Background: In 2014, enterovirus D68 (EV-D68) was responsible for an outbreak of severe respiratory illness in children, with 1,153 EV-D68 cases reported across 49 states. Despite this, there is no commercial assay for its detection in routine clinical care. BioFire® Syndromic Trends (Trend) is an epidemiological network that collects, in near real-time, deidentified. BioFire test results worldwide, including data from the BioFire® Respiratory Panel (RP).
Objectives: Using the RP version 1.7 (which was not explicitly designed to differentiate EV-D68 from other picornaviruses), we formulate a model, Pathogen Extended Resolution (PER), to distinguish EV-D68 from other human rhinoviruses/enteroviruses (RV/EV) tested for in the panel. Using PER in conjunction with Trend, we survey for historical evidence of EVD68 positivity and demonstrate a method for prospective real-time outbreak monitoring within the network.
Study design: PER incorporates real-time polymerase chain reaction metrics from the RPRV/EV assays. Six institutions in the United States and Europe contributed to the model creation, providing data from 1,619 samples spanning two years, confirmed by EV-D68 gold-standard molecular methods. We estimate outbreak periods by applying PER to over 600,000 historical Trend RP tests since 2014. Additionally, we used PER as a prospective monitoring tool during the 2018 outbreak.
Results: The final PER algorithm demonstrated an overall sensitivity and specificity of 87.1% and 86.1%, respectively, among the gold-standard dataset. During the 2018 outbreak monitoring period, PER alerted the research network of EV-D68 emergence in July. One of the first sites to experience a significant increase, Nationwide Children's Hospital, confirmed the outbreak and implemented EV-D68 testing at the institution in response. Applying PER to the historical Trend dataset to determine rates among RP tests, we find three potential outbreaks with predicted regional EV-D68 rates as high as 37% in 2014, 16% in 2016, and 29% in 2018.
Conclusions: Using PER within the Trend network was shown to both accurately predict outbreaks of EV-D68 and to provide timely notifications of its circulation to participating clinical laboratories.
Keywords: Enterovirus D-68; Epidemiology; Machine learning.
Published by Elsevier B.V.