Contents
In line with its mission, the Editorial Board of MedTvoiLokony makes every effort to provide reliable medical content supported by the latest scientific knowledge. The additional flag “Checked Content” indicates that the article has been reviewed by or written directly by a physician. This two-step verification: a medical journalist and a doctor allows us to provide the highest quality content in line with current medical knowledge.
Our commitment in this area has been appreciated, among others, by by the Association of Journalists for Health, which awarded the Editorial Board of MedTvoiLokony with the honorary title of the Great Educator.
Jacques Raubenheimer, a scientist at the University of Sydney, decided to slightly disenchant the mysterious world of statistics. Using graphs describing the development of the coronavirus pandemic, he showed how to read and understand them. Thanks to his advice, it is easier to separate truth from falsehood, and at the same time, we will shine with professional knowledge.
- Jacques Raubenheimer is an Australian scientist who works at the Faculty of Medicine and Health at the University of Sydney. His scientific interests include biostatistics
- In his analysis, Raubenheimer indicates the most common errors that the media make in interpreting the statistics on the coronavirus, and explains where the false conclusions come from.
- Among the most common mistakes, the scientist lists considering the number of new infections as the actual number of patients. Meanwhile, the latter is always much larger
- Another common mistake is to compare the numbers of infected and dead on the same day
Coronavirus, a global epidemic and restrictions introduced by other countries are one of the most exciting topics today. Many of us have had emotional discussions about safe functioning in the “coronavirus” reality and about the decisions made by politicians.
Are we protecting ourselves well enough? Do the prohibitions make sense? Or maybe the information provided by the media is exaggerated and exaggerates the problem? In such discussions, it is worth using data. Use the statistical data provided by the national authorities, observe the graphs and follow the statistics.
But how to do it right?
- Coverage of the COVID-19 coronavirus [MAP + CHARTS]
After all, the easiest thing for us is to accept the data that paints the picture of a pandemic in line with our ideas. We automatically reach for statistics that support our point of view and quickly ignore those that contradict it. But can we interpret the charts correctly?
Dr. Jacques Raubenheimer shows how to avoid common errors in data interpretation, and at the same time impress your family and friends with knowledge in such a hermetic field.
What matters is the percentage of infected, not dead
Looking at some of the data from the thousands of posts posted daily on social media, you get the impression that the death rate from the coronavirus is actually not high. It is enough to look at the graphs comparing the number of deaths from COVID-19 with other diseases, such as the flu.
However, one very important factor has been omitted from some of these pictures. It’s all about the reproduction rate. To figure it out, we need to look at the infection fatality rate (IFR), which is the number of COVID-19 deaths divided by the total number of infected (a number that we can only estimate at this stage).
The pandemic has yet to say its last word, but we are already seeing COVID-19 having an IFR higher than the flu. Meanwhile, many publications suggest the opposite is true. Dr. Raubenheimer thinks we simply underestimate the coronavirus.
- Why is the coronavirus not like the flu? Just look at the death statistics
So let’s go further. First, if we compare the typical flu IFR of 0,1 percent. with the most optimistic estimate for COVID-19 at 0,25%, it is still more than twice as deadly as the flu.
Second, and most importantly, we need to look at the basic reproduction rate of the virus (R₀). It determines the number of people that can become infected by one infected.
The R₀ of influenza is around 1,3. Although estimates for COVID-19 vary, its R₀ value is around 2,8. As the number of infections is growing exponentially, the jump from 1,3 to 2,8 means COVID-19 is much more contagious than the flu.
When we put together all the information that all the statistics carry, we’ll see why the real experts are pushing for “containment”. Not only because COVID-19 is a lethal disease, but because it is both lethal and highly contagious.
- A record mild flu season in Australia. Will it be similar in Poland?
Exponential growth and graphs that are misleading
A simple graph can show the number of new COVID-19 cases over a specific time period. However, because new cases are reported irregularly, statisticians prefer to report the rate of increase in their number over time. The steeper the graph ascends, the more to worry about.
When it comes to COVID-19, we are tracking an exponential increase in the number of cases. Otherwise – if we do not limit the incidence, their number is constantly growing. This gives us a graph that is flat at first, but then curves sharply upward. And that’s the curve we should flatten.
- Is this the second wave already? “One Great Wave of COVID-19” is sweeping the world
On social media, COVID-19 charts are typically compared to those describing deaths from other causes. And the latter show:
– more linear charts (numbers increase over time but at a constant pace)
– the much slower growing number of deaths due to influenza, or
– small numbers in the early stage of the epidemic, thus ignoring the effect of exponential growth.
Additionally, even scientists talking about exponential growth can mislead us.
A widely disseminated online analysis by Professor Yitzhak Ben-Israel found that the exponential increase in COVID-19 cases “disappears after eight weeks.” Well, the professor got it wrong. But why?
His model assumes that the number of COVID-19 cases increases exponentially over the course of a few days, but does not take into account the numerous transmissions that each take several days. This approach led Professor Ben-Israel to conclude that we are dealing only with a chaotic spike in cases early in the epidemic.
Correct graphs ignore the first cases, e.g. starting with the hundredth, or indicate how many days it takes for their number to double (about 6 to 7 days).
Not all infections are accidental
Australian statisticians emphasize that you need to differentiate between the number of infections and the number of cases. In epidemiological terms, a “case” is someone who has been diagnosed with a positive COVID-19 test.
But there are many more infections than there are cases. Some are asymptomatic, some are mild, and some people think they just have a cold. Additionally, tests are not always available to detect all infections.
Thus, infections “create” cases, and these tests detect cases. President Donald Trump came close to the truth when he said that the number of cases in the US was high because a lot of testing had been done. But had he bothered to explore the subject? After all, more tests performed do not result in a greater number of cases, but only allows for a more accurate estimation of their true number.
- Ministry: most patients in Poland are asymptomatic
From an epidemiological point of view, the best strategy is to test as many people as possible. In this way, let’s minimize the discrepancy between cases and infections.
We cannot compare the deaths with those on the same day
Estimates vary, but the time from infection to eventual death can be up to a month. However, the time that passes until recovery is even longer. Some people get seriously ill and recover for a long time, others have an infection asymptomatically.
Thus, the deaths recorded on the day the chart was drawn reflect those recorded a few weeks earlier. Back then, the number of cases could have been half the number on the day in question.
- Do children get seriously ill and die from the virus? New research results
The short time to double the number of cases and the long recovery time lead to a large discrepancy between the number of sick and recovering. However, according to Dr. Raubenheimer will not know the real data soon.
Our data is disordered, incomplete and subject to change
Some social media users get frustrated to see the stats being updated on the fly. Various conspiracy theories proliferate. And only a few realize how gigantic and complex the task is to track the statistics on the coronavirus.
Different countries and even regions may count their cases and deaths differently. Information gathering takes time, which means that retroactive adjustments are made.
We will have a long time to wait before we know the truth about this pandemic. On the other hand, the graphs that were created at the beginning are wrong, not because the people who created them wanted to deceive the world, but because they did not have enough data, argues Dr. Raubenheimer.
See also:
- Are we dealing with a second wave of COVID-19 in Europe?
- Prof. Ryszarda Chazan: “I find it hard to believe how easily it was possible to stop the whole world with top-down decisions”
- How many tests are done in individual provinces? We have the latest data
- More and more coronavirus infections in Europe. More countries with records