Differential clinical characteristics and performance of home antigen tests between parents and children after household transmission of SARS-CoV-2 during the Omicron variant pandemic

Objectives The SARS-CoV-2 Omicron variant pandemic struck Taiwan in April 2022. Rapid antigen tests (RATs) play an important role in providing rapid results during a pandemic. However, self-collected samples by the children's caregivers without the supervision of medical personnel raise some concerns. Methods This study was performed to investigate household transmission, clinical characteristics, and antigen performance in a special COVID-19 family clinic in a children's hospital. The performance of at-home RATs was evaluated based on reverse transcription-polymerase chain reaction. Results We included 627 patients in our study between May 11 and June 10, 2022. The COVID-19 full vaccination rate was significantly higher in adults (98.5%) than in children (5.9%, P <0.001). The transmission rate was significantly higher in children (91.3%) than in adults (76.6%, P <0.001). Infected children had more incidents of fever (82.4% vs 22.4%, P <0.001) and a higher peak fever than adults. Based on the reverse transcription-polymerase chain reaction, the negative predictive rate of the home RAT was only 38.7% (95% confidence interval: 31.9-46.0%) in children. The cycle threshold value of those with false-negative antigen tests was significantly lower in children. Conclusion Children had a higher transmission rate, more fever, and higher peak fever than adults. Home RAT has a suboptimal negative predictive rate in children.


Introduction
The SARS-CoV-2 Omicron variant (B.1.1.529) BA.2 pandemic struck in 2022. Children seemed to have a lower rate of infection than adults at the beginning of the COVID-19 pandemic, but the clinical scenario changed, especially after the Omicron variant outbreak [1] . The transmission of SARS-CoV-2 may vary according to different viral variants, settings, and individuals, and understanding the transmission rate and factors associated with transmission may help further control COVID-19. To this end, the difference in clinical symptoms and overall household transmission rate between children and adults needs further investigation. Nasopharyngeal reverse transcription-polymerase chain reaction (RT-PCR) has been the gold standard for the diagnosis of COVID-19, but the need for special equipment, personnel qualifications, and labor time and the high cost represent considerable medical burdens during the pandemic [2] . Rapid antigen tests (RATs) may be an easy way to provide rapid results and reduce the laboratory burden. However, the sensitivity of RATs varies, ranging from 87.1-100%, and is related to the timing of specimen collection, prevalence, and SARS-CoV-2 viral load [3][4][5][6] . To the best of our knowledge, pediatric data are limited, and most real-world performance assays focus on the difference between RATs and RT-PCR tests under the circumstances of standardized sample collection by medical personnel [ 5 , 7 ]. Self-collected samples and those collected by children's caregivers without the supervision of medical personnel might raise some concerns, such as inappropriate sampling processes or contamination of samples. Failure to collect samples of the nasopharynx might be more common in pediatric sampling performed by caregivers owing to the lack of anatomical knowledge or fear felt by the children. Although it has been reported that RAT samples from the midturbinate and nasopharynx can yield similar results in adults under certain circumstances, the evidence in children is limited. Thus, our aim was to investigate the transmission, clinical characteristics, and home RAT performance of the Omicron variant among household children and their caregivers [ 8 , 9 ].

The status of COVID-19 and definition of confirmed cases in Taiwan
Taiwan, with a population of approximately 23.9 million, was struck by the SARS-CoV-2 Omicron variant (B.1.1.529) beginning in April 2022 and peaking in May, with 94,855 daily new cases (peak daily incidence rate of four per 10 0 0 individuals) on May 27, 2022. In Taiwan, the total number of children younger than 20 years with COVID-19 has exceeded 80 0,0 0 0 in 2022 [10] .
The Central Epidemic Command Center of Taiwan was established in January 2020 and is responsible for the majority of national policies related to COVID-19, including case definitions. Initially, a COVID-19 RT-PCR test was performed on all patients to confirm the diagnosis of COVID-19. With growing evidence of RAT performance and considering the cost-effectiveness of these tests, the Central Epidemic Command Center of Taiwan revised the case definition as either a positive RT-PCR or RAT on May 26, 2022. A positive antigen test was defined as when an individual, regardless of age, tested positive for COVID-19 with an at-home RAT kit and whose test result was confirmed by a medical professional or who tested positive with a RAT performed by a medical professional.

Study population and study period
A COVID-19 family clinic was set up on May 11, 2022, at the National Taiwan University Children's Hospital, specifically for children and their household caregivers who had confirmed (through RT-PCR or RAT) or suspected COVID-19 infection (based on symptoms or contact history). We retrospectively reviewed the charts of patients who visited this clinic between May 11 and June 10, 2022. There were 635 visits within this 1-month period. We excluded six patients who were immediately referred to the emergency department and the second visit of two patients. In total, 627 patients were included in our study ( Figure 1 ). This study was approved by the institutional review board of the National Taiwan University Hospital (202206053RINC).

Diagnostic procedures
RATs were performed by patients themselves or by their caregivers at home. The patients could use only an authorized specimen collection kit purchased over the counter in a pharmacy or at authorized stores. A standard written instruction in Chinese for the RAT were included in each commercial authorized kit. A name-based rationing system for at-home COVID-19 test kits was launched on April 28, 2022, which allowed individuals to purchase test kits with their national health insurance card at a regulated price in Taiwan. For RT-PCR, nasopharyngeal swabs were taken by pediatricians in the clinic after the agreement of the patient or the parent accompanying the child. RT-PCR was performed by a Roche cobas® SARS-CoV-2 assay (Roche Molecular Systems, Branchburg, NJ, USA), with three unique TaqMan probes targeting conserved regions within the ORF 1a/b and E genes [11] . Cycle threshold (Ct) values were provided by the testing platform based on manufacturerprovided interpretation criteria.

Clinical data collection
We collected clinical symptoms, COVID-19 vaccination history, and history of contact with a patient who is COVID-19-positive or any symptomatic individual from all patients and guardians of the children. Demographic (age and sex) and clinical variables and Ct values of RT-PCR results of the study population were obtained from electronic medical records. The vaccination status of our patients was obtained from the National Immunization Information System.

Vaccination
During the study period, four COVID vaccines were available in Taiwan, including those from Moderna and Pfizer-BioNTech (messenger RNA [mRNA] vaccines) and Oxford-AstraZeneca (viral DNA vector), and the Medigen Biologics MVC COVID-19 vaccine (protein subunit). The national vaccination for adults commenced in March 2021 and is free for all Taiwanese individuals and long-term residents. The first booster dose was available for all adults beginning in December 2021. mRNA vaccines were approved for children aged 12-18 years in September 2021, and the second dose was approved on November 28, 2021. Children aged 6-11 years could receive the mRNA COVID-19 vaccine after April 2022. COVID-19 vaccination for children younger than 5 years was still not available during the study period in Taiwan, considering insufficient clinical evidence.
Patients who were "fully vaccinated" were those who received at least two doses of COVID-19 vaccination and at least 2 weeks had passed. Patients with "partial vaccination" were those who received at least one dose of COVID-19 vaccination and at least 1 week had passed but who did not meet the criteria for full vaccination. A status of "no vaccination" was assigned to those who did not receive any COVID-19 vaccination or those who received one dose of vaccination within 1 week before the clinic visit.

Statistical analysis
Characteristics are presented as the number (%) for categorical variables and as the median with interquartile range (IQR) for continuous variables. The infection rate was defined as the number of newly confirmed COVID-19 cases, divided by the total number of patients who had symptoms and visited the clinic. Sensitivity and specificity, positive and negative predictive values, and accuracy were calculated to evaluate the performance of the home RAT based on the SARS-CoV-2 RT-PCR test. Sensitivity was defined as the probability that the RAT result was positive when RT-PCR was positive, and specificity was defined as the probability that the RAT result was negative when RT-PCR was negative. The positive predictive value was the probability that RT-PCR was positive when the RAT result was positive, and the negative predictive value was the probability that RT-PCR was negative when the RAT result was negative.
We compared the clinical characteristics and overall infection rate between the children and the adults within families. Differences between groups were compared using the Mann-Whitney U test or chi-square test. A two-sided P -value of 0.05 or below was considered statistically significant. Statistical analysis was performed using SPSS v28.0 (IBM Corp., Armonk, NY, USA) and the SciPy package (version 1.7.1) of the Python programming language version 3.9.7 (Python Software Foundation, Fredericksburg, VA, USA).

Patient characteristics
A total of 627 patients were included in this study, comprising 426 (68%) children and 201 (32%) adults ( Table 1 ). The median age of the children was 5 years (IQR: 2-8 years), and the median age of the adults was 40 years (IQR: 36-44 years). Among all patients, 35 (5.6%) had underlying diseases, and 484 (77.2%) had a known contact history with patients who were COVID-19positive or suspected patients. The COVID-19 full vaccination rate was significantly higher in adults (98.5%) than in children (5.9%, P < 0.001). Children had more school or occupational contact than their household adults ( P < 0.001). The infection rate of COVID-19 in the cohort was 86.6% (91.3% in pediatric subjects and 76.6% in adults). The household children had a significantly higher infection rate than their household adults ( P < 0.001).
We further compared the clinical symptoms of children infected with COVID-19 according to their vaccination status ( Table 3 ). The average age was significantly different (fully vaccinated: 15.5 years [13.0-17.0]; partially vaccinated: 8.0 years [6.5-10.0]; not vaccinated: 4.0 years [2.0-7.0], P < 0.001), which was in line with the national vaccination policy. Those without vaccination or who received only a partial vaccination had a higher rate of fever than those who had received a full vaccination (50% of the fully vaccinated group; 74.2% of the partially vaccinated group; 89.5% of the nonvaccinated group, P < 0.001).

Performance of RATs based on RT-PCR
Overall, 397 patients received both rapid antigen testing and RT-PCR. The numbers and Ct values of patients with different test results are shown in Table 4 . Interestingly, the Ct value of those who had false-negative antigen results was significantly lower in the pediatric group than in the adult group (21 [19,24] vs 29 [19,33], P = 0.013). In children, the Ct value of true-positive cases was similar to that of false-negative cases ( P = 0.28). However, the Ct value of true-positive cases was significantly lower than that of false-negative cases in adults ( P < 0.001).
Based on RT-PCR, the overall sensitivity and specificity of home RAT were 79.8% (95% CI: 75-84%) and 90.9% (95% CI: 81.3-96.6%), respectively ( Table 5 ). The sensitivity and specificity were significantly higher in pediatric subjects (83%, 96%) than in adults (72.9%, 87.8%). Among PCR-positive cases with RAT and the available documented onset day of fever or clinical illness, we analyzed the sensitivity according to the day after the infection. The sensitivity varied (68-93% according to the day after the fever and 73-91% according to the day after the clinical illness, respectively) with the day after the infection, and it was the highest on the third day after the fever or clinical illness ( Supplementary Tables 1 and 2). The positive predictive value was also significantly higher in pediatric subjects (99.5%, 95% confidence interval [CI]: 96.5-99.9%) than in adults (94%, 95% CI: 87.2-97.3%) ( P < 0.001), but the negative predictive value was only 38.7% (95% CI: 31.9-46.0%) in children and 55.4% (95% CI: 47.1-63.4%) in adults ( P = 0.001). The overall accuracy of the RAT was 81.6% (95% CI: 77.4-85.3%), without a statistically significant difference between the adult and pediatric groups.

Discussion
Younger children who seldom receive the COVID-19 vaccine and have difficulties in maintaining social distancing and wearing masks may play an important role in the spread of COVID-19. Without strict social distancing measures ( e.g ., school closures) and COVID-19 vaccination, children could be index cases at home, resulting in household transmission. Although the SARS-CoV-2 vaccination coverage rate of the first dose in the entire population      was over 92% in Taiwan, the Advisory Committee on Immunization Practices at the Ministry of Health and Welfare of Taiwan did not suggest that children aged younger than 12 years receive the COVID-19 vaccination before April 2022 due to the lack of an available pediatric COVID-19 vaccine in Taiwan, concern over vaccine adverse effects, and limited clinical evidence at that time [10] . Children aged 6-11 years could receive the COVID-19 vaccine after May 9, 2022. The lack of vaccination makes children more vulnerable to the Omicron variant than adults who have had access to vaccines for months; so, this study shows that the infection rate and severity were significantly higher in children than in adults. The overall COVID-19-confirmed rate was significantly lower in the adult group, and their rates of fever and peak body temperature were also significantly lower than those in the pediatric group. The difference in vaccination rates could explain the different symptomatic ratios across different age groups [12] . Higher rates of fever were also noted in the subgroup analysis of children infected with COVID-19 between vaccination statuses ( Table 3 ). We only compared the reported fever peak but not the average body temperature during the COVID-19 infection, which may explain why the peak body temperature difference did not reach statistical significance. The most commonly reported symptoms, such as fever and cough, were similar to those previously reported [13][14][15][16] . Sore throat was reported to be more prevalent among Omicron cases than among Delta cases, regardless of vaccination status, and our cohort indeed had a higher rate of sore throat than that in a previous report [ 14 , 17 ]. However, the evaluation timing after disease onset and expression ability in different age groups could interfere with the reported rate of symptoms.
Laboratory testing is necessary to confirm the diagnosis of COVID-19 because no single symptom or combination of symptoms reliably differentiates SARS-CoV-2 from other community-acquired respiratory viruses [18] . As anticipated, patients with a close contact with a patient with COVID-19 were more likely to test positive with both RT-PCR and RAT. The overall specificity in this study was high, at 90.9% (96% in children vs 87.8% in adults), which was consistent with previous studies describing the excellent specificities of different RATs, ranging from 87.1% to 100% [ 3-6 , 19 ]. The overall sensitivity was 79.8% (83% in children vs 72.9% in adults), which was optimal in a cohort of symptomatic patients [20] . It was suggested that serial testing is critical in clinically suspected patients to compensate for the lower sensitivity of the RAT because asymptomatic patients with high Ct values at an early stage of infection may not be detected with Ag tests until later in the course [ 5 , 21 ].
The negative predictive value of the home RAT was quite low (47.2%) in our study, especially in children (only 38.7%). Notably, among cases with a false-negative RAT, children had significantly lower Ct values than adults. The Ct value of false-negative cases was similar to that of true-positive cases in children. This finding indicates that it is difficult for some parents to obtain a good nasopharyngeal swab sample from children, resulting in false-negative test results. Adherence to the standard instructions of the home RAT could also vary and may affect the RAT performance. Therefore, proper evaluation/examination is indicated through physicians for children with symptoms or high clinical suspicion of COVID-19 and negative home RAT results.
The strengths of this study are the comparison of children and adults among household members with COVID-19 contact and the real-world performance of the RAT carried out at home. Notably, all the nasopharyngeal swabs for RT-PCR in our study were sampled by licensed pediatricians who were well trained and familiar with pediatric anatomy. Taiwan's National Health Insurance covers most of the medical costs, including RT-PCR tests, when clinically indicated. This policy supports the ability of all patients, regardless of economic status, to receive standardized medical treatment. Compared with previous studies that focused on comparing the performance of self-test RAT with RT-PCR, all of our patients were evaluated based on clinical needs instead of study volunteers. In addition, we had a younger age on average in the pediatric group and had nasopharyngeal specimens instead of specimens of the anterior nares [ 19 , 21 , 22 ].
Nevertheless, the results of the study should be considered in light of its limitations. First, we only analyzed symptoms reported during the clinical evaluation. In addition, symptoms present at the time of infection might have been related to viral or bacterial coinfections at the time of SARS-CoV-2 infection. The definite time of infection or symptom onset was difficult to clarify in all of our cases, and most of the adults arrived at the clinic with children who already had obvious symptoms. The discordance might be related to different times since the onset of symptoms. However, if we used Ct values as a surrogate for viral load, there was no statistically significant difference between the adult and pediatric groups. Second, the COVID-19-confirmed case definition was based on the Central Epidemic Command Center's rolling wave planning; so, RT-PCR was not performed in all of our patients, which could also impact the overall performance of the RAT. Incorrect compliance with the instructions of the home test can be interpreted as a possible limitation but is also an important confounding factor that physicians need to consider in the real-world setting. Third, the use of the Ct value needs to be interpreted with caution even though there was no statistically significant difference between the adult and pediatric groups. Although our hospital had its own testing platforms and testing sites under strict quality control performed by highly trained laboratory technologists under standard operating procedures, the variation among different users, specimen collection, and different infection processes need to be considered. Finally, the infection rate was high (approximately 80%) in our cohort due to the nature of the COVID-19 family clinic, which could impact the overall diagnostic performance of the RAT.
In conclusion, household children with a lower vaccination rate had a higher SARS-CoV-2 infection rate, higher incidence of fever, higher peak fever, and more gastrointestinal symptoms than their adult caregivers. The home RATs had a suboptimal negative predictive rate in children in our cohort. Proper evaluation/examination is indicated through physicians for children with a contact history and symptoms of COVID-19 but a negative Ag test.

Declaration of competing interest
The authors have no competing interests to declare.

Funding
This study was supported by the Ministry of Science and Technology (MOST 111-2314-B-002-146 and NSTC 111-2321-B-002-017) and National Taiwan University Hospital (NTUH. W1_112-03). The funders of this study had no role in the study design, data collection, analysis, interpretation, or writing of the report.

Ethical approval
This study was approved by the National Taiwan University Hospital Institutional Review Board (202206053RINC). The need for written informed consent was waived by the regulation of the National Taiwan University Hospital Institutional Review Board. All methods were carried out in accordance with the guidelines of the Declaration of Helsinki.

Author contributions
S-H Chen performed the literature search. L-Y Chang and S-H Chen designed the study. S-H Chen, J-L Wu, T-Y Yen, and C-Y Lu collected the data. S-H Chen, Y-C Liu, W-T Lee, J-M Chen, P-I Lee, L-M Huang, and L-Y Chang performed the data analysis. The data were interpreted by all authors. S-H Chen wrote the original draft of the manuscript, which was then reviewed and edited by all authors. L-Y Chang and Y-C Liu assessed and verified the data. All authors had full access to all of the data. L-Y Chang had final responsibility for the decision to submit the manuscript for publication.

Data availability statement
All data relevant to this study are included in the article. Please contact the corresponding author, Dr. Luan-Yin Chang, to request the study data.

Supplementary materials
Supplementary material associated with this article can be found, in the online version, at doi: 10.1016/j.ijid.2023.01.014 .