Accuracy and Usability of a New Rapid Diagnostic Test for the Diagnosis of Dengue in Vientiane, Laos.
Phimolsannousith V., Sibounheuang B., Chanthongthip A., Vongsouvath M., Mayxay M., Mukaka M., Luangraj M., Newton PN., Ashley EA., Hopkins H., Dubot-Pérès A.
BACKGROUND: An estimated 390 million dengue virus (DENV) infections occur per year, 70% in the Western Pacific and Southeast Asia. A prototype rapid diagnostic test (RDT) (NS1, immunoglobulin M [IgM]/immunoglobulin G [IgG]; DengueDx, Mologic) was evaluated for accuracy and ease of use for dengue diagnosis at Mahosot Hospital, Vientiane, Lao People's Democratic Republic. METHODS: Five hundred twenty-six sera collected in 2017 and 2018 from patients hospitalized with suspected dengue were included. Prototype RDT results were compared to NS1 and IgM enzyme-linked immunosorbent assays and/or real-time reverse-transcription polymerase chain reaction as reference assays and to Bioline dengue RDT (NS1, IgM/IgG) as comparator. The accuracy of the image-based result interpretation (RDT scan or photograph) at day 0 and after a delay of 7 days was evaluated. Five technicians were observed when performing RDTs and interviewed on their perception of the ease of use. RESULTS: The overall sensitivity and the sensitivity for NS1 detection alone were significantly lower for the prototype RDT (65.2% [95% confidence interval {CI}, 59.3%-70.8%] and 44.6% [95% CI, 38.0%-51.4%], respectively) than for the Bioline RDT (76.7% [95% CI, 69.1%-83.2%], z-test P = .014; and 66.7% [95% CI, 57.5%-75.0%], z-test P < .001, respectively). The positive predictive value of the prototype RDT for detection of NS1 was >90%. Only half of the positive results were found positive when reading results from images. Correlation of >98% for positive results was obtained between reading at day 0 and reading at day 7. During ease of use assessment, when performing the Bioline RDTs the participants made fewer errors and were more satisfied than when performing the prototype RDTs. CONCLUSIONS: Improvement in test sensitivity and in device design would adapt the prototype RDT for Laos and similar contexts. Optimization of the prototype RDT is warranted and further evaluation in different healthcare settings would provide a more accurate estimation of the value of this RDT for dengue surveillance in remote areas.