A diagnostic decision-making protocol combines a new-generation of serological assay and PCR to fully resolve ambiguity in COVID-19 diagnosis

2020 
The capacity to accurately diagnosis COVID-19 is essential for effective public health measures to manage the ongoing global pandemic, yet no presently available diagnostic technologies or clinical protocols can achieve full positive predictive value (PPV) and negative predictive value (NPV) performance. Two factors prevent accurate diagnosis: the failure of sampling methods (e.g., 40% false negatives from PCR testing of nasopharyngeal swabs) and sampling-time-dependent failures reflecting individual humoral responses of patients (e.g., serological testing outside of the sero-positive stage). Here, we report development of a diagnostic protocol that achieves full PPV and NPV based on a cohort of 500 confirmed COVID-19 cases, and present several discoveries about the sero-conversion dynamics throughout the disease course of COVID-19. The fundamental enabling technology for our study and diagnostic protocol-termed SANE, for Symptom (dpo)-Antibody-Nucleic acid-Epidemiological history-is our development of a peptide-protein hybrid microarray (PPHM) for COVID-19. The peptides comprising PPHMCOVID-19 were selected based on clinical sample data, and give our technology the unique capacity to monitor a patient9s humoral response throughout the disease course. Among other assay-development related and clinically relevant findings, our use of PPHMCOVID-19 revealed that 5% of COVID-19 patients are from an "early sero-reversion" subpopulation, thus explaining many of the mis-diagnoses we found in our comparative testing using PCR, CLIA, and PPHMCOVID-19. Accordingly, the full SANE protocol incorporates orthogonal technologies to account for these patient variations, and successfully overcomes both the sampling method and sampling time limitations that have previously prevented doctors from achieving unambiguous, accurate diagnosis of COVID-19
    • Correction
    • Source
    • Cite
    • Save
    • Machine Reading By IdeaReader
    28
    References
    0
    Citations
    NaN
    KQI
    []