Generally, coronavirus genomes encode four major structural proteins, namely spike (S), envelope (E), membrane (M), and nucleocapsid (N). Current serologic tests have been developed to target antibodies directed against these antigens, showing that spike protein-based detection was more sensitive than nucleocapsid protein-based detection 26,27. In our study, we used the C.I.A. method, which was based on the recombinant SARS-CoV-2 S-RBD protein to detect serum IgG and IgM, and its analytical performance was successfully evaluated by Wan Y et al. They reported the performance verification of the SARS-CoV-2 IgM (82% sensitivity and 93.85% specificity) and SARS-CoV-2 IgG (86% sensitivity and 96.92% specificity) detection kits among COVID-19 patients 25.
However, due to the problems of immunological detection methods, there have been interferences attributed to certain pathological factors, biological factors, and cross-reactions, resulting in false-positive results that were inconsistent with clinical manifestations and epidemiological characteristics. Some of these factors identified by previous studies included inadequacy during any step of the testing process, presence of cross-reactive antibodies, other endogenous interference factors, and other viral infections 24,28−29. Additionally, some false-positive cases likely did not result from problems with the sample, procedure, or other random factors, which was supported by obtaining repeated positive results with similar S/CO values on repeat testing (data was not shown) in our study, making a transient response to antigen less likely. Furthermore, this phenomenon regarding endogenous interference factors in SARS-CoV-2 antibody testing was also reported in our previous study 24. Although we can use the electronic medical records and laboratory results, including N.A.T.s, as a source to determine true anti-SARS-CoV-2 status, the procedure is labor-intensive and time-consuming. Therefore, we attempted to seek an effective strategy to solve such problems when confronted with false-positive results in the SARS-CoV-2 antibody screening test.
To elucidate these issues, we retrospectively analyzed the false-positive cases of SARS-CoV-2 IgM, and IgG detected using C.I.A. This study showed that the false-positive rate of the single SARS-CoV-2 IgM positive results was 95.88%, which was significantly higher than those of the single SARS-CoV-2 IgG positive results (67.50%) and SARS-CoV-2 IgM & IgG positive results (29.55%). Therefore we concluded that the possibility of false-positive of the single SARS-CoV-2 IgM positive and single SARS-CoV-2 IgG positive results was high, and the combined detection of SARS-CoV-2 IgM and IgG antibody was better than the single detection in terms of a positive detection. Previous investigations have shown that SARS-CoV-2 IgM and IgG antibodies could be detected as early as the 4th day following symptom onset 30. The positive rates of the single SARS-CoV-2 IgM, single SARS-CoV-2 IgG, and SARS-CoV-2 IgM and IgG positive results among COVID-19 patients were 1.72%, 3.45%, and 94.83%, respectively, concluding that the combined detection of IgM and IgG had better practicability and sensitivity than IgM or IgG alone 31. Interestingly, most false-positive signals were detected in the SARS-CoV-2 IgM assays, which other studies have also noted 28,32. Thus, the combined detection of SARS-CoV-2 IgM and IgG should be given high priority in its implementation as the standard serological test in clinical and public health practice during the pandemic.
In this study, we found that the S/CO values of the IgM false-positive results were mainly between 1.0 and 3.0, whereas the S/CO values of the IgG false-positive results were mainly between 1.0 and 2.0. These results indicated that the SARS-CoV-2 IgM and IgG false-positive results detected by C.I.A. mainly existed in the low-value area. The false-positive results, which were low positive or low-value, needed to be confirmed further. Therefore, the S/CO ratio may be a helpful indicator in differentiating false positives from true positives. Aside from the S/CO values, we also compared the false-positive proportions of the single SARS-CoV-2 IgM, single SARS-CoV-2 IgG, and SARS-CoV-2 IgM & IgG positive results in different sexes and ages. Our study showed no significant differences for the false-positive proportion in different sexes and ages.
After SARS-CoV-2 invades the human body, the time and duration of producing IgM and IgG antibodies are different. Due to the dynamic change of specific antibodies, cases with a single positive SARS-CoV-2 IgM/IgG antibody test can be determined by dynamically monitoring them over some time5. Despite this, the question of how long the supposed "antibody" level could last in the false-positive cases remains unknown. To investigate this, we reviewed the subsequent results of four true-positive cases and 25 false-positive cases, including 18 cases with a SARS-CoV-2 IgM positive result and seven cases with a SARS-CoV-2 IgG positive result, and dynamic monitoring of the serum "antibody" levels after the first test. It was found that the time of conversion to seronegativity in IgM false-positive cases was 4 to 19 days, with a median seroconversion time of 9 days. Moreover, the "antibody" duration in the IgM false-positive cases was significantly shorter than that in COVID-19 patients with a duration of 2 to 6 months, as reported in other studies 13,30. Meanwhile, the conversion time to seronegativity in IgG false-positive cases was 2 to 8 days, with a median seroconversion time of 5 days. Compared with previous studies, the "antibody" duration in the IgG false-positive cases was also significantly shorter than the duration of serum IgG antibodies in COVID-19 patients (6 months) 13,30. For other cases observed during the monitoring period and those that did not turn seronegative, "IgM antibody" levels showed a downward trend or remained unchanged. Similarly, "IgG antibody" levels in those same cases also showed a downward trend. Due to the lack of blood samples collected from the false-positive cases in the later stage, the time of conversion in their "antibodies" remained unknown.
In this study, the results suggest that the dynamic monitoring of serum antibody level was also of practical value in distinguishing between true-positive and false-positive results. When confronted with positive results in SARS-CoV-2 specific antibody testing, these findings should be comprehensively judged. First, it is crucial to observe the antibody pattern. If it is a single SARS-CoV-2 IgM positive pattern, the probability of a false-positive is higher than that of a single SARS-CoV-2 IgG positive pattern. Meanwhile, if it is a SARS-CoV-2 IgM & IgG positive pattern, the probability of a true-positive is higher than that of a single positive antibody result. Second, one must observe the S/CO value distribution of the antibody results as follows: (1) if the S/CO value of a single SARS-CoV-2 IgG positive result was between 1.0 and 2.0, the probability of a false-positive is approximately 94.74% (Table 5); (2) if the S/CO value of a single SARS-CoV-2 IgM positive result was between 1.0 and 3.0, the probability of a false-positive is approximately 100.0% (Table 4); (3) if a single SARS-CoV-2 IgM has a positive result, and the S/CO value is very high (> 20.0), the result needs to be judged in combination with the S/CO value of the SARS-CoV-2 IgG result; (3a) if the S/CO value of the SARS-CoV-2 IgG result approaches 0, there is a high possibility of a false-positive result; (3b) if the S/CO value of SARS-CoV-2 IgG result is close to 1.0, it is more likely to be a true-positive result. Lastly, antibody level changes should be observed dynamically. In cases of a single SARS-CoV-2 IgM positive status, IgM antibody increases and IgG antibody turns positive (i.e., IgM and IgG positive status), with which we can judge if the patient truly has a SARS-CoV-2 infection. Otherwise, it is a false-positive result. On the other hand, in the dynamic monitoring of SARS-CoV-2 positive IgM and positive IgG status, the IgM antibody is the first to increase and subsequently decrease, whereas the IgG antibody titer has a 4-fold increase, with which we can judge if the patient truly has a SARS-CoV-2 infection 5. Otherwise, these are false-positive results. Furthermore, the single SARS-CoV-2 IgG positive status shows a rapidly decreasing trend in dynamic monitoring, which may indicate a false-positive result.
Despite the findings of our study, several limitations were noted. First, due to the insufficient conditions of our laboratory, we failed to determine the interferences or factors causing false-positive results. Second, the number of samples included in this study was limited, leading to some deviation in the analysis results. More cases should be included in further studies of the same topic. Third, since the majority of the SARS-COV-2 specific antibody detection was performed using the C.I.A. platform due to its high throughput, our research analysis was only focused on the C.I.A. Methods and molecules used for generating and detecting signals. The epitopes and specificities of antigens and antibodies are different between the assays. Thus, the characteristics of the false-positive results analyzed in this study may not apply to other SARS-COV-2 antibody detection methods. Regardless, this study can still provide a reference strategy for other researchers.