Analyzing Conflicting Results in Rapid Point-of-Care COVID-19 Testing

Abstract

Rapid point-of-care (POC) testing for the SARS-CoV-2 virus provides an ability to quickly test individuals, facilitating near real-time action if infection is determined, such as isolation and initiation of contact tracing. Rapid POC testing also facilitates more frequent testing in individuals, which is crucial for screening programs and for environments where as up-to-date results are essential, for example to facilitate safe entrance into nursing homes. The Abbott ID NOW is a rapid POC test that delivers an accurate result in 5-13 minutes and that has demonstrated a high level of performance in variety of patient populations. However, recent anomalous results by Basu et al. have raised important questions about the sensitivity of the Abbott test. Understanding the validity of these anomalous results cannot be overstated. The ID NOW is currently used widely as a mainstay of many COVID-19 testing programs: urgent-care clinics, hospitals, businesses and other institutions rely on the test to provide fast and accurate results every day. Here, we assess the anomalous findings and find significant limitations in the study design, comparisons used, and patients evaluated. We describe how these limitations can account for essentially all of the losses in sensitivity described by Basu et al. We also discuss interim results from an ongoing multi-site clinical study in urgent-care clinics that indicate that the ID NOW COVID-19 rapid test is highly sensitive (≥94.7%) and specific (≥98.6%). Rapid POC testing in outpatient clinical settings and acute care facilities offers the ability to quickly diagnose and isolate infected patients and enhance safeguards against virus transmission. It is important that clinicians received accurate information about rates of sensitivity and specificity in order to feel confident using current POC rapid-testing systems.