RESUMO
A prospective study of dengue infected patients at Dr.Soetomo Hospital pediatric ward was carried out from October 2008 to April 2009 to evaluate the revised dengue classification system proposed by the Dengue Control (DENCO), for early detection of severe dengue infected patients using the WHO classification system for comparison, with the addition of clinical interventions as a tool to grade for severity. One hundred forty-five patients were included in the study. Using the WHO classification system, 122 cases (84.1%) were classified as having non-severe dengue, of which 70 (48.3%) were classified as having dengue fever (DF), 39 (26.9%) as having dengue hemorrhagic fever (DHF) grade I, and 13 (9%) as having DHF grade II. Twenty-three (15.9%) were classified as having severe dengue, of which 16 (11%) were classified as having DHF grade III and 7 (4.8%) as having DHF grade IV. With clinical interventions included, 8 cases (6.6%) originally classified as having non-severe dengue infection were reclassified as having severe infection (sensitivity = 74%, specificity = 100%, likelihood ratio (-) = 0.26). Using the new dengue classification system, 117 cases (80.7%) were classified as having non-severe dengue infection, of which 79 (54.5%) were classified as having dengue without warning signs and 38 (26.2%) were classified as having dengue with warning signs, while 28 (19.3%) were classified as having severe dengue infection. Using clinical intervention, 4 cases (3.4%) which were originally classified as having non-severe dengue infection were reclassified as having severe dengue infection (sensitivity = 88%, specificity = 99%, likelihood ratio (+) = 98.88, likelihood ratio (-) = 0.13). Binary logistic regression showed the revised dengue classification system (p = 0.000, Wald:22.446) was better in detecting severe dengue infections than the WHO classification system (p = 0.175, Wald:6.339).
