Clinical Spectrum and Predictors of Severe Dengue Among Hospitalized Children in a Tertiary Care Hospital in Bangladesh.

Introduction Dengue is a major public health concern in tropical regions, with children at increased risk of severe disease. Early identification of predictors is essential for timely intervention. Methods This hospital-based cross-sectional study was conducted from January to December 2024 at a tertiary care hospital in Dhaka, Bangladesh. A total of 346 children aged 0-14 years with confirmed dengue infection were included. Clinical, laboratory, radiological, and comorbidity data were analyzed. Variables associated with severe dengue were identified using the chi-square test and multivariate logistic regressio. Results Severe dengue was observed in 53 (15.3%) cases, with one mortality (0.3%). Common clinical features included fever in 346 (100%), vomiting in 97 (28.0%), abdominal pain in 73 (21.1%), and diarrhea in 59 (17.1%) patients. On univariate analysis, several clinical, laboratory, and radiological variables were significantly associated with severe dengue. Multivariate analysis identified elevated hematocrit (adjusted odds ratio (AOR): 9.2) as the strongest predictor, followed by pleural effusion (AOR: 7.4), ascites (AOR: 6.3), thrombocytopenia (AOR: 5.6), hypoalbuminemia (AOR: 4.8), elevated aspartate aminotransferase (AST, AOR: 4.2), and elevated alanine transaminase (ALT, AOR: 3.5).  Conclusion Severe dengue in children is primarily associated with markers of plasma leakage and organ involvement. Early risk stratification using laboratory and radiological parameters can improve clinical outcomes.