Diagnostic performance of LumiraDx SARS-CoV-2 rapid antigen test compared to PCR for diagnosis of COVID-19 in Liberia

Abstract

BackgroundDiagnosis of SARS-CoV-2 infection in Liberia has primarily relied on polymerase chain reaction (PCR)-based testing at the country’s National Reference Laboratory. This centralized approach caused reporting delays, prompting evaluation of point-of-care antigen-based tests. We assessed the test performance of the LumiraDx™ SARS-CoV-2 Ag test (LumiraDx™, London, UK) in this setting.MethodsWe tested ambulatory individuals screened for enrollment into an observational cohort study of COVID-19 sequelae in Monrovia in 2021. We compared the results of LumiraDx testing of anterior nasal swab specimens to the results of PCR BioFire® R2.1P (bioMérieux, Salt Lake City, Utah) on eluent from nasopharyngeal swabs.ResultsWe evaluated 348 individuals. Among the 274 persons with symptoms, 49.3% were PCR-positive and 36.5% were antigen-positive. The sensitivity, specificity, negative predictive value (NPV), and positive predictive value (PPV) of LumiraDx in this group were 72.6% (95% CI: 64.3%-79.9%), 98.6% (95% CI: 94.9%-99.8%), 78.7% (CI: 71.9%-84.6%), and 98.0% (CI: 93.0%-99.8%), respectively. Among the 74 asymptomatic individuals, 12.2% were positive by PCR, and 5.5% by antigen testing, resulting in a sensitivity, specificity, NPV, and PPV of LumiraDx of 44.4% (95% CI: 13.7%-78.8%), 100% (95% CI: 94.5%-100%), 92.9% (CI: 84.1%-97.6%), and 100% (CI: 39.8%-100%), respectively.ConclusionAlthough the specificity and PPV of LumiraDx for diagnosing SARS-CoV-2 were high among persons with and without symptoms, the sensitivity was unacceptably low in both groups, and much less than that reported by the manufacturer. Before new point-of-care diagnostics are adopted, test performance needs to be assessed in the local setting.