A Comparison of COVID-19 Point-of-Care Tests
How available molecular, antigen, and serology tests that can be performed at the point of care stack up
The emergence of the severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) disease in late 2019 created an urgency to map its genome in an effort to diagnose and control the global threat. That relentless pace of scientific discovery spurred the development of laboratory based tests to diagnose infection, and led to numerous commercially available point of care tests (POCT) that provide rapid and mobile test results at the site of patient care.
The World Health Organization (WHO) designated COVID-19 a global pandemic in March 2020, ensuring that all POCT received not only an emergency use authorization (EUA) but also a Clinical Laboratory Improvement Amendment (CLIA) waived certification. As a result, COVID-19 POCT fall short of the usual Food and Drug Administration mandates for minimum rigor in testing accuracy, forgoing the stringent conformation to laboratory protocols required for high and moderate complexity CLIA certifications. Emergency CLIA waiver certification is granted to sites as soon as the application is submitted and can be used at additional off-site locations.
What are COVID-19 POCT options?
Molecular testing
The test of choice for acute diagnosis of COVID-19 infection is the nucleic acid amplification test (NAAT). Specifically, the gold standard is RT-PCR from an upper respiratory specimen.
Additionally, loop mediated isothermal amplification (LAMP) and palindromic repeat (CRISPR based) techniques are used for rapid diagnosis. These POCT tests require minimal human power from laboratory personnel, are rapid (most are completed within 30 minutes) and provide excellent sensitivity and specificity. However, these mobile tests are limited by their cost, as each test runs many fewer samples than the facility-based laboratory counterparts. Additionally, they are not standardized to provide the relative viral concentration in a sample and may not differentiate replicating virus from remnant viral RNA/DNA. As viral load and replication may affect disease severity and infectivity, these limitations should be considered when directing patient management.
Antigen testing
Antigen testing, the majority of which is performed at the point of care, detects viral proteins—most commonly the nucleocapsid protein—from respiratory samples and employs immune based detection techniques such as microfluid immunofluorescence assay, lateral flow, and chromatographic digital immunoassay. The myriad of detection techniques leads to lower sensitivity, and a wider sensitivity range, compared to molecular tests, with an average sensitivity of 56.2 percent (0–94 percent) for antigen POCTs. Antigen based tests are further affected by the stage of infection and disease prevalence in the local community. They provide results within 30 minutes, making them especially useful when a rapid diagnosis is required for purposes of isolation or treatment or when molecular tests are not available. The specificity of antigen POCT is high (98.9 percent; CI 97.3-99.5 percent) but there is cross reactivity with other human coronaviruses, such as Middle East Respiratory Syndrome (MERS) and other SARS viruses. Due to poor sensitivity, negative tests require confirmation using a molecular based test when clinical suspicion of infection remains high.
Serological testing
Serologic tests detect antibodies against nucleocapsid or spike proteins and can diagnose current or past COVID-19 infections. They lack utility in detecting active infections as antibody levels are low at the height of viral replication or infectivity. As such, serologic testing is most useful at three to four weeks after the onset of symptoms, when a robust antibody response has occurred.
For POCT serological testing, the available technique is lateral flow immunoassays (LFIA). Meta-analysis suggests that mean sensitivity for LFIA based tests is lower (66 percent) compared to ELISA (84.3 percent) or chemiluminescent immunoassay methods (97.8 percent). The significance of specific immunoglobulin testing is unclear, with some studies indicating that immunoglobin G (IgG) or total immunoglobins tests are the most accurate, while others suggesting no difference.
Like other available diagnostic tests, cross reactivity occurs with other human coronaviruses. For this reason, a positive antibody test typically requires confirmation with a subsequent antibody test. Regardless, there is considerable concern about the accuracy of serological testing in managing individual patients and it remains unclear if antibody detection confers immunity since the immunological memory response is not directly assessed.
Etiologies of inaccuracies
Population prevalence
All diagnostic tests are affected by the disease prevalence. Diagnostic accuracy is improved in high prevalence locations such as long-term facilities or sites of new outbreaks. In low population settings, the positive predictive value of tests is diminished.
SARS-CoV-2 variants
With the detection of new COVID-19 variants, such as the B.1.1.7 variant, there is concern that genome mutation may lead to specific mutations of the sequences targeted and amplified by molecular techniques for detection. So far, the sensitivity of several tests has been impacted, and potential for false positives and negatives has been noted, but the information is still rapidly evolving.
Specimen collection
Sampling for both molecular and antigen tests is done via the respiratory tract, with most specimens obtained from the posterior nasopharynx. As many tests are self-administered, there is concern regarding individual ability to obtain adequate specimens. With evidence that the sensitivity of salivary specimens is comparable to nasopharyngeal specimens, there is increasing use of combined nasopharyngeal and oropharyngeal samples.
Timing of testing
The viral load affects the sensitivity for all POCTs. Virus is best detected around 5–7 days after exposure and in symptomatic patients. Individuals are well served by POCT during this time, but viral loads may be undetectable earlier after exposure and may decline in the week following symptom onset. For example, one study showed that the false negative rate for RT-PCR testing is 100 percent on the day of exposure, as low as 20 percent on the eighth day of exposure—when symptoms are present—then increases to 66 percent at day 21.
During the incubation period and especially in asymptomatic patients, false negatives abound. For this reason, EUA for COVID-19 POCT was originally intended for symptomatic individuals, although the Center for Medicare and Medicaid Services (CMS) has allowed use of rapid antigen tests on asymptomatic individuals. Antibodies, on the other hand, take weeks to develop resulting in negative results soon after symptom onset. Additionally, antibody lifetime is unknown with some studies showing persistence for up to four months while others indicate antibodies resolve around eight weeks.
Table 1. Accuracy of POCT COVID-19 molecular tests
Molecular Tests | ||||
---|---|---|---|---|
Type | LoD copies/ml | Cross reactants | CLIA* setting | |
Abbot ID Now | RT, isothermal amplification | 125 | none | H, M, W |
Cue COVID-19 | RT, isothermal amplification | 4000a | SARS-CoV-1 | H, M, W |
Cepheid Xpert Omni SARS-CoV-2 | Real-time RT-PCR | 400 | Human, bat coronavirus | H, M, W |
CLIA: Clinical Laboratory Improvement Amendment Certification
*H= high complexity lab, M= moderate complexity lab, W= waived (point of care testing)
RT: reverse transcriptase; RT-PCR: reverse transcriptase polymerase chain reaction
LoD: level of detection: ≥ 95 percent with 95 percent of confidence. a no 95 percent CI given
Table 2. Accuracy of POCT COVID-19 antigen tests
Antigen Tests | |||||
---|---|---|---|---|---|
Type | Sn (percent) | Sp (percent) | TCID50 (x 102/ml) | CLIA* setting | |
Quidel Sofia SARS | Lateral flow, fluorescence | 96.7 (83.3-99.4) | 100 (97.9-100) | 1.1 | H, M, W |
BD Veritor | Chromatographic digital immunoassay | 84 (67-93) | 100 (98-100) | 1.4 | H, M, W |
Abbott BinaxNOW | Lateral flow, visual read | 84.6 (76.8-90.6)a | 98.5 (96.6-99.5)a | 1.4 | H, M, W |
CLIA: Clinical Laboratory Improvement Amendment Certification
*H= high complexity lab, M= moderate complexity lab, W= waived (point-of-care testing)
Sn: Sensitivity; Sp: Specificity; (compared to RT-PCR)
TCID50: median tissue culture infectious dose
afor patients presenting within seven days of symptom onset
Table 3. Accuracy of POCT COVID-19 serology test
Serology Tests | ||||
---|---|---|---|---|
Type | Sn (percent) ≥15 days | Sp (percent) ≥15 days | CLIA* setting | |
Advaite RapCov | IgG lateral flow | 90 (73.6–97.3) | 95.2 (89.2-97.9) | H, M, W |
Hangzhou RightSign | IgM & IgG lateral flow | IgM: 92.5 (89-94.9) IgG: 91.6 (88-94.1) | IgM: 98.1 (95.1-99.2) IgG: 99.5 (97.4-99.9) | H, M, W |
CLIA: Clinical Laboratory Improvement Amendment Certification
*H= high complexity lab, M= moderate complexity lab, W= waived (point of care testing)
Sn: Sensitivity; Sp: Specificity (compared to RT-PCR)
The role of COVID-19 POCT
POCT is critical during the global pandemic to provide far reaching evaluation at doctors’ offices, urgent care facilities, and pop-up testing sites. COVID-19 POCTs extend laboratory capabilities by providing public health access to rural areas, long-term care facilities, and new areas of emerging outbreaks. Still, POCT cannot replace laboratory-based tests that handle more samples and maintain higher standards for laboratory processing.