by Frank Schnittger
Fri May 15th, 2020 at 12:14:18 PM EST
The data above (sorted by total deaths per country) is taken from Worldometer with the exception of two calculated column (in red) of confirmed case mortality rates (deaths/confirmed cases) and % positive tests (confirmed cases/tests) which I have added to illustrate the degree of testing which is happening in each country. Obviously the more widespread and intensive the testing regime, the lower the % positive rate you are going to get. Conversely, countries which only test the most obvious cases, usually on admission to hospital, will return a much higher % positive rate.
With the world average coming in at 8.36% for % positive tests (confirmed cases/tests), any countries below that are testing more intensively than average, and those above, less so. The outliers for low levels of testing are Ecuador, Mexico and Brazil. Mexico and Brazil are seeing a massive spike in deaths, while Ecuador, mysteriously, is seeing hardly any new cases or deaths at all. Perhaps their rate of testing is so low they are not recording many new illnesses or deaths as Covid-19 related.
Surprisingly, the USA and many of the worst hit countries in Europe are testing less intensively than the world average, perhaps because of political disorganisation, or because their health systems have been overwhelmed. Some may be pursuing a de facto if unacknowledged "herd immunity" strategy, in which case testing for immunity is more important than testing for the disease.
Of the hardest hit countries, only Italy, Canada, Germany, India, Russia, Portugal and Romania are testing more intensively than the world average, not all renowned for the quality of their health services. Perhaps we need to adjust some of the preconceptions the mainstream media would have us have of them.
Of course if you measure the testing regime by tests per million population, a different pattern emerges, but that reflects unfairly on countries with small or non-existent outbreaks, where testing is much less necessary. The level of testing required is proportionate to the size of the outbreak in each country, and on that measure - % of tests results that are positive - many of the worst hit countries come out rather badly.
It looks like the WHO mandated strategy of test, test, test, followed by contact tracing and quarantine is being vindicated the world over. That is if you really are trying to contain the outbreak. Some countries - the UK, Netherlands - aren't even bothering to report the number of recoveries from the disease, which any reasonable observer would regard as perhaps the most important metric of all.
Why are death rates rates so different in different countries, ranging from 16.4% of those testing positive in Belgium to 2.6% in neighbouring Luxembourg? Again testing rates may be a factor, with Belgium's 53,839 tests per million little more than half the Luxembourg rate of 94,021 per million population. But that can only explain part of the variation; other factors include the degree to which a health service has been overrun, and the stage the pandemic has reached. New deaths have been declining in most European countries for some time.
Table II: Sorted by deaths/million
Sorting the data by deaths/million (Table II above) gives some idea of how hard, in relative terms, different countries have been hit, with European countries to the fore, but with the USA, Brazil, and Mexico catching up fast (Table III) below. Russia is to the fore in having large numbers of new cases, but so far this has not been reflected in the mortality figures, with their confirmed case mortality rate a remarkably low 0.9%. That figure has habit of trending upwards, though, as the Pandemic progresses, and Russia appears to be in the early stages of what will end up being one of the largest outbreaks in the world.
Table III: In order of new deaths
Finally, we must note that there have been a lot of problems with false negatives in testing:
Research this week raised new questions about tests that are widely used to diagnose Covid-19, including the Abbott Laboratories test used by the White House, suggesting that false negative results may be fairly common.
The Abbott test, which is used for rapid diagnoses, could miss between one-third and half of positive cases, according to a New York University study that hasn't been reviewed by other scientists. (Abbott disputes the findings.) A separate, peer-reviewed study from Johns Hopkins researchers looked more broadly at a widely used group of diagnostics for current infections, called polymerase chain reaction tests. It found that false negatives were common when the tests were performed early in infection and that patients are likely to get the most accurate results when tested just a few days into having symptoms.
From a Bloomberg e-mail
So the numbers we are seeing above may be a gross under-estimate, even where widespread testing has taken place. It looks like this pandemic is going to be with us for some time yet, especially as it is only beginning to take hold in many third world countries with less comprehensive health care systems and perhaps more immunocompromised people.
Covid-19's characteristic of being asymptomatic for a period in all infections, and asymptomatic for the entire infection in some cases has proved to be a very effective transmission strategy that no testing strategy can overcome, unless we were to test large parts of the population on a regular basis, something no country has attempted.
Sifting through the data can reveal some important patterns and point to more effective management strategies, but in the absence of an effective vaccine and treatment options may may well have to accept some level of ongoing infections and deaths for some time to come.