Who ride bugs? A detailed look at the results of a large test room

A study of immunity to covid-19 revealed only 107 positive cases from about 26,500 people tested. Without the flag of them, the design was roughly aunty. We will find out how many people in the test room could fail from the next phase of the study.

At the press conference on May 6, we learned about the first expected preliminary results of the study of immunity plots, which took place at the end of April. As you no doubt know, the resulting value was very low, in breaths of percentage units. This corresponded to both ordinary Czech experts and results from other countries.

According to the people with counterparts, even in the most affected localities it reaches low units of percent. The data can be changed to a certain extent, because the process of their breeding is still going on, but the cars still have their quality for a while, so let’s not expect big changes.

The test, which indicates an outbreak in the past, was positive in only 107 people. None of them thought they had an acute SARS-CoV-2 infection. If they sold the disease, 70 percent of those who tested positive wanted to know. This was a little more (six percent) not in the whole groups of respondents, according to the survey. Twenty percent of them were abroad this year when they could become infected.

According to nurses, the population is promoted only in per mille. Ladislav Duek, Director of the State of Health Information and Statistics, had to think about and prepare adequately for the next waves of the pandemic, presenting the result. The Minister of Health, Adam Vojtch, has even changed the fact that, despite the obvious outbreaks of the disease, the people will have to stumble in the coming years.

The group of positive testers was not statistically particularly interesting. Both sexes were represented by dark toton among people with a positive test. In terms of age, they formed the same group from 40 to 49 years and from 60 to 69 years. According to Ladislav Duk, the bottom statistical value cannot be attributed to this bag. A sample of 107 people is more than that, he said.

Somewhat more interesting was the data on the past health status of people who had a positive test result. The occurrence of covid-19 was reported by less than half of those tested positive. Mostly lo o kael and called temperature. Two fifths of the positive testers then experienced a sore throat, a difficult breath.

The remaining 51 percent of those who unknowingly wished for the disease did not have any symptoms. The overall pilot estimate of asymptomatic courses is between 27 and 38 percent, estimated by the Duek study.

What about St.

What about the state of health was reported by people who, in the Czech study, had a positive effect on the presence of counterattacks. When evaluating, keep in mind that some people in this group probably didn’t actually own the infection due to test errors.

Of course, I certainly don’t give this a damn. Because the study went back, the cars had to rely on what people remember about their health. They did not have an objective way to find out how they get (slightly increased temperature, kael, etc.) so far. We are not sure, I think, because of this relatively small time two people pay close attention to their health and watch their health, said one of the study leaders Marin Hajdch from Palacký University in Olomouc for iDNES.cz.

If we are to accept this result, it means that there are more insignificant symptoms than with how many epidemiological models of the advisory darkness of the government. The ZIS model assumed that the number of asymptomatic people who themselves probably could not be ill is about 10 percent of the total sweat. If the ratio should be different, the model will completely change this.

We can show it on the treasures that Ladislav Duek showed at the press conference on April 8. At that time, in addition to the common model, he also presented to the public a hypothetical model for a situation where 30 percent of them are. That is, a value that, according to the findings of the study, is relevant (not certain).

Model modified

The model then assumed that the total number of nakaench would increase by the end of April. At the end of April, according to him, there could be somewhere between 15 and 25 thousand people with a confirmed disease. Let’s compare it with a graph from the same prints of the presentation, which ran with 10 percent of nakaench. He estimated that by the end of April there should be about 9 and 12.5 thousand cases in R:

Prediction chart in

Undoubtedly, both models varied, the detected case of the disease was about seven thousand, ie below the lower limit of the stated variance. It is interesting to note that both of them completely differently determined the development of sweat-free infections (in both graphs ed kivka), and thus the future of SARS-CoV-2 virus infection in the Czech Republic.

In the case of a 10% proportion of healthy patients, all cases of infection should be gradually revealed. At 30%, the undetected case of the disease should increase steadily, even relatively quickly. The reproductive rate (ie the number of people who on average infected one patient) should be above 1, more precisely between 1.17 and 1.35. This is clearly not the case: the sweat of the disease and the sweat of hospitalized patients with confirmed covid-19 disease would have to be severe, not today.

The monch is explained in full. For example, data from astnk studies on their health are not really accurate. I’m wrong somewhere else. For example, it will be interesting to see what percentage of asymptomatic cases ZIS models will work in the coming weeks.

Superpesn tests?

A very low number of positive tests indicate that there were very few cases in the study where the test mistakenly showed that a person who had not contracted the infection had antibodies. These so-called false positive results can completely distort the results of large studies, because their number can significantly strengthen the number of accurate results even with relatively reliable tests, and it seemed that this could be the case even in the case of a Czech project.

Prior to the start of the test, only information was available that the tests used initially show at least 95 percent of the negative cases as negative (so-called specificity). Which means that even five percent of cases should prove wrong. In the case of a test facility of 27,000 people, this would mean about 1,300 false results. Assuming that I had the occurrence of detectable opposites between the tested, the number of these so-called fallen positive results would be significantly higher than the number of actual catches.

What obviously did not happen, because the positive result was only 107. How is it possible? The reason, according to data published in the press conference with the preliminary results, is very simple: the test was much more accurate, not the original.

The manufacturer guaranteed a specificity of about 95 percent, and this data was so published. In the dark around the study, he performed a pouitch test from Wantai on several hundred blood samples from the time when SARS-CoV-2 jet did not occur. In none of the cases did the tests show a positive result. In this particular case, the rapid tests used for the study proved to be 100% accurate.

Let’s add now, in practice it will be so. In fact, the tests will undoubtedly be inaccurate and false alarms will undoubtedly show. Due to the large number of test subjects, it cannot be ruled out that most of the positive results of the test are false or all of them. From a statistical point of view, it is therefore correct to state that the test results in the city show the rate of occurrence of the disease somewhere between 0 and X percent (even in the case of 0, X percent). So do the following:


Preliminary results according to people who had a positive result in the rapid test for the presence of antibodies against SARS-CoV-2 in a pre-test in R at the end of April.

From a practical point of view, however, the difference between 0 and 0.5 percent is negligible. According to the positive results, it is clear that the problem will not be great and these specific tests do provide a very small number of false alarms. It is a leap from the first tried-and-tested speed tests in the Czech Republic, which had a significant reliability. This indicates that there is a quality technology on the market against the new coronavirus, which is also improving rapidly in many respects.

An interesting question at this point is whether the rapid tests will actually detect everyone who has contracted the SARS-CoV-2 virus infection. i spe, how many do not catch.

How many did not appear in the test?

Meanwhile, the test could pass the infection due to the unreliability of the test. Both simply do not show that the tested counterparts m. But this is especially a negligible problem. Accuracy (first sensitivity) in this regard, according to the manufacturer, should again be around 95 percent. It is also possible that the test passed the people who had the counterparts.

Interesting, and for the further course of the epidemic, the question is how many people have survived the infection so that they have not developed protests. Otherwise, how many of them developed so easily with the infection that they did not even allow for the formation of antibodies. Rapid tests revealed by counterparts (so-called IgM and IgG) arise in the body with a certain bottom.

Count p

Graph of the presence of antibodies (specifically IgM) in a group of hospitalized patients with proven SARS-CoV-2. As can be seen, protests appear and less than two weeks after the expected infection.

According to the results of tests performed for the study (also confirmed in other countries), it can be roughly estimated that in most patients patients develop anesthesia the day after infection.

However, it is possible that a number of patients get rid of the infection for two, ie before that, with counter-emergence. In such a case, it is significantly more difficult to prove the disease, it requires a simple and thorough testing, so we do not yet have accurate information on how many such cases there are. Which is, of course, from the point of view of the previous development of the epidemic and the set of measures against them to a small extent. The preliminary results of the study cannot help to estimate the total number of patients in the Czech Republic, as it sometimes appears in the first results. They can estimate the number of people with protests that are not tot.

There are these signs, including the unconfirmed data of some Czech figures, that according to them may be quite significant. According to some hypotheses, there will be many cases, especially among children and adolescents, their immune systems work a little differently than in adults and who deal with SARS-CoV-2 infection in most cases very easily.

The SARS-CoV-2-CZ-Preval study did not provide any answers in this regard, but it should in the future. In Olomouc, Litovel and Uniov, not only speed tests were performed, but blood vessels were collected. One of them was obtained, which first ends up in freezing samples, so that sometime in the near future it will go to the laboratory.

Then the samples should be thoroughly examined, which, in contrast to the rapid tests, determines not only whether the antibodies are present, but (at least to a certain extent) and how many of them the hunter has (which indicates how much the infection progresses). First of all, they should reveal whether someone sold the infection without a rapid test of the detected antibodies (ie IgG and IgM).

We don’t know exactly when to do it [vyeten] occurs, to Marin Hajdch. At this point, we still decide what technology to use. It is not for us to do it as quickly as possible, but to do it right. Otherwise, it is said that researchers are still looking for reliable and accurate methods, similar to how they used the use of rapid tests before the start of the study.

In this case, it is not only a matter of accuracy, but also a matter of labor: a thorough test for the presence of antibodies is relatively laborious, and thus looks for at least some automated procedures that would allow 10 thousand blood samples to pass quickly.

It is not clear when the round could have resulted. The authors of the study make it quite clear that it will be more in the case of two units of units than units of weeks. Until then, our curiosity will have to give coronavirus uncertainty. Due to the very favorable development of the last few weeks, perhaps the time will not be too long.