"Antigen testing needs to be measured again after a day to ensure a greater detection rate."
Written by | Ling Jun
Source | "Medical Community" public account
"Blow off the snot, take out the nasal swab from the box, insert it into the nasal cavity and rotate it around. The slightly wet swab head also carries mucus, immerses itself in the preservation solution and rotates, pressing the outer wall of the sampling tube until the transparent liquid gradually becomes cloudy. ”
"Drop the liquid vertically into the test card sample hole and wait... Anxiously waiting... Until the test card shows a bar, breathe a long sigh of relief. ”
Over the weekend, tens of millions of Shanghai residents were doing the same thing: poking themselves in their nostrils at home for antigen testing. On March 27, according to a press conference on the prevention and control of the new crown pneumonia epidemic in Shanghai, more than 14 million people had been tested for antigens as of that day.
But earlier, on March 15, Li Jinming, deputy director of the Clinical Laboratory Center of the National Health Commission, reminded at a press conference that antigen testing cannot replace nucleic acid testing. Antigen testing should be used in high-risk, high-prevalence clusters of infected populations.
Is antigen testing accurate? How much can it play in epidemic prevention and control? Will the problem of false positives and false negatives make antigen testing a "chicken rib"? In recent days, the domestic epidemic situation has continued to spread at many points, which has also become a topic of concern for most people.
300,000 false positives?
Concerns about "false positives" are mainly based on the extrapolation of theoretical data.
Taking a reagent used in this round of antigen detection in Shanghai as an example, the sensitivity disclosed in the English instructions was 97.01%, and the specificity was 98.8%.
Specificity, i.e., accuracy in correctly diagnosing an uninfected population. According to the current cumulative number of confirmed cases in Shanghai, assuming that the new crown infection rate is about 0.08%, that is, about 20,000 people out of 25 million people are infected, combined with 98.8% of the reagent specificity, the antigen test of all employees will cause about 300,000 false positives. Based on sensitivity calculations, about 19,000 real infected people can be detected.
This is similar to Li Jinming's view at the press conference:
The prevalence of the new coronavirus in most parts of the mainland is less than one in a million, if you take a kit with 85% sensitivity and 97% specificity, if you screen in cities with tens of millions of people, you will get 300,000 positives, but only 9 of these 300,000 positives are true, and the negative results are reliable.
On the 26th, a street in Shanghai self-tested 110,000 people, using the reagent of Icon Biotech, 10 samples were positive, of which 1 sample was positive for nucleic acid, that is, 9 samples were false positive, and the false positive ratio was one in ten thousand.
Another region used the reagent of Oriental Biology, which tested 730,000 people, detected 146 positive samples, 58 samples were positive for nucleic acid, and the false positive ratio was also about one in 10,000.
What does this data mean? The "SIFIC Infection Horizon" article interprets and assumes that the covid-19 infection rate in Shanghai is 0.04%.
According to the following calculation formula (with nucleic acid detection as the gold standard):
Icon Bio reagent
False positive rate = [b/(b +d)] * 100% = 0.0082% (specificity = 1 - false positive rate = 99.9918%);
False negative rate = [c/(a+c)] * 100% = 97.73% (sensitivity = 1 - false negative rate = 2.27%)
Oriental Biological Reagent
False positive rate = [b/(b+d)] * 100% = 0.012% (specificity = 1 - false positive rate = 99.988%)
False negative rate = [c/(a+c)] * 100% = 80.14% (sensitivity = 1 - false negative rate = 19.86%)
Although the total COVID-19 infection rate in the above calculation model is only an estimate, the antigen test produced only 9 and 88 positive and false samples in 110,000 and 730,000 people, respectively, which are very low according to the current degree of COVID-19 prevalence in mainland China, and the specificity is far better than the data on the reagent specification.
Chang Rongshan, one of the earliest virology experts in China to call for "antigen self-testing", told the "medical community" that the questioning of theoretical high false positive data is in line with the international academic view that the use of "antigen self-testing" is less effective in an environment with extremely low incidence.
However, unlike overseas, Chinese mainland is currently adopting a "dynamic zeroing" epidemic prevention strategy. Chang Rongshan believes that "false positives" are not terrible, even if the theory that Shanghai may produce hundreds of thousands of false positives, it only needs to pass a second review.
"In the early stages of the epidemic, the 'initial screening' of antigen detection can quickly find sporadic transmission points. Once the targeted secondary nucleic acid test finds that it is a false yang, the relevant risk is released. This is the advantage of nucleic acid testing for all employees who queue up for sampling and take an average of 2 days to know the results. Chang Rongshan said, "At the same time, one more round of nucleic acids for all employees means that there may be an extra risk of cluster infection, and it will also consume a lot of manpower and material costs." ”
The key to dynamic zeroing is to quickly buy time, and in this regard, "antigen self-test" has an irreplaceable advantage. Wu Qianyu, a first-level inspector of the Shanghai Municipal Health Commission, also said at a previous press conference that residents in key areas carried out nucleic acid screening, and residents in non-key areas carried out antigen testing. The aim is to increase detection rates and detect hidden infected people in a timely manner.
But it is inevitable that a false positive will cause panic in the tester himself, and even the entire community residents. Therefore, Chang Rongshan reminded that in case it is found positive in the antigen self-test, residents do not have to be too anxious and nervous, just report in time and wait for the second review according to the relevant requirements.
In the face of the Amikeron mutant strain,
Antigen detection accuracy may be reduced
More noteworthy is "false negative", because sensitivity = 1 - false negative rate, high false negative rate means low sensitivity.
According to the 0.04% new crown infection rate in Shanghai, about 44 and 292 people should be infected among 110,000 and 730,000 people, respectively, but the actual number of infections detected by the two reagents was only 1 and 58 people, respectively, and the false negative rate reached 97.73% and 80.14%, which was much higher than the values on the reagent instructions (2.99% and 2.75%).
The above picture is taken from the English instruction manual of Aikang Biology, the source: "Zhongzhong Kanyao" public number
The above picture is taken from the antigen self-test instruction manual exported by Oriental Biologics to Japan, the source: "Zhongzhong Kanyao" public number
If the false negative rate is really so high, it means that large-scale antigen detection is not only difficult to play a role, but also consumes a lot of social costs. The main possible reasons for the above "prudent" data include, but are not limited to:
There is a problem with the quality of both reagents.
Residents lacked supervision of self-tests and problems with operations, making it impossible for infected people to be detected.
The 0.04% COVID-19 infection rate does not apply to all areas of Shanghai. The two regions mentioned above have done a good job in epidemic prevention, and the actual number of infected people is very small.
Characteristics of the Omikejong mutant strain.
Because the marketing of the two antigen detection reagents has passed the formal approval, and the numerical difference is too large, the possibility of "quality problems" is very small.
At the same time, Chang Rongshan told the "medical community" that when the new crown virus was infected on the 3rd, 4th and 5th days, the viral load was very high, even if it was improperly operated, it only needed a little sample to detect it.
"Two days before infection, the viral load was still very low, and it was difficult to detect a single antigen test, but repeated testing 2 days apart could be done, and this should be taken into account when testing at home." Chang Rongshan therefore suggested that antigen testing should be retested after one day to ensure the detection rate to a greater extent.
For the third possibility, according to the epidemic information released daily by the Shanghai Municipal Health Commission, most of the infected people are found in isolation and control. This means that the infection rate on the social level may be much lower than 0.04%, and the above two areas may actually have few infected people.
According to the sensitivity of the two reagents, the infection of 1 person and 58 people determined by the final secondary review is basically equivalent to the actual number of infections.
But regardless of the true false negative rate, the sensitivity of antigen detection may have declined in the face of the onset of the Aomi Kerong mutation. As early as the beginning of this year, overseas academics have begun to discuss "how much effect antigen detection reagents have in the face of the Aomi Kerong mutant strain".
On January 15, 2022, American scholars published an article on the preprint platform medRxiv, facing the Omiljung variant, two widely used antigen detection reagents overseas seem to lag by a few days, and many people who have been infected still test negative.
On the 22nd of the following month, German scholars published a paper on Immuned detection of omicron by SARS-CoV-2 rapid antigen tests, which evaluated 9 antigen detection kits and showed that most of the reagents had a decrease in their ability to detect the Semicroton strain compared with the Delta strain.
Some overseas scholars pointed out that the Omikejung variant may be more easily detected in the throat, and in January 2022, The Reuters news agency reported that the Israeli Ministry of Health said that when using rapid antigen testing, it was also necessary to wipe the throat.
However, the relevant basis has not been widely confirmed by the academic community.
Li Jinming previously mentioned in an interview with Xinhua News Agency that for those who have negative antigen self-tests but have symptoms, it is recommended to go to a medical institution that has set up a fever clinic as soon as possible for nucleic acid testing.
If it is inconvenient to seek medical treatment, you should self-isolate at home, avoid going out, and conduct a self-test of antigens once a day for 5 consecutive days.
bibliography:
4.Discordant SARS-CoV-2 PCR and Rapid Antigen Test Results When Infectious: A December 2021 Occupational Case Series，https://www.medrxiv.org/content/10.1101/2022.01.04.22268770v1
Source: Medical community
Editor-in-charge: Zheng Huaju
Proofreader: Zang Hengjia
Plate making: Xue Jiao