epidemic deaths in Japan (mean age 48) are 13 times higher than in Uganda (mean age 17). However, in absolute terms, the mortality rate among the young and poor population is much higher than that of the rich population of the same age group. And for the elderly in poor countries, the situation is clearly dire. The number of excess deaths (red) and the number of officially announced deaths by week in the US. Photo: The Economist Meaning of model The fact that mortality rates are relatively low in the developing world seems to be due to age several different meanings, as the virus is spreading more easily among younger people, a finding that has been reinforced by investigations. serum value. This implies many cases of non-fatal illness. It also means that the virus has a lot of opportunities to mutate. There is a documented exception where in some Southeast Asian countries, mortality rates appear to be very low. It is possible that residents here are benefiting from “cross-immunity”, the level of protection against SARS-CoV-2 caused by infection in other virus outbreaks in the area in the past. But, unfortunately, there are now signs that the numbers are on the rise. The Economist’s estimation model is the first of its kind, though not the only way to deduce the total number of deaths from Covid-19. However, Ariel Karlinsky, a statistician at Israel’s Kohelet Economic Forum think tank, notes that either way, the estimates are no substitute for official data. Tuan Anh