Résumé
A prospective study was carried out to evaluate the sensitivity of dengue NS1 antigen-capture ELISA in comparison with dengue virus isolation, conventional RT-PCR and real-time RT-PCR for laboratory confi rmation of acute dengue based on single-acute serum samples. Four primary healthcare centres were involved to recruit patients with clinical diagnosis of dengue illness. Patient’s demographic, epidemiological and clinical information were collected on a standardized data entry form and 5 ml of venous blood was collected upon consent. In the laboratory, six types of laboratory tests were performed on each of the collected acute serum sample. Of the 558 acute serum samples collected from 558 patients with clinical diagnosis of dengue from mid-August 2006 to March 2009, 174 serum samples were tested positive by the dengue NS1 antigen-capture ELISA, 77 by virus isolation, 92 by RT-PCR and 112 by real-time RT-PCR. A total of 190 serum samples were tested positive by either one or a combination of the four methods whereas, only 59 serum samples were tested positive by all four methods. Thus, based on singleacute serum samples, 190 of the 558 patients (34.1%) were laboratory-confi rmed acute dengue. The overall test sensitivity was 91.6%, 40.5%, 48.4% and 58.9% for dengue NS1 antigen-capture ELISA, virus isolation, conventional RT-PCR and real-time RT-PCR respectively. Statistically, dengue NS1 antigen-capture ELISA was the most sensitive and virus isolation was the least sensitive test for the laboratory confi rmation of acute dengue based on single-acute serum specimens. Real-time RT-PCR was signifi cantly more sensitive than the conventional RT-PCR.