COVID-19 Ag/Ab Diagnostic
For the COVID-19 (SARS-CoV-2) response, rapid diagnostic tests (RDTs) of COVID-19 antigen (Ag) and antibodies (Ab) are expected to complement core molecular tests, particularly in low-resource settings. This review assesses the requirements for the implementation of the COVID-19 PDRs in Sub-Saharan Africa.
Review of articles published in PubMed evaluating the PDRs of COVID-19 supplemented with Instructions for Use (IFU) of the products.
In total, 47 articles were recovered on two PDRs of COVID-19 Ag and 54 PDRs of COVID-19 Ab and IFU of 20 PDRs of COVID-19 Ab. Only five COVID-19 Ab PDRs (9.3%) were evaluated with capillary blood sampling at the point of care; none of the studies was conducted in sub-Saharan Africa. Sampling: Challenges for COVID-19 Ag PDRs include nasopharyngeal sampling (technique, biosafety) and sample stability; For COVID-19 Ab RDT, the equivalence of whole blood vs. plasma/serum needs further validation (assessed for eight (14.8%) products only). Sensitivity: specificity: the sensitivity of COVID-19 Ag and Ab RDT depends on the viral load (antigen) and the period of time (antibody), respectively; COVID-19 Ab tests have lower sensitivity compared to laboratory tests platforms, and the kinetics of IgM and IgG are very similar.
The specificity reported was high but has not yet been evaluated against tropical pathogens. Kit setup: for COVID-19 Ag RDTs, flocked swabs should be added to the kit; For COVID-19 Ab RDT, fingerstick sampling materials, transfer devices, and controls must be added (currently only supplied in 15, 5, and 1/20 products). Usability and robustness: some COVID-19 Ab PDRs showed high proportions of weak lines (> 40%) or invalid results (> 20%). Deficiencies were reported for the buffer vials (spills, air bubbles) and their instructions for use.
Stability: the storage temperature was ≤ 30 ° C for all the PDRs except one, the stability in use and of the result was maximum at 1 hour and 30 minutes, respectively. Integration into the healthcare environment requires a targeted product profile, an overview of the technology landscape, certified manufacturing capacity, a sustainable marketplace, and strict but timely regulation. Deployment in the country depends on integration into the national laboratory network.
Discussion / Conclusion:
Despite these limitations, successful implementation models have been proposed in triage, contact tracing, and surveillance, particularly for COVID-19 Ab RDT. There is valuable experience from the implementation of other disease-specific RDPs in sub-Saharan Africa.
Sample and sampling
COVID-19 Ag RDT
So far, nasopharyngeal secretions are the preferred sample for rapid diagnostic tests for COVID-19 Ag, as well as for reference NAAT tests. The equivalence of CORIS COVID-19 Ag Respi-Strip samples was evaluated in one study, demonstrating the equivalence of nasopharyngeal aspirates and nasopharyngeal swabs. In addition, the product’s instructions for use list nasopharyngeal washes as an eligible sample; however, this information was not supported by published evidence. The BIO CREDIT COVID-19 Ag Test was evaluated on saliva, nasopharyngeal swabs, nasopharyngeal aspirates, pharyngeal swabs, pharyngeal swabs, and sputum.
The EASY 2019-nCoV Rapid Antigen Test Kit has published an evaluation of nasal/nasopharyngeal swabs and oropharyngeal swabs as eligible specimens; The product’s instructions for use also list sputum as a sample with no published data. Given the reluctance of patients to take diagnostic samples in LRS, alternative samples (such as saliva) would be more acceptable than a nasopharyngeal swab or aspirate. COVID-19 has been detected in saliva samples collected by patients themselves using NAAT methods, but this needs further study.
For COVID-19 antigen detection, sample stability is a concern: in published studies on CORIS COVID-19 Ag Respi-Strip, BIOEASY 2019-nCoV Rapid Antigen Test Kit, and BIO CREDIT COVID-19 Test Ag, samples were stored at 4 ° C or −70 ° C when testing could not be performed immediately, indicating the need for a cold chain. The CORIS COVID-19 Ag Respi-Strip IFU indeed confirms the need to freeze at -20 ° C if immediate sample testing is not possible and mentions a loss of signal intensity when samples are stored at 4 ° C.
Consequently, the sample stability of COVID-19 Ag RDTs is a concern. For comparison, the WHO draft specifications for IVD POC COVID-19 that can be implemented in the classification list as a minimum (“acceptable”) requirement a pre-test sample stability of 30 min to 10- 35 ° C, 2-4 h at 2 at 8 ° C and 8 h in a generic preservative at 2-8 ° C.
To facilitate logistics and prevent patients from being lost to follow-up, the sample or sample and buffer mixture for the COVID-19 antigen test should be appropriate for subsequent NAAT testing (sufficient volume, assay-compatible RDT buffer NAAT, stability preserved, and contained in a leak-proof tube). In the publications on both COVID-19 Ag RDTs mentioned above, the same sample was used for the detection of NAAT and Ag, indicating the possibility of downstream NAAT.
COVID-19 Ab RDT
For COVID-19 Ab RDT, fingerstick capillary blood samples stand out as the preferred sample, as fingersticks are minimally invasive, safe, and easy to perform. Additionally, at SSA, healthcare workers and patients are familiar with fingerstick sampling, particularly in malaria-endemic areas. Probably explained by the use of stored (leftover) samples, published assessments of COVID-19 Ab PDRs were performed only on serum or plasma for half (29/54 products, 53.7%) of COVID-19 products. Ab; 19 (35.2%) with venous blood and five (9.3%) with capillary whole blood were also evaluated, all of them in a POC environment. Only eight COVID-19 Ab PDRs (14.8%) in four studies have published evidence on the equivalence of whole venous blood with serum or plasma.
In these studies, plasma was obtained by centrifugation of whole blood with EDTA and a concordance of more than 97% was found between both types of samples. Only one article (evaluating a single product) studied the equivalence of samples between plasma, whole venous blood and fingerstick blood and found no differences in the 10 paired samples (seven COVID-19 patients and three healthy controls) evaluated. Although venous whole blood and serum appear to be equivalent so far, further studies are needed to validate the equivalence of the sample, as serum and plasma are expected to have higher antibody titers compared to whole blood. None of the retrieved studies had evaluated the equivalence of different anticoagulants.
Sample type can affect the diagnostic performance of PDRs: for example, for HIV 1/2 PDRs, a higher number of false positives were shown in whole blood compared to plasma samples. Furthermore, the antibody concentration is higher in serum and plasma than in whole blood, which can lead to differences in sensitivity and specificity if the same volume is used. In their IFUs, all 20 COVID-19 Ab PDRs mentioned both serum and plasma. Two products had only plasma and serum listed as eligible samples in their IDUs, and two products indicated the use of serum, plasma, and whole blood, but specified that stab blood was not recommended.