Positive Predictive Value of Rapid Antigen Tests in School During SARS-CoV-2 Omicron Variant Surge

Abstract

In Korea, we conducted a national observational study to calculate the positive predictive value of SARS-CoV-2 rapid antigen tests in K-12 schools during the Omicron variant surge in March 2022. The weekly positive predictive value ranged from 86.4% to 93.2%. The positive predictive value was the highest among elementary school students with symptoms (95.7%) and lowest among teachers/staff without symptoms (70.9%).

The coronavirus disease 2019 (COVID-19) has led to the closure of schools worldwide, resulting in significant disruption in children’s learning.¹ In South Korea, schools K-12 and universities opened in the spring of 2022 during the Omicron variant surge.² The Ministry of Education introduced guidelines outlining the use of random screening in the rapid antigen test (RAT) for symptomatic and asymptomatic students and teachers/staff.³ RAT aims to provide timely identification of infection, but there are concerns regarding the positive predictive value of testing.⁴ This study investigated the true and false positives of the RAT in a large sample of students and teachers/staff serially screened throughout South Korea during the SARS-CoV-2 Omicron variant surge.

METHODS

Since March 2022, nationwide RAT screening for SARS-CoV-2 has been implemented in kindergartens (4–6 years of age), elementary schools (grades 1–6), middle schools (grades 7–9) and high schools (grades 10–12). Students with and without symptoms were tested twice weekly, and teacher/staff were screened once weekly. Those who had a close contact history with patients with SARS-CoV-2 were screened 3 times per week voluntarily. Each student was asked to report about symptoms (multiple choice), including fever, chills, cough, headache, myalgia, sore throat, runny nose, loss of taste/smell, fatigue, nausea, vomiting or diarrhea. The RAT results were reported through the Student Health Self-Diagnosis App developed by the Ministry of Education. If the RAT result was positive, the patient would be referred for a confirmatory polymerase chain reaction (PCR) test within 72 hours. We merged the symptoms and RAT results data with the PCR data collected between March 2 and April 3, 2022. Individuals with incomplete information (ie, input errors, blank dates) and a history of SARS-CoV-2 infection were excluded from the analysis. The RAT results were compared with the PCR results, and the positive predictive value and binomial 95% confidence intervals were calculated. The weekly positivity rate of COVID-19 in the RAT and PCR was estimated by dividing the number of enrolled students and teachers/staff with COVID-19 by the total number of tested individuals.

The study was reviewed and approved by the Korea Disease Control and Prevention Agency Institutional Review Board (2021-12-03-PE-A).

RESULTS

Between March 2 and April 3, 2022, of the 2,822,544 RAT results collected, 136,902 (4.9%) were positive, of which 126,042 were confirmed by PCR, resulting in a positive predictive value of 92.1% (Table 1). The RAT positivity rate ranged from 68.7% to 83.8% among individuals with symptoms, while the rate ranged from 0.3% to 0.7% in individuals without symptoms (Table 1). The weekly SARS-CoV-2 RAT positivity rate in South Korea ranged from 4.9% to 5.9%, whereas the positive predictive value of RAT ranged from 86.4% to 93.2% (Figure 1, Supplemental Digital Content 1, https://links.lww.com/INF/E851). The positive predictive value was the highest among elementary school students with symptoms (95.7%) and lowest among teachers/staff without symptoms (70.9%) (Table 1). In all groups, the RAT positive predictive value was higher in patients with symptoms than that in those without symptoms (Fig. 1).

TABLE 1. - Result of Nationwide School COVID-19 RAT Screening Reflexed to PCR Confirmation During SARS-CoV-2 Omicron Variant Surge in South Korea, March 2 to April 3, 2022

Symptom*	Status		Tested Persons	RAT Screening				Reflex PCR Confirmation†
				Positive		Negative		Positive		Negative
				n	%	n	%	n	%	n	%
Symptomatic	Students	Kindergarten	8974	6872	76.6	2102	23.4	6499	94.6	373	5.4
		Elementary school	79,435	56,246	70.8	23,189	29.2	53,820	95.7	2426	4.3
		Middle school	36,551	25,119	68.7	11,432	31.3	23,770	94.6	1349	5.4
		High school	28,679	20,083	70.0	8596	30.0	18,228	90.8	1855	9.2
	Teachers/staffs		13,060	10,943	83.8	2117	16.2	9088	83.1	1855	17.0
Asymptomatic	Students	Kindergarten	172,746	1133	0.7	171,613	99.3	969	85.5	164	14.5
		Elementary school	1,087,744	7534	0.7	1,080,210	99.3	6659	88.4	875	11.6
		Middle school	614,195	4568	0.7	609,627	99.3	3773	82.6	795	17.4
		High school	474,160	3392	0.7	470,768	99.3	2519	74.3	873	25.7
	Teachers/staffs		307,000	1012	0.3	307,000	99.7	717	70.9	295	29.2
Total			2,822,544	136,902	4.9	2,686,654	95.1	126,042	92.1	10,860	7.9

^* Symptom refers to fever, chills, cough, headache, myalgia, sore throat, runny nose, loss of taste/smell, fatigue, nausea, vomiting, and diarrhea.

^†All positive results from RAT screening are reflexed to undergo PCR testing for confirmation.

FIGURE 1.: The positive predictive value of rapid antigen test screening among students and teachers/staff by presence or absence of symptoms (fever, chill, cough, headache, myalgia, sore throat, runny nose, loss of taste/smell, fatigue, nausea, vomiting, and diarrhea).

DISCUSSION

The overall rate of true positive results among the RAT tests for COVID-19 was high, especially in students with symptoms, which is consistent with other studies.⁵ Establishing an early diagnosis of COVID-19 requires interpreting test results appropriately. Currently, RAT is widely available and its appropriate use may improve case detection and therefore could reduce transmission and absenteeism.⁶ As this study was conducted parallel with the policy of returning to school if the RAT result was negative, the results may inform the policymakers to minimize school closure due to the COVID-19 outbreak.

The probability of infection depends on the SARS-CoV-2 prevalence in the community, which was high in South Korea with the Omicron variant surge during the observation period, which increased the performance of RAT in schools.^7,8 During the peak of the COVID-19 community outbreak, a positive RAT result is highly likely to represent a true SARS-CoV-2 infection. Furthermore, a negative test result suggests a low possibility of COVID-19. Here, most students presenting with respiratory symptoms and fever with a positive RAT result would have COVID-19, and conversely, if the prevalence is low, positive RAT results are more likely to be false-positive. The study limitations include the convenience sample based on passive surveillance using volunteered symptoms and RAT positivity, therefore may lack clear generalizability. Moreover, reported symptoms were aggregated as symptomatic versus asymptomatic, therefore, we were unable to assess the most significant symptom suggestive of RAT or PCR positivity.

Interpreting SARS-CoV-2 RAT appropriately can provide useful information for decision-making in the educational setting. RAT screening during the community outbreak of SARS-CoV-2 provides a high positive predictive value; therefore, provides helpful information to policymakers to optimize the school health policy.

ACKNOWLEDGMENTS

We thank the Ministry of the Interior and Safety, Si/Do and Si/Gun/Gu, medical health centers staff, and medical facilities for their efforts in responding to the COVID-19 outbreak.