Comparative performance study of three Ebola rapid diagnostic tests in Guinea

Background: The 2014/15 Ebola outbreak in West Africa resulted in 11,000 deaths and massive strain on local health systems, and the ongoing outbreak in Democratic Republic of Congo has afflicted more than 3000 people. Accurate, rapid Ebola diagnostics suitable for field deployment would enable prompt identification and effective response to future outbreaks, yet remain largely unavailable. The purpose of this study was to assess the accuracy of three novel rapid diagnostic tests (RDTs): an Ebola, an Ebola-Malaria, and a Fever Panel test that includes Ebola, all from a single manufacturer. Methods: We evaluated the three RDTs in 109 Ebola-positive and 96 Ebola-negative stored serum samples collected during the outbreak in Guinea in 2014/15, and tested by real-time polymerase chain reaction (RT-PCR). Sensitivity, specificity, and overall percent agreement were calculated for each RDT using RT-PCR as a reference standard, stratified by Ct value ranges. Results: All tests performed with high accuracy on samples with low Ct value (high viral load). The Fever Panel test performed with the highest accuracy, with a sensitivity of 89.9% and specificity of 90.6%. The Ebola and Ebola-Malaria tests performed comparably to each other: sensitivity was 77.1 and 78% respectively, and specificity was 91.7% for the Ebola test and 95.8% for the Ebola-Malaria test. Conclusions: This study evaluated the accuracy of three novel rapid diagnostic tests for Ebola. The tests may have significant public health relevance, particularly the Fever Panel test, which detects seven pathogens including Ebola. Given limitations to the study resulting from uncertain sample quality, further evaluation is warranted. All tests performed with highest accuracy on samples with low Ct value (high viral load), and the data presented here suggests that these RDTs may be useful for point-of-care diagnosis of cases in the context of an outbreak. Restrictions to their use in non-severe Ebola cases or for longitudinal monitoring, when viral loads are lower, may be appropriate. Highlighting the challenge in developing and evaluating Ebola RDTs, there were concerns regarding sample integrity and reference testing, and there is a need for additional research to validate these assays.