|| Checking for direct PDF access through Ovid
The aim of this study was to analyze the characteristics of Dengue virus (DENV)-infected children and the accuracy of dengue diagnosis based on clinical presentations. The inclusion criteria were children ≥1-year-old presenting febrile illness with 1–7 days of onset. Children (n = 110) aged 2–15 years were included in this study. DENV infection was confirmed with virological tests using serum, salvia, and/or urine samples. The attending pediatricians classified 56/110 (50.91%) of the children as suspected dengue cases. The DENV infection was confirmed by specific laboratory tests in 52/56 (92.9%) of the suspected dengue cases but also in 44/54 (81.5%) of the unsuspected dengue cases; total of 96/110 (87.27%) confirmed dengue cases. The clinical diagnosis gave an overall sensitivity of 54.2% (52/96) and a specificity of 71.4% (10/14). The positive predictive value of the clinical diagnosis was 92.8% and negative predictive value was 18.5%. After the third day of onset of symptoms, the DENV genome detection rate was similar in serum and saliva samples, suggesting that saliva samples represent an alternative to blood samples for early dengue diagnosis. Vaccination against Yellow fever virus did not influence the antibody response against DENV-1, DENV-2, and DENV-3, which circulated during the study period. Although the signs and symptoms were compatible with dengue, the attending pediatricians did not suspect the disease in several children. Therefore, the inclusion of virological tests for early diagnosis in the protocols for dengue surveillance would help in the implementation of prompt treatment of patients and epidemic containment strategies.