Researchers observe that a combination of oral and nasal swabs would yield better results for detecting SARS-CoV-2
In a recent study posted to medRxiv*, researchers observed that a combination of oral and nasal swabs would yield better results for detecting severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2).
Early detection of SARS-CoV-2 infection is critical to reduce its transmission and minimize the spread of new variants. The utility of rapid antigen tests (RATs) has been increasing for SARS-CoV-2 screening. RATs provide quick results, are portable, and are less expensive, improving accessibility in low-resource settings.
Rapid antigen/molecular tests require high viral loads and have lower analytic sensitivity than reverse-transcription quantitative polymerase chain reaction (RT-qPCR). Previously, the authors observed a delayed increase in the viral loads of anterior-nares swab (ANS) samples compared to oral cavity samples and that viral load differed with specimen type for the same individuals at the same sampling time.
About the study
In the present study, researchers tested whether the previously observed differences in viral loads from different specimen types and delayed increase in ANS viral loads would lead to poor performance of a specific RAT. About 228 individuals were enrolled, and 90 were SARS-CoV-2-infected, ascertained by RT-qPCR at enrolment. Participants self-collected saliva, ANS, and oropharyngeal swab (OPS) samples for RT-PCR testing.
RT-qPCR was performed using a quick RT-qPCR testing kit, with a reported limit of detection (LOD) of 250 copies/ml. The team also performed viral sequencing and variant determination. Lab operators and supervisors were blinded to participants’ samples, infection status, and RAT results. Participants took an at-home RAT, the Quidel QuickVue At-Home OTC COVID-19 Test, using ANS immediately after packaging specimens for RT-qPCR.
Test results were interpreted by the participants, who also shared photographic evidence of test strips. Participants collected test specimens at 2215 time points; they were considered infected if any specimen was positive on RT-qPCR tests, non-infected if all specimens were RT-qPCR-negative, and inconclusive if one specimen was inconclusive while others were negative by RT-qPCR. Overall, results from 2118 timepoints had paired ANS RAT and RT-qPCR results, of which 847 were deemed infected.
The team observed a positive percent agreement of 47% for RAT-positive results relative to PCR-positive results, much lower than reported by the manufacturer (83.5%). Although PCR-positive and RAT-negative results were anticipated for time points with low viral loads in the ANS, more than half of the PCR-positive ANS specimens were RAT-negative. They observed a negative percent agreement of 97% for RAT-negative results relative to PCR-negative results, marginally lower than the manufacturer’s reports (99.2%).
The antigen test missed more than half of the time points with infectious viral loads (> 104 copies per ml). Of the 90 infected individuals, 79% had at least one RAT-positive result during the study. The team identified 17 individuals who enrolled and began sample collection, with a quantifiable viral load in all specimen types. Each participant reported a minimum of one COVID-19-associated symptom within three days of the first detectable viral load.
The sensitivity of the RAT for symptomatic individuals was significantly higher than that for asymptomatic individuals. Almost all individuals attained ‘presumed infectious viral loads’ a day before the ANS sample was RAT-positive. The delay between the first infectious sample and RAT positivity was one/two days (for six subjects), three days (five individuals), and five/eight days for one participant. Notably, two subjects presumably had infectious viral loads for several consecutive days but never tested RAT-positive.
In subsequent analyses, the authors observed that ANS RATs were poor at detecting pre-infectious and infectious subjects. They assessed the clinical sensitivity of different sample types tested with low/high-analytic-sensitivity assays to detect individuals during the infectious phase. None of the single specimen type, i.e., ANS, OPS, or saliva, reached 95% clinical sensitivity with low/high-analytic-sensitivity assays.
The researchers observed that participants frequently achieved infectious viral loads in OPS or ANS samples first; this led to speculation that a specimen combining OP and AN sampling on one swab might augment test performance. A computationally-contrived combination of AN and OP swab samples was created. The combination specimen was predicted to perform better than any single specimen type. The OP-AN specimen combination also featured improved performance than the other possible two-specimen combinations.
The study demonstrated lower clinical sensitivity (44%) of the Quidel QuickVue RAT to detect infected individuals at any infectious stage. The observed sensitivity to detect even at symptomatic time points was still low (< 50%). Individuals exhibited infectious viral loads in at least 33% of asymptomatic time points, and various participants were asymptomatic on the day of peak viral load. The team concluded that an AN-OP combination swab would be significantly more effective in detecting all stages of infection than a single swab sample, even during the earliest infectious phase.
medRxiv publishes preliminary scientific reports that are not peer-reviewed and, therefore, should not be regarded as conclusive, guide clinical practice/health-related behavior, or treated as established information.