Study design
This study was a single-center, cross-sectional investigation that employed a prospective cohort approach to assess the diagnostic accuracy of an Ag-RDT when collected through self-administration and by healthcare professionals for detecting SARS-CoV-2 infection. The reference standard employed to determine SARS-CoV-2 infection was RT-PCR. The study protocol, case records form, and consent form were reviewed and approved by the Ethics Committee for Human Research, Chulabhorn Research Institute (reference number: 155/2564).
Participants
The inclusion criteria were individuals aged 18 years or older who provided informed consent to participate in the study. The study had two target populations: (1) participants with respiratory tract symptoms such as a runny nose, cough, sore throat, or breathlessness with or without fever for 14 days; (2) participants with no respiratory tract symptoms or fever but who had close contact with COVID-19 patients within 7 days. Participants without RT-PCR results were excluded from the study. The study was conducted at the respiratory infection clinic of Chulabhorn Hospital in Thailand from October 2021 to January 2022, during which the delta variant of SARS-CoV-2 was the most commonly circulating strain 6. All potentially eligible participants who came to respiratory infection clinic were invited to participate in the study.
Test methods and principles
Ag-RDT test
After enrollment, all participants in the study underwent nasal swabbing twice, one for the Ag-RDT and one for real-time RT-PCR testing. Two types of Ag-RDTs were used in the study: 5T, which came in a pack of 5 tests designed for self-testing by patients, and 25T, which came in a pack of 25 tests designed for professional use. The assignment of participants into the 5T or 25T groups was randomized based on simple randomization. For both tests of Ag-RDT, the specimen collection process was similar. However, for the 5T-test, participants collected the specimen themselves under supervision of the trained nurse, while for the 25T-test, a trained nurse collected the specimen. The collection process involved removing the swab from the packaging, tilting the patient's head slightly back, inserting the sterile swab into the nostril with the most secretion until resistance was met at the turbinate, rotating it for around 15 seconds, and then removing it. This same process was repeated for the other nostril with the same swab. After specimen collection was complete, the swab was placed into an extraction buffer tube, and the buffer tube was squeezed while stirring the swab more than 10 times. The swab was then removed while squeezing the side of the tube to extract the liquid from the swab. Four drops of the extracted sample were then applied to the test device, and the result was read in 15–30 minutes.
The SARS-CoV-2 Ag-RDT is a qualitative test that uses an immunoassay technique to detect the presence of the SARS-CoV-2 antigen in a patient's specimen. This test kit employs color particles to aid in the detection process. Initially, there are two lines on the test paper made from a nitrocellulose membrane. The first line is the "T" test line, which contains a mouse monoclonal anti-SARS-CoV-2 antibody (mAb anti-COVID antibody). The second line is the "C" control line, which contains a mouse monoclonal anti-Chicken IgY antibody (mAb anti-Chicken IgY). When a specimen containing the SARS-CoV-2 antigen is added to the test kit's reagent, it forms a complex with the color particle (mAb anti-COVID-19 antibody-gold conjugate). This complex then spreads across the test paper and is captured by the mAb anti-COVID antibody at the T line. Within the same reagent, there is a control consisting of purified chicken IgY-gold conjugate, which also spreads across the test paper and is captured by the mAb anti-Chicken IgY at the C line.
Therefore, the presence of two visible lines, one at the T line and one at the C line, indicates a positive test result, even if the test line is faint or not uniform. A single line at the C line indicates a negative result, while the absence of a line at the C line suggests an invalid or falsified test.
SARS-CoV-2 RT-PCR test
After completing the Ag-RDT, participants underwent another specimen collection process for RT-PCR testing, using a more flexible swab with a polyester tip. The swab was inserted deeper into the nasopharynx, rotated against the mucosa for approximately three seconds, and then withdrawn. The swab was placed immediately in cobas® PCR media tube, and the remaining portion of the swab was broken off and left in the tube. The cobas® PCR media tube was then transported to the central laboratory center of the hospital in a box with ice, where the cobas® SARS-CoV-2 RT-PCR test was performed. Prior to the RT-PCR test, the cobas® PCR media tube and specimen was stored at 2-8°C, and the test was conducted within 48 hours. The test tubes are already labeled with the patients' names and are then arranged into the test cartridge. From this point, the RT-PCR process is fully automatic. The Cobas machine carries out the necessary steps for the test, and at the end of the process, the results, whether positive, negative, or invalid, are displayed and printed out from the machine.
The SARS-CoV-2 RT-PCR test is a quantitative test that detects specific viral RNA sequences. The process involves collecting a sample from the patient and adding it to a transport medium. An enzyme called proteinase is then used to extract the viral RNA and separate it from other debris that could affect the test results. At the same time, internal control RNA (RNA IC) is included in the process.
Next, reverse transcription is performed to convert the viral RNA into double-stranded viral DNA, which is ready for polymerization. Specific forward and reverse primers, designed to target the ORF1 a/b non-structural and E gene regions of the virus, are added to initiate the DNA polymerase reaction. Additionally, a separate set of forward and reverse primers with a completely different sequence is used specifically for the internal control RNA sequence.
Afterward, a mixture containing detection probes with specific sequences for SARS-CoV-2 and RNA IC is added. Each probe contains a reporter dye and a quencher dye. During DNA hybridization, if the probes detect the specific sequences, the quencher dye is cleaved, and the reporter dye emits a fluorescent signal. The fluorescence signals are measured at different wavelengths for SARS-CoV-2 and RNA IC simultaneously during each amplification cycle. A positive test result indicates the presence of a fluorescence signal for SARS-CoV-2, while a valid test result indicates the presence of a fluorescence signal for the internal control.
The Ag-RDT and RT-PCR tests were performed at different sites, and each of the result readers did not know each test result.
Analysis
Summary statistics were presented as medians with interquartile ranges (IQRs) or as means with standard deviations (S.D.) where applicable. To compare continuous data, independent t-tests or the Mann-Whitney Test were employed as appropriate. For categorical data, comparisons were made using the chi-squared test or exact test as needed.
To compare the diagnostic performance accuracy of the Ag-RDT test between the two types of Ag-RDT tests (5T vs 25T) using the receiver operating characteristic curve (AuROC) as a primary outcome. In addition, the other parameters also provide to demonstrate the diagnostic performance accuracy of each Ag-RDT test, including the area under the AuROC, sensitivity, specificity, positive predictive value (PPV), negative predictive value (NPV), and likelihood ratio of a positive test (LR+). Additionally, we provided information on the prevalence of the disease in the study population. Any missing data were excluded from the analysis to ensure the accuracy of the results. We also conducted a comparison between participants with symptoms and those with high risk but without symptoms using the AuROC.
The sample size for this study was carefully determined to meet the requirements set by the Thai Food and Drug Administration. The goal was to have a minimum of 50 participants to evaluate the sensitivity of the rapid antigen test kit with a sensitivity of at least 90%, and 100 participants to evaluate the specificity with a specificity of at least 98%. To estimate the required sample size, we assumed a prevalence of SARS-CoV-2 infection during the pandemic to be 50%, and the sensitivity of RT-PCR was considered to be 99%. Based on these assumptions, we calculated that a total of 112 participants would provide the trial with 80% power and an alpha error of 5%. As the study involved evaluating two types of antigen test kits (5T and 25T), we aimed to include at least 224 participants with symptoms to assess the sensitivity and 224 participants without symptoms but with confirmed contact with SARS-CoV-2 to evaluate the specificity.