Immunological detection of serum antibodies in pediatric medical workers exposed to varying levels of SARS-CoV-2

Dear Editor,

Since the initial reports of COVID-19 disease outbreak in Wuhan, China, it has continued to spread rapidly with cases identified in virtually all countries, worldwide.¹ The population is generally susceptible to SARS-CoV-2, including children and pregnant women, and medical staffs are a high-risk population for this disease. In this journal, Chen et al. have reported the high SARS-CoV-2 antibody prevalence among healthcare workers exposed to COVID-19 patients.² Here we would like to share our finding about the serum antibodies analyzed in a special group of pediatric medical workers exposed to varying levels of SARS-CoV-2 after Wuhan severe epidemic of COVID-19.

A preliminary study suggests children can be infected with SARS-CoV-2 like adults but are less likely to be symptomatic or develop severe symptoms.^{3 , 4} The asymptomatic or mildly symptomatic children might transmit the disease.⁵ Therefore they are tested for SARS-CoV-2 less often than adults, leading to an underestimate of the true numbers of children infected.⁶ Laboratory tests play a pivotal role in the diagnosis and management of COVID-19; the current gold standard being real-time reverse transcription polymerase chain reaction (rRT-PCR) on respiratory tract specimens.⁷ The measurement of specific COVID-19 antibodies (both IgG and IgM) should serve as an additional, non-invasive tool for disease detection and management, especially in patients who present late, with a low viral load. Due to the high infection rate of medical workers and the uncertainty of child-to-person transmission, we chose a special group of pediatric medical workers as the research subjects to investigate their infection status with SARS-CoV-2 and analyze possible causes. This study also helps clarify the potential of different immunological techniques for antibody detection as an auxiliary diagnosis of COVID-19.

On March 19–20, 2020, pediatric medical workers (n = 325) in one hospital but not the designated hospital for COVID-19 in Wuhan were recruited. They were divided into three groups depends on their level of contact with confirmed and/or suspected COVID-19 cases during the outbreak: i. close contact group (contact with confirmed and/or suspected cases of COVID-19), ii. non-close contact group (contact only with non-COVID-19 patients), and iii. non-contact group (no contact with any patients). Three different immunological detection methods were used to measure SARS-CoV-2 serum antibodies: colloidal gold-based detection, enzyme-linked immunosorbent assay (ELISA), and dual-target immuno-fluorescence assay (DTFA) (details in the Supplementary methods). The overall positive rate for SARS-CoV-2 IgG and IgM antibodies in the pediatric medical workers was 43.08 and 5.85%, respectively. For the close contact, non-close contact, and non-contact groups, respectively, the DTFA positive rates for IgG were 41.36, 14.68, and 12.50% (p < 0.05), and the ELISA positive rates for IgG were 34.55, 10.91, and 4.17% (p < 0.05) and 8.38, 0.91, and 0% for IgM (p < 0.05). Colloidal gold detection results were negative for IgG and only two participants tested positive for IgM, both in the close contact group (Table 1 ). It suggests the colloidal gold detection kit used in this research is not sensitive enough to be useful in accurate antibody detection, whereas the DTFA and ELISA positive rate performed similarly.

Table 1.

Test results of serum antibodies in pediatric medical workers exposed to different levels of SARS-CoV-2.

			DTFA			ELISA				Colloidal Gold Detection
	Total positive rate of IgG%	Total positive rate of IgM (%)	rN-IgG	rRBD-IgG	Total positive rate of IgG (%)	rN-IgG	rN-IgG positive rate (%)	rRBD-IgM	rRBD-IgM positive rate (%)	IgG positiverate (%)	IgM positive rate (%)
All participants	43.08 (140/325)	5.85 (19/325)	1076.52±1153.14	738.42±988.99	30.25 (98/324)	0.13±0.16	24.31 (79/325)	0.06±0.04	5.23 (17/325)	0.00 (0/325)	0.62 (2/325)
Close contact group	58.12 (111/191)	9.42 (18/191)	1308.98±1323.92	958.49±1188.43	41.36 (79/191)	0.16±0.20	34.55 (66/191)	0.06±0.05	8.38 (16/191)	0.00 (0/191)	1.05 (2/191)
Non-close contact group	22.73 (25/110)	0.91 (1/110)	784.02±791.23	434.45±463.47	14.68 (16/109)	0.08±0.06	10.91 (12/110)	0.06±0.04	0.91 (1/110)	0.00 (0/110)	0.00 (0/110)
Non-contact group	16.67 (4/24)	0.00 (0/24)	587.58±362.93	385.63±282.21	12.50 (3/24)	0.07±0.04	4.17 (1/24)	0.05±0.02	0.00 (0/24)	0.00 (0/24)	0.00 (0/24)
F or χ2	43.02	10.80	10.28	12.35	27.29	9.95	26.93	1.36	9.28	–	1.41
p	0.00	0.01	0.00	0.00	0.00	0.00	0.00	0.26	0.01	–	0.49

We further conducted a multivariate logistic regression analysis using antibody results as the independent variables to investigative the relationship of positive serum antibody results, with the performance of aerosol procedures, exposure levels to COVID-19 cases, clinical symptoms (including fever, cough, headache, stuffy nose, runny nose, sneezing, pharyngalgia, diarrhea, fatigue, etc.), chest CT imaging changes, and age of participant (Table 2 ). The results showed that participants who had performed an aerosol procedure had a 2.70-fold higher risk of testing positive, and with each additional level of exposure to COVID-19, the risk of testing positive for antibodies increased 5.26-fold. None of the antibody positive participants contained neutralizing antibodies in their serum maybe cause of the low viral load exposure.

Table 2.

Multivariate logistic regression analysis of positive antibody tests.

Variables in the equation
		B	SE	Wald	df	Sig.	OR	95% Confidence interval of OR
								Lower	Upper
Step 1	Aerosol operation	.992	.297	11.118	1	.001	2.696	1.505	4.828
	Exposure levels	1.660	.298	31.142	1	.000	5.262	2.937	9.428
	Clinical symptoms	.109	.281	.150	1	.698	1.115	.643	1.935
	Chest CT imaging changes	.492	1.028	.229	1	.632	1.636	.218	12.267
	Age	.015	.017	.754	1	.385	1.015	.982	1.048
	Constant	−5.846	.871	45.015	1	.000	.003

B: Regression coefficients; SE: Standard error; Wald: Chi-square value; df: Degrees of freedom; Sig: Significance; OD: Odds Ratio.

After one more month at the end of April, 70 of the 325 participants who had a positive result using any of the above test methods volunteered to participate in a retest of IgG/IgM for SARS-CoV-2. Positive results were observed for 33 of the 70 (47.14%) participants when they were first tested for rN-IgG and rRBD-IgG by DTFA, and 30 of the 33 (90.91%) participants became negative or weakly positive by the same detection method one month later. Meanwhile, 47 of the 70 (67.14%) participants tested positive when they were first retested for rN-IgG by ELISA; 41 of the 47 (87.23%) became negative or weakly positive when similarly tested again one month later. Eleven of the 70 (15.71%) participants tested positive when they were first tested for rRBD-IgM by ELISA, 4 of the 11 (36.36%) became negative or weakly positive one month later. Although we cannot clearly track antibody kinetics for asymptomatic infections, we can observe that the majority of participants with positive IgG antibodies had a significant decline in antibody levels after one month. That means the SARS-CoV-2 antibodies diminish to near undetectable levels within two months.

This research revealed that pediatric medical workers are a high-risk group for infection by SARS-CoV-2, and the higher the exposure levels to COVID-19 patients and aerosol production, the greater chance of being infected. Meanwhile, we found that pediatric workers had lower levels of IgG antibodies than patients with COVID-19.² Children-to-person transmission is almost inevitable, but pediatric medical workers often have no or only mild clinical symptoms and also cannot produce enough antibodies to neutralize the virus. The antibody protection that healthcare workers obtained after infection by SARS-CoV-2 in this study, could not be maintained for a long time and no NAbs were detected to provide them with sufficient protection.

Declaration of Competing Interest

None.