A better way to test for COVID-19

Efficiency and economy are pivotal to successful disease control strategies. The use of reverse transcription polymerase chain reaction (RT-PCR) to test for COVID-19 has not delivered in terms of efficiency and mitigating the cost burden. From Asia to Europe and the Americas, medical workers fixated on using RT-PCR as the main workhorse for coronavirus detection have failed to measure up to the demands of pandemic testing, generating long lines of specimen backlogs and patients in distress waiting for RT-PCR results. Inexplicably, orthogonal algorithmic testing for COVID-19 has not gained much traction to date.

The RT-PCR’s analytical sensitivity (ability to detect a pathogen in a clinical specimen) is nearly 100 percent and attests to this test’s very high positive predictive value (PPV). However, its clinical sensitivity (ability to identify a patient’s infection status, reportedly ranging from 59-80 percent) detracts from its overall performance, making this test a poor choice for ruling out COVID-19. Paradoxically, most medical institutions worldwide are using this test precisely for this purpose.

Antibodies against SARS-CoV-2 have been detected in a study using enzyme-linked immunosorbent assay (Elisa, a low-cost immunoassay for antibodies distinct from rapid antibody test) in 75 percent, 94.7 percent, and 100 percent of patients on the first, second, and third week post-symptom onset by testing 535 plasma samples from 173 patients, while only 66.7 percent, 54 percent, and 45.5 percent over the same time periods, respectively, could be detected by RT-PCR in the same set of patients. This serologic trend post-symptom onset has been independently corroborated in another study of COVID-19 outpatients. Historically, Elisa can pin down target analytes in parts-per-billion or even in 10-100 parts-per-trillion concentrations, exhibiting a very high sensitivity and negative predictive value (NPV) ideally suited for ruling out non-COVID-19 cases. However, a positive Elisa test is only a presumptive evidence of COVID-19 illness, as it lacks enough specificity.

To optimize test outcomes, the World Health Organization recommends, among other strategies, an orthogonal testing algorithm where two independent tests, each with unique diagnostic platforms, are employed in tandem when the first test yields a positive result. Given current diagnostic tests for COVID-19, worth exploring is a double-layered test algorithm using a well-validated Elisa (high sensitivity, high NPV) as upfront screening test, and a SARS-CoV-2-specific RT-PCR (high specificity, high PPV) as the second test to verify/confirm only those positive by Elisa.

Theoretically, the value of a high sensitivity test is largely when the result is negative, while the value of a high specificity test is largely when the result is positive. Therefore, even without RT-PCR, COVID-19 can be ruled out by a negative Elisa test, assuming absence of symptom/history of possible/actual exposure, or by two negative Elisa tests 15 days apart under quarantine condition. It should not be the role of an RT-PCR test to rule out COVID-19 in previously untested patients, because Elisa, when used as an upfront test, is expected to do a much better job of detecting true negatives than RT-PCR. However, RT-PCR can definitively rule in COVID-19 when target RNAs are detected.

Using this orthogonal testing algorithm, the cost burden can be dramatically scaled down. In a population with 10-percent COVID-19 prevalence, cost savings are approximately equal to the cost of

RT-PCR tests intended for the 90-percent Elisa-negative subjects that no longer need RT-PCR verification. Further, both the sensitivity and specificity metrics for detecting COVID-19 are enhanced in this double-platform algorithm that circumvents the

zero-sum relationship between specificity and sensitivity inherent in single-platform tests. Likewise, ruling out negative cases, decongesting quarantine spaces, etc. are accomplished more quickly due to the shorter turnaround time of 2-3 hours for Elisa results.

Efficiency-wise, Elisa is scalable to high-throughput screening, with processing capacity of at least half a million serum samples monthly, assuming around 30 diagnostic laboratories are commissioned for the undertaking.

* * *

Faustino C. Icatlo Jr. ([email protected]) holds a Ph.D. degree from the University of Tokyo and is a consultant for diagnostics at the Research Institute for Tropical Medicine-Department of Health, where he is helping develop a COVID-19 Elisa system. The article reflects the author’s own views.